Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Biomass and root attributes of eight of New Zealand’s most common indigenous evergreen conifer and broadleaved forest species during the first 5years of establishment

Biomass and root attributes of eight of New Zealand’s most common indigenous evergreen conifer... Background: Tree allometric equations are critical tools for determining tree volume, biomass and carbon stocks. However, there is a lack of species-specific biomass equations for juvenile trees of many of New Zealand’sindigenous species. The aim of this study was to provide allometric equations for total above- and below-ground biomass and total root biomass and length for eight common evergreen conifer and broadleaved species. Methods: In a plot-based field trial, growth metrics of conifers Prumnopitys taxifolia (matai), Agathis australis (kauri), Prumnopitys ferruginea (miro), Podocarpus totara (totara), Dacrycarpus dacrydioides (kahikatea) and Dacrydium cupressinum (rimu) and broadleaved species Alectryon excelsus (titoki) and Vitex lucens (puriri) were measured annually. These species were selected based on their potential role as a long-term solution for mitigating erosion in areas of marginal land proposed for new afforestation/reforestation and as an important carbon (C) sink. Results: Root collar diameter (RCD) provided the best fit for tree height, total above-ground biomass (AGB) and total below-ground biomass (BGB), and all regressions were highly significant (P = 0.001). Most species showed significant increases in annual growth and, by year 5, the BGB ranged between 21 and 42% of total biomass and decreased with increasing plant age. Of the conifers, Podocarpus totara had the greatest mean maximum root spread (2.2 m) exceeded only by the broadleaved Vitex lucens (2.5m).Forallspecies,andin each year of the trial, 100% of the BGB remained confined to within 0.5 m of the ground surface. With the exception of Vitex lucens and Podocarpus totara, > 90% of the total root length remained within a 0.5-m radius of the root bole. The species-specific mean tree biomass of 5-year-old plants ranged from 0.32 to −1 4.28 kg plant . A mixed-species forest established at 1000 stems per hectare (spha), consisting of 200 of −1 each of the best performed of the trialled species, would amass ~ 2.3 t ha of biomass and a forest carbon −1 stock of 3.8 t CO ha within 5 years. Conclusions: Inter-species differences in the allocation of BGB and AGB appeared to be age dependent. The root-growth metrics of these common indigenous forest species, as candidates for erosion control, have improved our understanding of their potential usefulness for stabilising marginal land. Whole-plant biomass of juvenile trees will greatly improve the accuracy of current estimates of forest carbon stocks for proposed new areas of indigenous afforestation/reforestation. Keywords: Biomass, Root systems, Conifer and broadleaved species, 5-year plot-based trial * Correspondence: mardenm@landcareresearch.co.nz Landcare Research, NZ, Ltd, PO Box 445, Gisborne 4010, New Zealand Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 2 of 26 Background Wardle 2002). The resultant forests typically comprise Before human settlement, New Zealand was almost long-lived conifers and broadleaved species that can completely vegetated with indigenous evergreen podo- achieve large stature and an active growth phase that ex- carp (mainly conifer) and broadleaved forest in the low- tends from 150 to 500 years (Hall 2001, Hinds and Reid lands (~ 200 m), transitioning to Nothofagus forest at 1957). As suggested by Hasselmann (1997), active net elevations above 600 m to the tree line at 1050 m and carbon accumulation over such time frames is consistent alpine-subalpine shrubland and grasslands on the high- with the prolonged effort likely required to effect signifi- est parts of the axial ranges (McGlone 1988; Wilmshurst cant reductions in atmospheric carbon dioxide (CO ) 1997). However, since Maori (beginning ~ 600 years BP) levels. and then European settlement (starting in the 1820s), Given the pressing need to quantify carbon fluxes as- much of New Zealand’s indigenous forest has been sociated with terrestrial vegetation dynamics, an increas- cleared for conversion to pastoral use, for the develop- ing number of researchers have sought to improve ment of a forest industry involving non-indigenous spe- estimates of tree volume, biomass and carbon stocks. cies and/or for urban development. By the time the first Many New Zealand-based studies have attempted to National Forest Survey began in 1945, only 25% of indi- quantify the biomass of indigenous forest stands, mainly genous forest remained on a total land area of 26.6 mil- for live-tree carbon sequestration (Carswell et al. 2012; lion hectares, mostly in mountainous (difficult to farm) Scott et al. 2000; Trotter et al. 2005; Beets et al. 2014; areas (Masters et al. 1957). In more recent years, the on- Schwendenmann and Mitchell 2014; Dale 2013 ). These going clearance of vegetation has led to the progressive use allometric functions based largely on figures derived loss of biodiversity and degradation of waterways and for individual specimens of a single species such as water quality through increased erosion, sedimentation kānuka (Kunzea ericoides var. ericoides (A. Rich.) J. and nutrient pollution. Thompson), and, where available, published values for a An increase in awareness of the poor health of New limited number of other species (Russo et al. 2010). The Zealand’s remaining areas of indigenous forest (as well few available biomass relationships are for a handful of as public and government pressure to maintain and en- old-growth indigenous tree species from a subset of New hance the indigenous biodiversity of New Zealand) (Par- Zealand forest classes and from a limited number of sites 3 4 liamentary Commissioner for the Environment 2002) (Beets 1980; Allen et al. 1998 ; Peltzer and Payton 2006 ; has encouraged reversion of cleared land back to forests. Beets et al. 2008 ). Thus, where species-specific and/or The planting of indigenous trees has become an integral regional values remain unavailable, congeneric values part of indigenous-forest restoration efforts. Further- have been used instead or in their absence, the mean more, since ratifying the Kyoto Protocol (Intergovern- of all published values (Beets et al. 2008). Further- mental Panel on Climate Change (IPCC) 2000; New more, the majority of these studies were of mature Zealand Climate Change Office 2003), New Zealand pro- trees though the age of the sample trees was not posed to reduce its carbon emission target in the short always determined. to medium term through afforestation/reforestation. Tree allometric equations are critical tools for deter- Exotic forests established on former grasslands since mining tree volume, biomass and carbon stocks and 1990 are likely to provide the major carbon offsets, but have the potential to improve our understanding about it has been recognised that an important additional car- carbon sequestration in woody vegetation to support the bon sink could be created through afforestation/refores- implementation of policies and mechanisms designed to tation of steep, erosion-prone pastoral hill country that mitigate climate change (Jara et al. 2014). In New environmentally is marginal for long-term agriculture by Zealand, substantial gains have been made in improving using indigenous shrubs and trees (Tate et al. 2000; Tate the accuracy of the above-ground live tree carbon stored et al. 2003, Trotter et al. 2005). Relatively little is known in the most widely distributed forest and shrubland about carbon sequestration by indigenous shrub species, ecosystems following the establishment of a nationwide but the density and diversity of species typical of natur- network of permanent sample plots, (Carswell et al. 2012; ally reverting stands suggests that relatively high levels Scott et al. 2000; Trotter et al. 2005; Peltzer and Payton of carbon storage are achievable. A range of additional 2006; Beets et al. 2014). Nonetheless, as identified by Pelt- objectives in sustainable environmental management will zer and Payton (2006), there remains a growing need for also be achievable. These include improving indigenous further age-specific and species-specific allometric rela- biodiversity, erosion mitigation and soil conservation tionships particularly for the 20 most abundant species and consequent improvements in water quality. Import- that comprise 90% of the total live-tree carbon in antly, from a carbon-sink perspective, indigenous shrubs old-growth forests. However, the below-ground biomass provide the first step to a successional pathway to a per- component has proven difficult to quantify for old-growth manent cover of indigenous tall forest (Hall 2001; trees. In particular, little is known about root allocation Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 3 of 26 except that there are interspecies differences in root distri- damage to infrastructure and the degradation of stream bution. Each species is known to allocate differing propor- habitats. However, more fundamental knowledge is re- tions of their total biomass to roots at different stages of quired on the differences in relative growth rates be- growth, but overall, species-specific allocations of dry tween indigenous shrubland, conifer and broadleaved matter to roots are generally poorly documented (Korner species and more specifically on the contribution of their 1998). Other than destructive sampling, there are few root systems towards shallow soil reinforcement, espe- techniques with which to assess either the below-ground cially during their formative years. Such data will assist biomass component or the architectural differences in in evaluating the potential effectiveness of various root system attributes between species. The use of species both as plantings on marginal land degraded by non-invasive techniques such as ground-penetrating radar erosion and as a longer-term land-use option for areas (GPR) is limited to coarse-root systems (Hruska et al. retired from exotic forest production. As is the case for 1999). The fine-root components are particularly difficult old-growth indigenous forests, such data for the and expensive to measure accurately. Several New early-growth period of the majority of indigenous shrub Zealand and international studies have shown that roots and forest species have not been collected to date. Such comprise approximately 20% of the above-ground live tree data are essential for the development of allometric biomass (Watson and Tombleson 2002; Beets et al. 2007). relationships for measuring the live-tree carbon seques- The precision of these root estimates, however, tend to be tration for: (i) planted areas of juvenile shrubs and low (Hall et al. 2001) but have nonetheless been used in mixed species of indigenous conifer and broadleaved the absence of better species- and age-specific data. Fur- forests; (ii) areas of abandoned pasture currently in the thermore, studies that have attempted to extract the ma- juvenile stage of reverting to shrubland; and (iii) areas of jority of the below-ground root biomass often exclude the exotic forest with ‘high environmental risk’ and likely to stump (Will 1966; Heth and Donald 1978; Watson and be converted (planted or by natural reversion) to a O’Loughlin 1990) and therefore underestimate the longer-term, and more sustainable, mix of indigenous below-ground to above-ground biomass ratio. Previous shrubland and conifer/broadleaved forest. authors (e.g. Cairns et al. 1997;Coomes et al. 2002)have This study is the first known attempt to systematically suggested a value of 20% can be applied generally, but quantify, as a time series, both the above- and more work is needed before this figure can be accepted below-ground growth metrics for eight of New Zealand’s internationally. most common indigenous conifer and broadleaved forest Nationally, about 1.45 million hectares (Mha) of mar- species during their early growth period (years 1 to 5). ginal pastoral land are suitable for afforestation/refores- The goals of this study were to: (i) develop species-specific tation by indigenous shrubs or forests (Trotter et al. allometric equations to better estimate above- and 2005). The enactment of the Permanent Forest Sink Ini- below-ground live-plant biomass/carbon storage for areas tiative (PFSI) and Afforestation Grant Schemes (AGS) in of newly planted marginal land established as a pathway 2008 and 2015, respectively (Ministry for Primary Indus- to a permanent indigenous forest and long-term carbon tries 2015), and the introduction of forestry to the Emis- sink; and (ii) provide the first-ever description and classifi- sions Trading Scheme (ETS) have strongly incentivised cation of their root morphology and architecture and the planting in new areas of indigenous shrublands measurements of their root metrics (biomass, length and and/or successional tall forest species as a means of fur- the spatial distribution of roots with increasing distance ther reducing greenhouse gas emissions and ensuring from and below the root bole). Such data form the basis that New Zealand meets both its short- and long-term for assessing their potential effectiveness as an alternative obligations under the Framework Convention for land-use option in areas degraded by erosion and deemed Climate Change (FCCC). Given the current interest in marginally sustainable for pastoralism and/or for areas the retirement of economically and environmentally un- currently unsustainable as short-term rotation exotic sustainable pastoral steep lands, the planting of succes- forest. sional indigenous forest species with a longer rotation This paper does not address: (i) the effect of contain- and higher wood value may become an alternative and erised versus bare-rooted seedlings on plant growth; nor viable land use option for these areas. Furthermore, a (ii) the impacts of the wider range of ecological and en- species change to indigenous shrubs and/or forest trees vironmental stresses known to influence plant develop- could potentially alleviate issues currently associated ment (e.g. competition for nutrients and water, soil with harvesting non-native species (predominantly Pinus texture and fertility; water-logging; drought; elevation; radiata D.Don) that were originally established for climate regime; exposure to wind and sun; shade; and erosion control, particularly in terrain with a high risk of grass suppression) since it is a site-specific study; nor storm-induced landslides. Issues include excessive (iii) silviculture and management practices other than amounts of sediment and associated woody debris, planting density requirements needed to achieve effective Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 4 of 26 mitigation of erosion. Information on ecological and site and in block three (plants to be extracted 3 years after requirements and growth parameters of mature trees of planting, i.e. at age 5) at 1.5–2.0-m spacing to allow for the selected species are provided in Additional file 1 and increases in canopy and root growth during the course has been sourced from previous research by Foweraker of the trial. (1929), McSweeney (1982), Pollock (1986), Salmon (1986), Poole and Adams (1994) and Wardle (2002). The ecology, Species selection establishment, growth and management of the studied The species chosen (Table 1) include those that succeed species for wood production is adequately discussed by early colonising species to become a significant compo- Bergin (2003), Bergin and Steward (2004), Bergin and Gea nent of areas of mature coniferous/broadleaf forests (2005, 2007) and Steward and Beveridge (2010). found in most regions throughout the North Island of New Zealand (Allan Herbarium 2000). The conifers Methods (softwoods) selected included rimu (Dacrydium cupressi- Study site num (Lambert)), totara (Podocarpus totara (D.Don)), The trial site was located on a low-lying, even-surfaced miro (Prumnopitys ferruginea (D.Don) Laubenf.) and alluvial terrace adjacent to the Taraheru River, in kahikatea (Dacrycarpus dacrydioides (A.Rich.) Laubenf.) Gisborne City (Fig. 1). The same site has also been used because they are among the most abundant species that to measure ‘plant growth performance’ of: (i) 12 early collectively contribute 90% of New Zealand’s total (colonising) indigenous species considered typical of ri- live-plant carbon by volume (Peltzer and Payton 2006). parian margins (Marden et al. 2005); (ii) different clones Together with matai (Prumnopitys taxifolia (D.Don) of poplar and willow (Phillips et al. 2014); and (iii) a Laubenf.), kauri (Agathis australis (D.Don) Lindley) and range of exotic forest species (Phillips et al. 2015). Tem- the broadleaved species titoki (Alectryon excelsus (Gaert- peratures over summer average 23 °C and over winter ner)) and puriri (Vitex lucens (Kirk)), these are among 12 °C, and mean annual rainfall is 1000 mm. The soil is the most common species planted in landscape restor- a naturally fertile, free-draining Typic Sandy Brown Soil ation projects (Bergin and Gea 2005, 2007). These spe- of the Te Hapara soil series (Hewitt 2010). Although not cies are considered as potential alternatives to the used, a drip-irrigation was installed to ensure survival of non-native species currently used in commercial planted the plants in the event that drought could jeopardise the forests to meet a range of economic, erosion-mitigation longer-term aims of this trial. and other environmental benefits. Advantages include: The soil has no physical or chemical impediments to tolerance to a wide range of climate, soil and topo- root development to about 1.2 m depth, other than a graphic site conditions; seedlings are widely available; variable-depth water table that fluctuates between > and their longer growing cycles are considered to be 1.5 m depth and within ~ 0.2 m of the surface (Phillips more sustainable than faster-growing non-native et al. 2014). species. Trial design Data collection The site (50 × 20 m) was tilled, and weed mat was laid A group of 1- and 2-year-old plants were removed from down before planting in autumn 2006 (Fig. 1). Weed their containers and destructively partitioned into their mat reduced weed competition and minimised efforts to component parts. At the trial site, another group of extricate intertwined roots belonging to weeds from the 2-year-old seedlings were planted and a subset of plants anticipated dense and fibrous root network of the was excavated and similarly destructively partitioned on trialled species. the anniversary of planting for three successive years. An Containerised seedlings sourced from a local nursery air lance at 240 kPa was used to remove the soil sur- consisted of 1.2-L polythene bags (0.12 m deep by rounding each root system without damage or signifi- 0.10 m diameter) filled with a 50:50 mix of bark and cant loss of the fine root mass. The aim was to sample pumice. To minimise root disturbance, seedlings were ten plants per species per year, but frost killed some pur- planted with the soil attached. Root binding was minimal iri plants and stem breakage reduced the sample size of and root pruning before planting was not required as titoki plants (Table 1). tap root length did not exceed the depth of the planting Tree height, canopy spread, root collar diameter bag. The site was subdivided into three blocks and all (RCD—measured overbark at ground level) and diameter three blocks were planted in 1 day. A randomised block at breast height (DBH—measured at 1.4 m above-ground design was used with equal numbers of each species in level) (where applicable) were measured before removing each block. Trees within blocks 1 and 2 (plants to be ex- the plants from the trial site (Additional file 2). Canopy tracted 1 and 2 years after planting, i.e. at age 3 and spread was taken as the mean of the diameter mea- 4 years old, respectively) were planted at 1-m spacing, sured in two directions at right angles. Trees were Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 5 of 26 Fig. 1 Location and layout of the trial site cut at ground level and the above-ground of the maximum root length measured from root tip to components separated into branches, foliage and stem root tip in two directions at right angles to each other. (Additional file 3). Root system extraction and meas- Root length (excluding fibrous roots) within each of the urement methods followed well-established procedures root diameter size classes was measured by laying pieces (e.g. Watson et al. 1999;Mardenetal. 2005; Phillips of root end for end alongside a tape measure (Watson et al. 2014, 2015). Once removed from the ground, et al. 1999). The mass of the roots and root bole were the root system of each plant was washed to remove calculated separately (Additional file 5). Root distribution adhering soil matter then photographed (an example by mass (Additional file 6) and length (Additional file 7) is shown in Fig. 2). It was then described using the relative to the root bole was obtained by systematically classification of Hinds and Reid (1957) before being dissecting the root systems into 0.5-m radial and depth partitioned into root bole (stump) and roots. segments then sorted by diameter size classes, (< 1 mm Below-ground growth parameters included mean root fibrous, 1–2, 2.1–5.0, 5.1–10.0 and 10.1–20.0 mm over depth and mean root spread of the lateral roots, here- bark) (Watson and O’Loughlin 1990). after referred to as root depth and root spread All plant components, both above- and below-ground, (Additional file 4). The latter was taken as the mean were oven-dried at 80 °C for 24 h or until no further Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 6 of 26 Table 1 Indigenous conifers and broadleaved species trialled and number of trees extracted each year of the trial Botanical name Common Root Number of plants extracted/species/year name system Plant age (years) 1 2 3 4 5 Species total type Coniferous species Agathis australis Kauri Tap 10 10 10 10 10 50 Dacrycarpus dacrydioides Kahikatea Tap 10 10 10 10 10 50 Dacrydium cupressinum Rimu Tap 10 10 10 10 10 50 Podocarpus totara Totara Tap 10 10 10 10 10 50 Prumnopitys ferruginea Miro Plate 10 10 10 10 10 50 Prumnopitys taxifolia Matai Plate 10 10 10 10 10 50 Broadleaved species Alectryon excelsus Titoki Plate 10 10 10 10 9 49 Vitex lucens Puriri Tap 10 10 10 6 5 41 Annual total 80 80 80 76 74 390 weight loss was detectable then weighed to the nearest Statistical analyses 0.1 g. The biomass and root:shoot ratio were calculated ANOVA was used to determine the effects of year and using dry mass. A carbon content of 50% of biomass is species on above- and below-ground parameters. The considered robust (Coomes et al. 2002). A conversion Student-Newman-Keuls post hoc analysis was used to de- factor of biomass × 1.65 (Williams 1978) was used to cal- termine differences among the species within a year and culate carbon stocks (t CO ). Carbon stock values do among years when all species were averaged within each not include woody debris or fine litter. year. The normality of each analysis was determined by Fig. 2 Vertical and plan views of 5-year-old Agathis australis (a) and Alectryon excelsus (b) root systems Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 7 of 26 visual assessment of residual plots. The root biomass and curves for roots with > 1- and > 2-mm diameter (P = root length data at 0.5-m radial intervals exhibited 0.741). Therefore, regressions based on roots > 1-mm non-normality due largely to the prevalence of radial seg- diameter were used for analyses of total root length and ments in which no or little root material was recorded for total root biomass. Unless otherwise stated, data are pre- some of the root diameter size classes. The radially seg- sented as a mean with standard errors, and statistical ana- mented root biomass and length data were therefore lyses were considered significant if P ≤ 0.05. transformed using the equation log (x + c), where x was Differences between the grouped conifers versus the root biomass or root length and c was 1. broadleaved species were not examined because they dif- Type-1 linear regression was used to fit RCD (over- fer significantly in tree form; there were clear overlaps in bark) and tree height data using the equation (y = mx + above- and below-ground metrics between species c). RCD was used instead of DBH in the analysis of tree across these groupings and because of the likely bias in allometry because many of the species were not ex- analyses resulting from too few and/or uneven sample pected to reach DBH height (1.4 m) for some years after numbers. the establishment of the trial. More importantly, it may serve as a better predictor of both above-ground and Results below-ground biomass quantities as root collar is mea- Allometric relationships sured at the intersection of above- and below-ground Tree height parameters and as stem taper in seedlings is minimal Linear regression fitted the RCD and tree height data well, (Wagner and Ter-Mikaelian 1999). The relationship be- with r ranging between 0.768 (Vitex lucens) and 0.916 tween RCD and foliage, branches and stem (Table 2), (Dacrydium cupressinum), and all regressions were highly total above-ground biomass (AGB), total below-ground significant (P < 0.001; Fig. 3). There were significant biomass (BGB), total root biomass (> 1- and > 2-mm differences among species with respect to height in each diameter) and total root length (> 1- and > 2-mm diam- year (Additional file 2). With the exception of eter) (Table 3) was analysed using two-parameter expo- Prumnopitus taxifolia, which had the same average height nential growth analysis using the equation y = a exp (bx), in years 1 and 2 (P = 0.05: data not shown), the tree height where b represented the slope of the regressions. The re- of the remainder of species increased significantly from gression analyses include data for the containerised 1- year to year. There was, however, no clear pattern as to and 2-year-old seedlings partitioned at the time the trial which species had performed best until year 5 when site was planted. All regression analyses were undertaken Prumnopitys taxifolia (1.3 m), Agathis australis (1.6 m) using SigmaPlot 12.5 (Systat Software, San Jose, CA). and Prumnopitys ferruginea (1.6 m) were significantly Depending on the method of root extraction and focus shorter (P = 0.05) than the tallest two species Podocarpus of the research, regressions of total root biomass and totara (2.8 m) and Dacrycarpus dacrydioides (2.7 m), length are often based on different groupings of root while Vitex lucens (2.4 m), Dacrydium cupressinum diameter size classes. For comparative New Zealand stud- (2.2 m) and Alectryon excelsus (2.2 m) reached an ies, some researchers have based their regressions on roots intermediary height significantly different from the other with a > 1-mm diameter (Phillips et al. 2015)while others species but not each other (Additional file 2). have included only roots with a > 2-mm diameter (Watson and Tombleson 2002, 2004;Marden et al. 2016). In this Root collar diameter (RCD) study, when regressed against RCD, a two-tailed, unpaired Differences in RCD among species in years 1–3 were t test found no significant difference in the slopes of the variable, with no consistent species-specific trends other Table 2 Allometric relationships between RCD against stem, branch and foliage biomass Species Stem Branch Foliage 2 2 2 r P Equation r P Equation r P Equation Agathis australis (kauri) 0.822 < 0.001 y = 6.958 exp (0.114x) 0.744 < 0.001 y = 2.793 exp (0.111x) 0.799 < 0.001 y = 5.661 exp (0.106x) Dacrycarpus dacrydioide (kahikatea) 0.942 < 0.001 y = 25.937 exp (0.058x) 0.908 < 0.001 y = 16.919 exp (0.064x) 0.941 < 0.001 y = 12.900 exp (0.062x) Dacrydium cupressinum (rimu) 0.928 < 0.001 y = 17.665 exp (0.076x) 0.798 < 0.001 y = 13.086 exp (0.074x) 0.887 < 0.001 y = 23.053 exp (0.083x) Podocarpus totara (totara) 0.930 < 0.001 y = 39.461 exp (0.052x) 0.699 < 0.001 y = 35.461 exp (0.058x) 0.788 < 0.001 y = 51.688 exp (0.049x) Prumnopitys ferruginea (miro) 0.899 < 0.001 y = 10.667 exp (0.104x) 0.868 < 0.001 y = 8.871 exp (0.115x) 0.880 < 0.001 y = 9.919 exp (0.108x) Prumnopitus taxifolia (matai) 0.775 < 0.001 y = 7.987 exp (0.098x) 0.790 < 0.001 y = 26.804 exp (0.106x) 0.760 < 0.001 y = 5.446 exp (0.106x) Alectryon excelsus (titoki) 0.840 < 0.001 y = 31.875 exp (0.065x) 0.840 < 0.001 y = 14.277 exp (0.0812x) 0.734 < 0.001 y = 26.199 exp (0.052x) Vitex lucens (puriri) 0.786 < 0.001 y = 48.929 exp (0.035x) 0.948 < 0.001 y = 15.611 exp (0.049x) 0.950 < 0.001 y = 25.903 exp (0.042x) Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 8 of 26 Table 3 Allometric relationships between RCD and total root length, and total root biomass for roots > 1 mm and > 2 mm diameter Species Root size class Root length Root biomass 2 2 r P Equation r P Equation Agathis australis (kauri) > 1 mm 0.780 < 0.001 y = 0.542 exp (0.113x) 0.879 < 0.001 y = 1.173 exp (0.137x) > 2 mm 0.758 < 0.001 y = 0.161 exp (0.127x) 0.860 < 0.001 y = 0.879 exp (0.144x) Alectryon excelsus (titoki) > 1 mm 0.797 < 0.001 y = 8.179 exp (0.057x) 0.851 < 0.001 y = 17.629 exp (0.062x) > 2 mm 0.763 < 0.001 y = 1.842 exp (0.057x) 0.851 < 0.001 y = 11.206 exp (0.066x) Dacrycarpus dacrydioide (kahikatea) > 1 mm 0.783 < 0.001 y = 12.159 exp (0.045x) 0.891 < 0.001 y = 12.816 exp (0.053x) > 2 mm 0.704 < 0.001 y = 3.090 exp (0.047x) 0.897 < 0.001 y = 8.136 exp (0.057x) Dacrydium cupressinum (rimu) > 1 mm 0.816 < 0.001 y = 2.743 exp (0.080x) 0.925 < 0.001 y = 3.514 exp (0.098x) > 2 mm 0.829 < 0.001 y = 0.527 exp (0.088x) 0.946 < 0.001 y = 1.793 exp (0.106x) Podocarpus totara (totara) > 1 mm 0.461 < 0.001 y = 11.481 exp (0.040x) 0.730 < 0.001 y = 15.073 exp (0.049x) > 2 mm 0.659 < 0.001 y = 1.921 exp (0.046x) 0.780 < 0.001 y = 8.786 exp (0.053x) Prumnopitys ferruginea (miro) > 1 mm 0.831 < 0.001 y = 2.731 exp (0.095x) 0.909 < 0.001 y = 3.627 exp (0.114x) > 2 mm 0.855 < 0.001 y = 0.357 exp (0.112x) 0.931 < 0.001 y = 1.673 exp (0.127x) Prumnopitus taxifolia (matai) > 1 mm 0.779 < 0.001 y = 2.699 exp (0.101x) 0.789 < 0.001 y = 4.321 exp (0.102x) > 2 mm 0.515 < 0.001 y = 0.359 exp (0.124x) 0.808 < 0.001 y = 2.204 exp (0.111x) Vitex lucens (puriri) > 1 mm 0.812 < 0.001 y = 13.037 exp (0.037x) 0.886 < 0.001 y = 25.990 exp (0.038x) > 2 mm 0.723 < 0.001 y = 4.002 exp (0.033x) 0.878 < 0.001 y = 21.548 exp (0.038x) than Vitex lucens, which, with the exception of year 4, ranging between 0.803 (Agathis australis) and 0.944 had a larger RCD than the remaining species. In years 4 (Dacrycarpus dacrydioides), and all regressions were and 5, Agathis australis, Prumnopitys taxifolia and highly significant (P < 0.001) (Fig. 4). There was, Prumnopitys ferruginea had a significantly smaller (P = however, no significant difference in mean total AGB 0.05) RCD (~ 20–24 mm) than Dacrydium cupressinum among species until years 4 and 5, the exception being and Alectryon excelsus (~ 37–39 mm) (Additional file 2). Alectryon excelsus which showed no further increase in Significantly larger still were Podocarpus totara and mean total AGB (P = 0.05; data not shown). By year 5, Dacrycarpus dacrydioides, with a similar RCD as each the mean total AGB amassed by Agathis australis other (~ 48–50 mm), while Vitex lucens was the best (251.1 g) and Prumnopitys ferruginea (381.4 g) were not performed of the trialled species with an RCD of significantly different from each other nor from 77.6 mm, almost twice that of the next best performed Prumnopitys taxifolia (581.0 g). However, their biomass species and more than three times that of the slowest was significantly less (P = 0.05) than that of Alectryon growing Prumnopitys ferruginea (Additional file 2). excelsus (1092.3 g), Dacrydium cupressinum (1165.2 g) and Dacrycarpus dacrydioides (1231.0 g) and was Diameter at breast height (DBH) exceeded only by Podocarpus totara (2142.9 g) and Vitex There were significant differences among species with lucens (3026.0 g) (Fig. 5, Additional file 3). respect to DBH in year 5 (Additional file 2). The DBH of Prumnopitys taxifolia and Prumnopitys ferruginea were Biomass allocation to foliage, branches and stem not significantly different from each other (2–3.9 mm) Two-parameter exponential regression analysis was a but Prumnopitys taxifolia had a significantly smaller (P good fit for RCD assessed against stem (r =0.775–0.944), 2 2 = 0.05) DBH than did Agathis australis (8.0 mm), Dacry- branches (r =0.699–0.948) and foliage (r =0.734–0.950) dium cupressinum (9.8 mm) and Alectryon excelsus biomass, and all regressions were highly significant (P < (11.2 mm). Dacrycarpus dacrydioides (16.8 mm), Podo- 0.001) (Table 2). There were significant differences (P = carpus totara (21.5 mm) and Vitex lucens (23.5 mm) 0.05) among species in the allocation of biomass between each had a significantly larger (P = 0.05) DBH than did foliage, branches and stem in each year of the trial the other species but not compared with each other (Additional file 3). At 5 years of age, Agathis australis, (Additional file 2). Dacrycarpus dacrydioides, Prumnopitys ferruginea, Alectryon excelsus and Vitex lucens had the highest Above-ground biomass (AGB) proportion of their total AGB allocated to the stem (35– Two-parameter exponential regression analysis was a 49%), branches (19–40%) and least to foliage (20–32%). good fit for the RCD and total AGB data, with r Exceptions to this general trend were Dacrydium Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 9 of 26 Agathis australis Alectryon excelsus y=0.061x+0.085 y=0.053x+0.150 2 2 r =0.845; P<0.001 r =0.901; P<0.001 Dacrycarpus dacrydioide Dacrydium cupressinum y=0.052x+0.312 y=0.051x+0.366 2 2 r =0.906; P<0.001 r =0.916; P<0.001 Podocarpus totara Prumnopitys ferruginea y=0.048x+0.538 y=0.061x+0353 2 2 r =0.863; P<0.001 r =0.793; P<0.001 Prumnopitus taxifolia Vitex lucens y=0.041x+0.290 y=0.024x+0.368 2 2 r =0.849; P<0.001 r =0.768; P<0.001 Root collar diameter (mm) Fig. 3 Linear regression analysis of RCD and tree height. Note that the scale on the y and x axes differs between species cupressinum with a disproportionate 51% of its total AGB The differences in mean total BGB among species in 1- allocated to foliage comprising scale leaves closely and 2-year-old seedlings were initially small, but it be- appressed to the stem (Wardle 2002). Also, for came more significant in years 4 and 5 (P = 0.05; data Prumnopitys taxifolia and Podocarpus totara with deep, not shown), when the majority of species showed a sig- narrow crowns and short lateral branches, 69 and 37% of nificant increase in mean total BGB. Exceptions to this their respective total AGB was allocated to branches general pattern include Agathis australis, with no in- (Additional file 3). crease in mean total BGB in years 1 and 2, a modest in- crease in year 3 and no increase in year 4, and in year 5, Below-ground biomass (BGB) the increase in mean total BGB was significantly greater Similarly, two-parameter exponential regression analysis than in any of the previous years. By year 5, Agathis aus- was a good fit for the RCD and mean total BGB data, tralis (72.0 g) had a significantly less (P = 0.05) mean with r ranging between 0.757 (Podocarpus totara) and total BGB than both Prumnopitys ferruginea (134.9 g) 0.946 (Dacrydium cupressinum), and all regressions and Prumnopitys taxifolia (134.6 g). For the remaining were highly significant (P < 0.001) (Fig. 6). Additionally, species, the mean total BGB of Dacrycarpus dacry- there were significant differences (P = 0.05) between dioides (371.2 g), Dacrydium cupressinum (397.4 g), species in the allocation of biomass between root mass Alectryon excelsus (402.2 g) and Podocarpus totara and root bole in each year of the trial (Additional file 5). (450.0 g) was significantly different again but not to each Tree height (m) Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 10 of 26 Fig. 4 Exponential growth analysis of RCD and total above-ground biomass. Note that the scale on the y and x axes differs between species other. Vitex lucens (1257.4 g) had the highest mean total Podocarpus totara (103.0 g) and Dacrydium cupressinum BGB and was significantly different (P = 0.05) from all (102.6 g). In contrast, the mean total root bole biomass the other species (Fig. 5, Additional file 3). of the conifers Agathis australis (17.2 g), Prumnopitys ferruginea (22.2 g) and Prumnopitys taxifolia (22.9 g), Biomass allocation to root and root bole though not significantly different from each other, are For the majority of species, there were no significant each an order of magnitude less than for the remainder interspecies differences in the allocation of mean root of the conifers and broadleaved species (Additional file 5). and root bole biomass until years 4 and 5, at which time Vitex lucens also had the greatest root biomass (970.1 g, both root and root bole biomass increased significantly 79.3% of total BGB), while Agathis australis had the (P = 0.05; data not shown), similar to the pattern for the least (54.8 g, 75.2% of total BGB). For all species, the allocation of mean total BGB and AGB. For Agathis proportion of the total BGB allocated to the roots australis, the increase in root and root bole mass oc- (69–84%) was greater than that allocated to the root curred in year 3, whereas for Prumnopitys ferruginea bole (16–31%). and Vitex lucens, it did not increase until year 5. By year 5, Vitex lucens had the largest root bole biomass Root biomass distribution by depth and radius (287.3 g) but was not significantly different to Dacrycar- For all species in each year of the trial, 100% of the total pus dacrydioides (112.7 g), Alectryon excelsus (105.8 g), BGB (P < 0.001) was confined to the 0–0.5-m depth Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 11 of 26 and Alectryon excelsus occurred within a 0.5-m radius of the root bole, Podocarpus totara 76.8% and Vitex lucens 65.9% (Fig. 7, Additional file 6). Root biomass allocation by root diameter size class Exponential regression analysis of RCD and total root biomass for roots > 1- and > 2-mm diameter size class showed very similar regressions (Fig. 8), with RCD explaining between 73 and 92% of the variation in root biomass for roots > 1 mm and between 78 and 94% for roots > 2 mm (Table 3). In year 5, fibrous roots accounted for the highest pro- portion of the total root mass of each of the plate-rooted species Alectryon excelsus (23.5%), Prumnopitys ferrugi- nea (56.4%) and Prumnopitys taxifolia (49.7%) (Fig. 9). Similarly, for the tap-rooted species, fibrous roots accounted for between 28.1% (Dacrycarpus dacry- dioides) and 49.3% (Dacrydium cupressinum) of their total root mass. For all species, the percentage of fibrous Fig. 5 Plant biomass above- (dark bars) and below-ground (light bars) root biomass decreased as root sizes increased with in- in year 5. Error bars represent one standard error of the mean. Bars creasing plant age. The largest root diameter size class within each parameter with a different letter were significantly different (P= 0.05) (50–100 mm) was recorded for Vitex lucens only and comprised < 2% of its total root biomass. For the re- mainder of the species, the largest root diameter size increment, and inter-species differences (P = 0.05) in the class was 20–50 mm and comprised just 0.1% of Prum- mean maximum rooting depth were minimal. For most nopitys ferruginea and 6.8% of the total root mass species, this had increased from ~ 0.1 m in year 1 to ~ (largely tap root) of Agathis australis. 0.3 m by year 5 (Additional file 4). For any individual seed- ling, the tap-rooted Agathis australis developed the BGB/AGB ratio deepest (0.7 m) root, while the plate-rooted systems of In year 1, although the BGB/AGB ratio was highest for Prumnopitys taxifolia, Prumnopitys ferruginea and Alectr- the tap-rooted Vitex lucens (0.89) and lowest for the yon excelsus had the shallowest root systems (~ 0.15 m). plate-rooted Prumnopitys taxifolia (0.25), there was no By year 5, the mean maximum root depth of each of the consistent relationship between this ratio and root type trialled species was similar, whether conifers or broad- from year to year (Additional file 3). By year 5, Podocar- leaved species and tap-rooted or plate-rooted species pus totara (0.21) and Prumnopitys taxifolia (0.23) had (Additional file 4). the lowest BGB/AGB ratio, and although for Vitex Similarly, for the first 3 years, 100% of the total BGB (P < lucens the BGB/AGB ratio remained the highest of the 0.001) of all species was confined to within a 0.5-m radius trialled species (Fig. 5), it had halved to 0.42 by year 5. of the root bole. By year 4, only Dacrycarpus dacrydioides, Similarly, a decline in BGB allocation with increasing Alectryon excelsus and Vitex lucens hadextendedtheir age was apparent in Podocarpus totara (0.47 in year 1, roots to a radius of between 1.0 and 1.5 m from the root decreasing to 0.21 in year 5) and Prumnopitys ferruginea bole (Additional file 7). The root mass and percentage of (0.5 in year 1, decreasing to 0.35 in year 5). Averaged total root mass of Dacrycarpus dacrydioides (0.8 g, 0.4%) across all eight species, the BGB and AGB was 0.32 and and Alectryon excelsus (0.9 g, 0.2%) at this distance were 0.68, respectively. not significantly different from each other but were signifi- cantly different from Vitex lucens (3.4 g, 1.7%). Vitex lucens Root length and root system dimensions (diameter) was the only species by year 5 to have extended its root net- Exponential growth analyses of RCD and total root work to a radius of between 1.5 and 2.0 m from the root length fitted least well for Podocarpus totara with RCD bole; however, the roots comprised only 8.3 g, 0.8% of its explaining 46 and 66% of the variation in root length for total root mass. Between years 1 and 5, Agathis australis roots > 1 and > 2 mm, respectively. In contrast, this rela- was the only species to have 100% of its root mass confined tionship fitted best for Prumnopitys ferruginea with RCD to within 0.5 m of the root bole. At year 5, > 90% of the explaining between 83 and 85% of the variation in root total root mass of Dacrycarpus dacrydioides, Dacrydium length for roots > 1 mm and > 2 mm, respectively. cupressinum, Prumnopitys taxifolia, Prumnopitys ferruginea (Table 3, Fig. 10). Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 12 of 26 Fig. 6 Exponential growth analysis of RCD and total below-ground biomass. Note that the scale of the y and x axes differs between species At year 1, none of the species had produced roots > excelsus and Vitex lucens hadextendedtheir roots intothe 1-mm diameter. By year 2, Alectryon excelsus and Vitex 0.5–1.0-m radial sector where Vitex lucens had a signifi- lucens each produced > 2.5 m of root, significantly more cantly greater proportion (P = 0.05) of its total root length (P = 0.05) than Dacrycarpus dacrydioides (~ 2 m), which (34%)atthis radialdistancethandid Dacrycarpus in turn was significantly more than that of each of the dacrydioides (19%) and Alectryon excelsus (17%). remaining species at < 1 m of root length (Additional file 4). Similarly, at 1.0–1.5-m radial distance from the root By year 5, the total root length for each species increased bole, Vitex lucens had significantly more (P = 0.05) of substantially and differences between species became its total root length (~ 5%) than did Dacrycarpus more pronounced. For example, Agathis australis pro- dacrydioides (1%) and Alectryon excelsus (0.4%). duced the least mean total root length (9.01 m) while For all species, in years 1 and 2, there was little Vitex lucens (347.45 m) at the same age produced two or- inter-species difference in mean root system dimensions ders of magnitude more root length (Additional file 4). (i.e. mean maximum root spread). However, by year 5, of In years 2 and 3 of the trial, and for each of the species the broadleaf species, Vitex lucens had developed a sig- trialled, 100% of the total root length occurred within a nificantly greater (P = 0.05) mean maximum root spread 0.5-m radius of the root bole, and for Agathis australis,this (2.5 m) than had Alectryon excelsus (1.6 m), and its root was the case through to year 5 (Fig. 11, Additional file 7). spread was also significantly greater than that of the lar- By year 4, only Dacrycarpus dacrydioides, Alectryon gest of the conifer root systems, i.e. that of Podocarpus Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 13 of 26 Agathis australis Alectryon excelsus Dacrycarpus dacrydioide Dacrydium cupressinum Prumnopitys ferruginea Podocarpus totara Vitex lucens Prumnopitus taxifolia Distance from the root bole (m) Fig. 7 Distribution of total root biomass with increasing radial increments from the root bole. Note that the scale on the y axis differs between species totara (2.2 m) (Additional file 4). The mean maximum comprised roots in the 2–5-mm size class, 20% greater root spread of both Alectryon excelsus and Dacrycarpus than for the remainder of species (Fig. 12). The percent- dacrydioides (1.7 m), while not different from each other, age of total root length decreased with increasing root was significantly larger (P = 0.05) than the more compact size class, and for the 20–50-mm size class, where root systems of Prumnopitys taxifolia (1.2 m), Prumnop- present, roots comprised ≤ 1.5% of the total root length. itys ferruginea (1.0 m) and Dacrydium cupressinum None of the trialled species had roots in the > 50-mm (1.1 m), which in turn were not significantly different from diameter size class. each other. The mean maximum root spread of Agathis australis (0.5 m) was significantly smaller than that of the Plant biomass and carbon accumulation remainder of species trialled (Additional file 4). For each of the species trialled, the mean annual total biomass of individual trees increased significantly Root length by root diameter size class (Additional file 3). By year 5, the slowest growing of the For all the trialled species, the highest proportion of the trialled species including Agathis australis, Prumnopitys total root length of 5-year-old trees was in the 1– ferruginea and Prumnopitys taxifolia each accumulating −1 2-mm-diameter size class and with the exception of the least total biomass of ~ 0.32 kg tree (0.16 kg of C), −1 −1 Agathis australis comprised > 70% of their total root ~ 0.56 kg tree (0.26 kg of C) and ~ 0.72 kg tree length. For Agathis australis, 39% of its total root length (0.36 kg of C), respectively. Conversely, Dacrycarpus Total root biomass (g) Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 14 of 26 Agathis australis Alectryon excelsus Dacrycarpus dacrydioide Dacrydium cupressinum Podocarpus totara Prumnopitys ferruginea Prumnopitus taxifolia Vitex lucens Root collar diameter (mm) Fig. 8 Exponential growth analysis of RCD and total root biomass. Note that the scale of the y and x axes differs between species dacrydioides, Dacrydium cupressinum and Alectryon produced the longest laterals extending a distance of −1 excelsus each accumulated ~ 1.56 kg tree (0.78 kg of 2.2 m, significantly longer (P = 0.05) than all the C), while the highest biomass accumulation was by other trialled species, and of the broadleaved species, −1 Podocarpus totara ~ 2.59 kg tree (1.30 kg of C) and Vitex lucens attained the greatest mean root spread −1 Vitex lucens ~ 4.28 kg tree (2.14 kg of C) (Table 4). (2.5 m) (Additional file 4). Plate-rooted species include the conifers Prumnop- Root system types itys taxifolia and Prumnopitys ferruginea and the Two types of root systems (tap-rooted and plate-rooted) broadleaved Alectryon excelsus. Their juvenile root were observed based on the descriptive classification of systems are compact and consist largely of abundant Hinds and Reid (1957) we recognised. fibrous and numerous short, gnarly, lateral roots of Of the trialled species that developed a tap root small diameter size (< 1 mm diameter) somewhat re- (Agathis australis, Dacrycarpus dacrydioides, Vitex sembling that of Agathis australis but lacking a tap lucens, Podocarpus totara and Dacrydium cupressi- root (Fig. 2b). Of the plate-rooted species, the broad- num), the seedling radicle of Agathis australis leaved Alectryon excelsus developed the longest lateral (Fig. 2a) was prominent earliest (3-year-old plants) roots by year 5 with a spread of 1.6 m, significantly and extended deeper (0.3 m) than the remainder of longer (P = 0.05) than the other plate-rooted species the trialled species. By year 5, Podocarpus totara (Additional file 4). Total root biomass (g) Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 15 of 26 DBH is preferential to RCD as RCD can be affected by root buttress. However, RCD was used in the current study as DBH height (1.4 m) of the juvenile species trialled was not attained until year 4. Indeed, as suggested by Wagner and Ter-Mikaelian (1999), RCD may serve as a better predictor of both the above- and below-ground biomass as it is measured at the intersec- tion of the base of the stem with the root bole. In fact, RCD provided the best fit for tree height and total AGB, and all regressions were highly significant in each of the 5 years of the current trial and for all the trialled species. To date, generic rather than species-specific allometric equations have been used to estimate the AGB of ma- ture indigenous forest and shrub biomass. Extrapolating these equations to juvenile plants with smaller diameters far outside the range of data used to develop them is questionable however (Claesson et al. 2001; Zianis and Mencuccini 2003). The allometric biomass equations presented in this paper indicate that there are clear dif- ferences in growth performance among species, at least at their juvenile stage, and that different fitted coeffi- cients are required. Furthermore, there is a dearth of allometric equations for estimating the below-ground biomass of indigenous species. The absence of empirical data means that a figure of 20% is widely accepted, both locally (Beets et al. 2007) and internationally (Coomes et al. 2002) despite the precision of these root estimates tending to be low (Hall et al. 2001). Studies that include measurements of both the above- and below-ground growth attributes of the juvenile form of different species have also found RCD to be a good fit for estimating height, AGB and also BGB. For example, Marden et al. (2018; in press) found RCD to be a good fit for estimating all three attributes of 12 of New Zealand’s indigenous riparian species with each species requiring a specific fitted coefficient. Furthermore, Fig. 9 Percent root biomass (%) by diameter size class (mm) of Marden et al. (2016) found that two-parameter exponen- 5-year-old trees tial regressions using RCD was successful in predicting the BGB of five seed lots of Pinus radiata between 1 Discussion and 4 years old with RCD explaining between 93 and Plant allometry 98% of the variability in BGB. Similarly, Phillips et al. The physiological age of all the species was similar at (2015) found it possible to develop simple relationships the time of planting, but there were initial and between RCD and growth attributes with reasonable r significant differences in the above-ground metrics values for below-ground biomass and total root length. among species. Once established, interspecies differences However, year-on-year growth performance data for in year-on-year growth remained but became increas- the juvenile stage of most of New Zealand’s indigenous ingly variable. Payton et al. (2009) suggested that AGB species is particularly rare, thus limiting comparisons should be estimated using height and percentage cover with other published works. Nonetheless, there is for juvenile trees with a DBH < 50 mm. More recently, evidence to show that the proportion allocated to BGB Mason et al. (2014) used tree height and basal area to decreases with increasing age despite significant determine AGB in New Zealand shrublands while Beets year-on-year gains in total plant biomass in many spe- et al. (2012) used a two-factor exponential model which cies, both indigenous and exotic (Marden et al. 2005). uses DBH or DBH × height to estimate AGB biomass in This concurs with other local (Watson et al. 1995, 1999) mature native forests. Magalhães (2015) suggested that and international (Abernethy and Rutherfurd 2001; Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 16 of 26 Agathis australis Alectryon excelsus Dacrycarpus dacrydioide Dacrydium cupressinum Podocarpus totara Prumnopitys ferruginea Prumnopitus taxifolia Vitex lucens Root collar diameter (mm) Fig. 10 Exponential growth analysis of RCD and total root length. Note that the scale of the y and x axes differs between species Easson and Yarbrough 2002) studies where a decrease in 2005) or for 3-year-old Eucalyptus fastigata (33%), the root/shoot ratio for small saplings has been related to Sequoia sempervirons (29%), Acacia melanoxylon an increase in plant size (Kira and Shidei 1967; Walters (37%) and Pinus radiata (36%)trialledatthe same and Reich 1996) and to an increase in plant height (Cao site (Phillips et al. 2015). Although the below-ground and Ohkubo 1998). Conversely, Marden et al. (2016) biomass of each of five different 5-year-old clones of found no clear indication that the root/shoot ratio of 1- to Pinus radiata (~ 22%) established on steep hill coun- 4-year-old Pinus radiata decreased over this period but try (Marden et al. 2016)isatthe lowerend of this instead remained relatively constant. Biomass allocation to range, it is nonetheless consistent with that estab- components of woody plants is also affected by factors in- lished for similar-aged pines planted at a range of cluding plant architecture and morphology and climatic sites in New Zealand (Watson and Tombleson 2002, and edaphic factors (Ketterings et al. 2001). 2004) and likely reflects harsher site and edaphic In this study, the range of biomass apportioned to and climatic factors. below-ground components by year 5 was 23–41%, There was no consistent year-to-year difference in the not dissimilar to that of 12 early colonising indigen- root/shoot ratio between the tap- and plate-rooted ous species at the same age with a below-ground species or between conifers and broadleaved species biomass of between 24 and 44% (Marden et al. (Additional file 3). Total root length (m) Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 17 of 26 Fig. 11 Distribution of total root length with increasing radial increments from the root bole. Note that for the species with the greatest total root length there is a break in the y axis scale Root metrics and architecture 0.001) would likely remain confined to the 0–0.5-m Published accounts on the architecture of root systems depth increment, to within a 0.5-m radius of the root of New Zealand’s commonest and longest-lived native bole even in the absence of barriers to root penetration. conifer and broadleaved forest species are scarce, de- Also expected was that root biomass and root length scriptive and generally based on observations of partially would decrease with increasing depth and horizontal exposed roots undermined along river banks (Cameron distance from the root bole. Similarly, fibrous roots (as a 1963) or of root boles of windthrown trees (Bergin and percentage of total root biomass) were expected to Steward 2004). At the time of establishing this trial, decrease as root sizes increased. there were no root-related data (photographic or de- Conversely, while excavations of the root systems of scriptive) for juvenile specimens of the trialled species. juvenile Agathis australis revealed the known presence While some of the general findings on root distribution of a well-developed tap root (Bergin and Steward (2004), (depth and spread) reported here were anticipated from the presence of a tap root in Dacrycarpus dacrydioides, the results of a previous trial at the same site (Marden Vitex lucens, Podocarpus totara and Dacrydium et al. 2005), other findings were not. For example, it was cupressinum was unexpected and has not to our know- anticipated that a high proportion of the total BGB (P < ledge been previously reported. Additionally, their many Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 18 of 26 remaining devoid of root mass. This was also observed during the excavation of root systems of non-native for- est species (Phillips et al. 2015) and non-native Populus and Salix species at the same trial site (Phillips et al. 2014). These results suggest that interspecies differences in the distribution of roots and their overall root system architecture are likely to be inherent rather than due to seedling propagation, planting practices or site factors given that there were no observable differences in the physical attributes of the soil across this essentially flat terrace surface and competing influences from adjacent well-spaced plants were unlikely. Conversely, the highly asymmetric root architecture of juvenile Pinus radiata reported by Marden et al. (2016) is considered typical of tree root systems developed on steep slopes where the greatest proportion of the total root length trends in diagonally opposite directions pre- viously referred to as bilateral fan-shaped architecture (Chiatante et al. 2003). Here, the larger and stronger lat- eral roots develop in the upslope (Schiechtl 1980) under compression (Stokes and Guitard 1997) and provide the greatest contribution to tree stability/anchorage (Sun et al. 2008). Roots located in the downslope direction develop under tension, and are therefore smaller and weaker, and likely contribute less to tree anchorage. Nicoll and Ray (1996) reported that prevailing winds can influence the distribution of the structural roots of Sitka spruce (Picea sitchensis (Bong.) Carr.) where, relative to the prevailing wind, there were more roots located on the leeward side than on the windward side of trees. Marden et al. (2016) considered the development of the bilateral fan-shaped distribution of structural roots of Pinus radiata to have been influenced by the predomin- ance of northerly winds as an adaptation to improve the rigidity of the soil-root plate and to counteract their vul- nerability to windthrow. Fig. 12 Percent root length (%) by diameter size class (mm) of The openness and exposure of the conifer and broad- 5-year-old trees leaved species trial site to strong winds resulted in stem breakage of Alectryon excelsus, and Vitex lucens was leading lateral roots, though initially of small diameter, defoliated during periods of heavy frost. However, site were highly symmetric; thus, few areas of soil were left factors are unlikely to have contributed to inter-species totally devoid of any roots. These species typically had a differences in their overall growth performance as all the high number of roots per unit volume of soil and, as trialled species tolerate a wide range of environmental expected for juvenile plants, the bulk of their root mass conditions. Furthermore, with weed control, unlimited and total root length was confined to a shallow depth water availability, a fertile alluvial soil and the absence of and within close proximity to the root bole. By compari- barriers to root development, the resultant growth son, the root systems of the juvenile plate-rooted species performance could be considered as optimal for these (including the conifers Prumnopitys taxifolia and species. However, as is common with nursery-raised Prumnopitys ferruginea and the broadleaved Alectryon juvenile plants, factors such as wrenching, root excelsus) typically had fewer laterals roots but were of trimming and planter bag constriction likely stunt the larger diameter and with limited branching tended to rate of root biomass accumulation, root spread and developed in preferential directions. Thus, the number depth within the first 3 years following planting (Marden of roots per unit volume of soil was lower than for the et al. 2005). To what extent minor root binding affected tap-rooted species with significant areas of soil the eventual distribution of root diameter size classes Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 19 of 26 Table 4 Mean tree biomass (dry weight), carbon content (kg) and carbon stock (t CO /1000 stems per hectare) for 5-year-old trees Botanical name Common name Mean tree biomass (kg) Carbon content (kg) Carbon stock (t CO /1000 spha) Coniferous species a a a Agathis australis Kauri 0.32 (0.03) 0.16 (0.02) 0.53 (0.05) bc bc bc Dacrycarpus dacrydioides Kahikatea 1.60 (0.24) 0.80 (0.12) 2.64 (0.40) c c c Dacrydium cupressinum Rimu 1.56 (0.13) 0.78 (0.06) 2.58 (0.21) d d d Podocarpus totara Totara 2.59 (0.32) 1.30 (0.16) 4.28 (0.53) a a a Prumnopitys ferruginea Miro 0.52 (0.09) 0.26 (0.04) 0.85 (0.15) ab ab ab Prumnopitys taxifolia Matai 0.72 (0.07) 0.36 (0.03) 1.18 (0.12) Broadleaved species c c c Alectryon excelsus Titoki 1.49 (0.23) 0.75 (0.12) 2.47 (0.39) e e e Vitex lucens Puriri 4.28 (0.92) 2.14 (0.46) 7.07 (1.52) Values in parentheses represent standard error of the mean. Values with different letters, within each parameter, were significantly different (Student-Newman-Keuls P =0.05). The carbon stock values do not include woody debris or fine litter and spatial pattern of root development of the species re- The high concentration of roots (mass and length) of ported in this paper remains uncertain. Nonetheless, it both the tap- and plate-rooted species at shallow depth was clear that the timing and proportion of total plant bio- was also anticipated. General observations of the root mass allocated to roots and root bole did not increase sig- systems of mature plate-rooted indigenous species nificantly until years 4 and 5, and this appears to be confirm that their root systems are largely confined to related to the period (years after planting) that plants re- the uppermost metre or two of the soil with the preva- quire to establish their root systems sufficiently to sustain lence of shallow rooting in forest soils likely reflecting an increase in growth rate. With the exception of the dependence on nutrient cycling through litter fall broadleaved and plate-rooted Alectryon excelsus,the (Yeates 1924; Cranwell and Moore 1936; Wardle 2002). tap-rooted and predominantly coniferous species allocated Furthermore, observations of the root systems of the more of their total BGB to the root bole than did the mature, plate-rooted Dacrycarpus dacrydioides indicate remaining plate-rooted species or the tap-rooted Agathis that their lateral structural roots are shallow and often australis. That said, there is also likely to be a relationship sub-aerial as an adaptation to imperfectly drained soils between plant size and the rate of root and root bole bio- and/or wetter flatland sites including alluvial floodplains mass accumulation with the smallest and slowest growing and terraces with a periodic high water table dominated of the conifers Agathis australis, Prumnopitys taxifolia by pasture grasses (Wardle 2002). Alternatively, as sug- and Prumnopitys ferruginea amassing the least BGB. gested by Nicoll and Ray (1996), a shallow and extensive Between years 1 and 3, there was no consistent lateral root network may also be an adaptation to im- relationship between the diameter of the canopy and prove the rigidity of the soil-root plate and counteract root system dimensions (diameter) of tap- versus the increasing vulnerability to windthrow as trees grow plate-rooted species or between conifer and broadleaved taller. Similarly, the development of an extensive species. However, by year 4 and thereafter, the diameter network of lateral structural roots in Populus and Salix of the root systems of the broadleaved Alectryon excelsus species makes them the preferred species for soil conser- and Vitex lucens and the conifer Dacrycarpus dacry- vation and slope reinforcement on poorly drained sites dioides exceeded the width of their respective canopies such as earthflows (Phillips et al. 2014). suggesting that the ratio of root spread to canopy dri- The maximum rooting depth recorded for an indigen- pline may be different for different species at the same ous forest species is for a mature tap-rooted Agathis age. From the few available descriptions of root system australis. Unless impeded by rock or a hard soil pan, tap diameters of mature trees, it is evident that the lateral roots have been recorded at depths in excess of 2 m roots of conifers Dacrydium cupressinum and Podocar- (Bergin and Steward 2004), and ‘sinker roots’ have been pus totara may extend > 9 m from the root bole (Bergin observed to descend from large, lateral roots to a depth and Gea 2005, 2007). Similarly, the lateral roots of ma- of 4 m and terminate in a network of smaller roots (R. ture Prumnopitys taxifolia are often seen exposed on the LLoyd, pers. comm. cited in; Bergin and Steward 2004). ground surface and extending to > 19.5 m from the root Agathis australis is predominately found growing in the bole before disappearing into the soil (Allan 1926), and more sub-tropical regions of New Zealand so the early thus, for the more conical-shaped conifers, their lateral development of a deep tap root and relatively large bio- roots likely extend beyond the canopy dripline. mass allocation to roots should be beneficial to surviving Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 20 of 26 droughts as suggested by Becker and Castillo (1990). It species, and for areas actively planted and managed for has also been suggested that a larger (and early) alloca- mānuka honey production, 1200–1300 spha would be tion to root biomass is beneficial for tree establishment required. (T Poulson, pers. comm. cited in Coutts 1987). For the best performed of the trialled species, a mono- culture of broadleaved Vitex lucens planted at 1000 spha Biomass and carbon accumulation could potentially achieve a level of forest carbon stock of −1 Nationally, there is 1.45 Mha of marginal pastoral land 7.07 t CO ha by year 5 (Table 4), which approximated −1 suitable for reforestation by indigenous trees and shrubs to the 7.8 t CO ha estimated by the Ministry for (Trotter et al. 2005). The industry standard for establish- Primary Industries (2017) for new areas of planted and ment plantings of non-native species (principally Pinus reverting shrubland. However, irrespective of the species radiata) on marginal land varies considerably depending mix and planting regime adopted for new areas of on the class of land and its erosion severity (National indigenous forest establishment, the potentially Water and Soil Conservation Organisation 1976). In the achievable forest carbon stock by year 5 is likely to be an −1 highly erodible East Coast region of the North Island, order of magnitude less than the ~ 91.6 t CO ha of the planting density is generally around 1000–1250 forest carbon stock recorded for 4-year-old Pinus −1 stems per hectare (spha). The establishment of a mixed radiata (from Marden et al. 2016) and the 77 t CO ha indigenous forest comprising 125 of each of the eight estimated by the Ministry for Primary Industries (2017) species trialled here (1000 spha) would potentially amass for 5-year-old post-1989 plantings in the East Coast −1 ~ 1.5 t ha within 5 years of planting, which would region. −1 generate a forest carbon stock of 2.5 t CO ha . With Notwithstanding, the planting of indigenous conifers the exception of Agathis australis, the remainder of the and broadleaved species on erodible marginal land will conifers (mainly tap-rooted species) individually ultimately afford a more continuous and permanent for- accumulated significantly more total plant biomass than est sink for carbon over an active growth phase that ex- did any of the plate-rooted species, while the broad- tends from 150 to 500 years (Hall 2001; Hinds and Reid leaved Vitex lucens (also tap-rooted) accumulated the 1957) that is consistent with the prolonged effort likely largest total biomass of any of the trialled species required to effect significant reductions in atmospheric (Table 4, Additional file 3). Selection of a species mix CO levels. comprising 200 of each of the best performing of the tap-rooted conifers (Dacrycarpus dacrydioides, Dacry- Towards implementation dium cupressinum, Podocarpus totara) and the broad- The second objective of this trial was to assess the po- leaved Vitex lucens, together with the best performing of tential effectiveness of the trialled species for erosion the plate-rooted broadleaved Alectryon excelsa would control on newly afforested/reforested marginal pastoral −1 amass ~ 2.3 t ha , a forest carbon stock of land and/or as replacement species best suited to pro- −1 3.8 t CO ha . Furthermore, an early colonising viding long-term stability to areas of exotic forest indigenous shrubland species mānuka (Leptospermum deemed at greatest risk to storm-related erosion and scoparium (Forster et Forster f.)) established at the same under consideration for retirement in the future. −1 density would amass ~ 3.7 t ha resulting in a forest The well-developed literature on the geomechanical −1 carbon stock of 6.1 t CO ha over the same 5-year effects of vegetation on hill slope stability clearly indicate period (Marden and Lambie 2016). Thus, the planting of that tree roots reinforce soils, contribute to soil strength Leptospermum scoparium on marginal land would store and stabilise eroding hillslopes. The mechanical interac- −1 an additional ~ 2.3 t CO ha over an equivalent period tions of woody roots with the soil medium play an of time than that gained from planting an equivalent important role in tree anchorage and thus in the preven- density of the best performed of the trialled broadleaved tion of slope failure. Thus, one of the most important and conifer species. Interestingly, the carbon stock for traits is rooting depth, which, when soil conditions are areas of indigenous forest 5 years after establishment not limiting, is species dependent. The most effective (planting), or from the date when land-use change to re- slope stabilisation results when root development occurs −1 version commences, is estimated at 7.8 t CO ha at different depths in the soil profile (Schiechtl and Stern (Ministry for Primary industries 2017). By implication, 1994; Czernin and Phillips 2005). This has been ob- to achieve a similar level of carbon stock within this served in exposed root boles of mature Dacrydium time frame would for areas of new plantings require ~ cupressinum and Agathis australis. 2000 spha of the best performed of the trialled In New Zealand, shallow landslides on typically steep broadleaved and conifer species. Similarly, for areas set slopes underlain by Tertiary-aged bedrock are generally aside to permit passive reversion to occur would require ≤ 1 m deep (Marden et al. 1991; Page et al. 1994)so an equivalent density of mixed broadleaved and conifer roots must cross the basal shear plane in order to anchor Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 21 of 26 the soil into more stable substrate and stabilise a slope Larger diameter roots at greater soil depths would be against the initiation of shallow landslides (Wu et al. desirable to stabilise the deeper forms of mass move- 1979). This study shows that, for most of the tree species ment; thus, root diameter is another important consider- examined, the root systems are concentrated predomin- ation on landslide-prone slopes as thick roots reinforce antly in the upper soil profile with mean maximum root soil in the same way that concrete is reinforced with depths by year 5 ranging between 0.3 and 0.4 m. Further- steel rods, and the thicker the root, the greater its ability more, at this stage of their development, there was no sig- to penetrate soil (Clark et al. 2008). The most common nificant difference in this metric among individual conifer barriers limiting the development of a deep rooting species, between conifer and broadleaved species or system include a permanent high water table, bedrock between plate-rooted and tap-rooted species. The results interface or a cemented iron pan and stoniness and also indicate that there is a rapid decline in roots with depth of colluvium. In general terms, the root:shoot increasing depth and distance from the stem, that there ratios of seedlings are depressed on such sites. As previ- are interspecies differences in root distribution and that ously found for other New Zealand woody species, root each species allocates differing proportions of their total depth was not necessarily correlated with tree age but biomass to roots at different stages of growth, a finding rather to the stoniness and depth of slope colluvium similar to other local (Watson et al. 1995, 1999) and inter- (Watson et al. 1995). national (Abernethy and Rutherfurd 2001;Eassonand In terms of root anchorage, tap-rooted species will, in Yarbrough 2002) studies on plant root distribution. This, time, likely provide a higher level of reinforcement together with the few published reports on the root depth directly under a tree and to a greater depth than for of some of New Zealand’s tallest native conifer and hard- plate-rooted species. Thus, for tap-rooted species, root wood forest species (Cameron 1963), with the exception reinforcement quickly tapers off laterally as root density of Agathis australis, indicate that the rooting depth of ma- declines with increasing distance from the tree. In con- ture trees rarely exceeded 2–3 m. By implication, their trast, plate-rooted species with their widespread devel- comparatively shallow rooting depth will likely be a limit- opment of a lateral root network will likely provide a ing factor in their ability to contribute to deep-soil high level of near-surface soil-root reinforcement but reinforcement. Similarly, the ability of widely spaced nonetheless remain more susceptible to uprooting than plantings to prevent the occurrence of shallow landslide will tap-rooted species because as the number of roots diminishes as root distribution decreases with increasing decline with increased depth (Gray and Sotir 1996)so distance from the root bole. does root anchorage rapidly decline (Dupuy et al. 2005; Interspecies differences in root distribution (depth and Norris et al. 2008). lateral spread) have implications for the planting As a consequence of their shallow-rooted habit, many densities required to provide full, near-surface root of New Zealand’s conifer and broadleaved trees will have occupancy of the soil and/or to maximise the root dens- limited effectiveness in floodplain reaches where channel ity to the depth requirement at specific sites (Phillips hydraulic conditions are likely to result in the undercut- et al. 2012). For example, a dense lateral network of ting of stream banks to a depth greater than that of the woody roots predominantly confined to the upper soil maximum penetration depth of root systems (Marden horizons often forms a membrane that stabilises shallow et al. 2005). If the potential for streambed degradation soils (Schmidt et al. 2001) and significantly reduces exists, additional protection in the form of structural shallow landslide potential on steep slopes (Ekanayake materials (e.g. gabion baskets and rip rap) will be re- et al. 1997), while the larger tree roots can provide quired along the toe of banks and to some depth below reinforcement across lateral planes of weakness bound- normal streambed level. The limitations of root depth ing potential slope failures (Sidle et al. 1985). However, aside, some of the trialled conifer and broadleaved deeper soil mantles (> 5 m) benefit little from such species have adaptations suitable for floodplain restor- reinforcement as root density decreases dramatically ation. Alluvial communities of some of New Zealand’s with depth and few large roots are able to anchor across oldest podocarps (Podocarpus totara, Dacrycarpus dacry- the basal shear plane (Stokes et al. 2008). The only bene- dioides) are tolerant of wet soils and swampy sites subjected fit of root strength to the stabilisation of deep-seated to periodic flooding and seasonally high water tables. They landslides, e.g. earthflows and slumps, is when very large have the ability to produce a new root system after flooding lateral woody roots cross planes of weakness along the and inundation with river silt, and Podocarpus totara is able flanks of potential failures (Sidle and Ochiai 2006). to produce tiers of adventitious roots from the buried part However, such reinforcement would only benefit of the stem should the tree topple or break. Dacrycarpus deep-seated landslides of small area extent whereas dacrydioides, however, is intolerant of siltation if greater lateral roots also offer protection against most shallow than ~ 60 cm (Foweraker 1929; McSweeney 1982; landslides (Swanston and Swanson 1976). Campbell 1984). Many species, including Dacrydium Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 22 of 26 cupressinum, Podocarpus totara and Agathis australis,are species (e.g. Agathis australis) have the potential to equally tolerant of harsh growing conditions on steep develop thicker and stronger roots able to penetrate slopes with skeletal soils deficient in nutrients that are across the basal shear plane to provide a deeper level of prone to drying during seasonal droughts. reinforcement, albeit only for the smaller deep-seated Although native conifers and broadleaved trees have failures, than would plate-rooted species. Nonetheless, not traditionally been used specifically for soil conserva- as was evident before the mature indigenous forest cover tion, their successional role meets most of the require- was removed from hill-country areas, given time to ma- ments for use in revegetation where long-term erosion ture, these same species will slow down but may not control is required (Pollock 1986). It is also important to completely arrest many of the deeper-seated forms of consider that, in addition to its erosion-control function, mass failure. a species mix may have greater associated ecological, For many restoration sites, the strategy adopted should cultural, aesthetic and economic benefits than a single ideally include a mix of early colonising and successional species stand. In addition, most are sufficiently diverse species with plate- and tap-rooted characteristics. On in rooting form to stabilise slopes prone to shallow land- less exposed sites, a mixture of shrub and tree species slides and dry ravelling, while those with the limitations can be planted concurrently with the longer-lived conifer of root depth will be less effective in stabilising the dee- and broadleaved species interplanted at near final spa- per forms of mass movements such as earthflows and cing (Bergin and Gea 2005, 2007). Another option that slumps as their failure plane tends to be deeper than the has found favour with many community groups is to re- vertical limit of root penetration. Plant density and spe- strict the number of colonising species (three or four) cies mix are also key factors in determining plant effect- and then carry out follow-up planting with the late suc- iveness; thus, options that promote the quickest canopy cession, longer-lived species after a few years to fill in closure and root development at all levels of the soil any gaps created by failures from the initial planting profile are likely to be the most effective in promoting (Fred Lichtwark pers. comm.). slope stability (Phillips et al. 2001). The faster the soil is During selection of suitable plant materials, the cap- occupied by roots, the greater the reinforcement of the acity of species to match specific site conditions, e.g. soil and the greater the chance of limiting landslide initi- those that are long lived and can overcome potential ation on erosion-prone slopes and dry ravelling of ex- limitations to successful establishment such as over- posed soil and rock. As an example, soil occupancy by crowding resulting in suppressed growth (less is often the larger root systems of the broadleaved Vitex lucens best in the long term), frost, fungal and insect attack, and Alectryon excelsus would occur sooner (~ 5 years browsing, wind throw and stem breakage, must also be after planting) than the best performed of the conifers taken into account. Potential limitations to a successful (Podocarpus totara) tested under optimal growth condi- establishment outcome of the trialled species, particu- tions and using the mean maximum root spread at year larly during the early growth period, include susceptibil- 5 as an indicator of the rate at which soil-root ity to frost (e.g. Alectryon excelsus, Vitex lucens, reinforcement might be considered as being effective Dacrydium cupressinum, Agathis australis), insect dam- against erosion. Conversely, the more compact root sys- age (e.g. Podocarpus totara, Dacrydium cupressinum, tems of Prumnopitys taxifolia, Prumnopitys ferruginea Dacrycarpus dacrydioides, Agathis australis), fungi (e.g. and Dacrydium cupressinum would take longer (~ 10– Agathis australis, Podocarpus totara), browsing feral 15 years) to be effective and the significantly smaller mammals (e.g. Dacrydium cupressinum, Podocarpus to- root system of Agathis australis, the longest. Phillips tara, Agathis australis) and stem snap (e.g. Alectryon et al. (2015) compared the lateral root spread of exotic excelsus) (Bergin and Gea 2005, 2007). species including alder (Alnus rubra), cherry (Prunus serrulatus), blackwood (Acacia melanoxylon) and red- Conclusions wood (Sequoia sempervirons) with reported values for This study has provided much-needed new information indigenous riparian species (Marden et al. 2005), for mā- on the overall growth performance of eight of New nuka (Marden and Lambie 2016) and for Pinus radiata Zealand’s more common conifer and broadleaved forest at a similar age (Watson and Tombleson 2002, 2004; species. However, the conclusions drawn are primarily Marden et al. 2016). The conclusion drawn was that the premised on their early growth period (first 5 years) at species with the larger root system dimensions would one trial site; thus, it must be acknowledged that provide earlier soil-root reinforcement and thus be more allometric relationships will likely change with increasing effective in mitigating the initiation of shallow landslides. tree age. The excavation of whole-plants has provided Similarly, using root depth as an indicator of species po- rare and valuable insights into inter-species differences tential for stabilising deeper-seated forms of erosion (e.g. in root architecture (dimensions), root sizes and biomass rotational slumps and larger earthflows), tap-rooted and their distribution relative to soil depth and distance Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 23 of 26 from the stem. Such data are integral to the selection of lands collected in 2002/03. Landcare Research Contract an appropriate species mix, planting density and as a Report LC0506/099. means of assessing the period (years after planting) Payton, I., Forrester, G., Lambie, S., Berben, P., Pin- juvenile plantings are likely to remain vulnerable to the kney, T. (2009). Development and validation of allomet- initiation of shallow landslides. Although the juvenile ric equations for carbon inventory of indigenous forests plants of the trialled species typically develop very shallow and shrublands. Landcare Research Contract Report root systems, observations and anecdotal descriptions of LC0910/004. Prepared for the Ministry of Agriculture the root systems of older trees show they have potential and Forestry. for mitigating shallow landslides, earthflows of small ex- tent and dry ravelling within a few years of planting. Con- Additional files versely, as the basal failure plane of the deeper forms of Additional file 1: Ecological and site requirements of the trialled mass movements such as earthflows and slumps tends to species. (DOCX 16 kb) be deeper than the vertical limit of root penetration of the Additional file 2: Mean above-ground attributes of six indigenous coni- majority of the trialled species, their stabilisation will fer and two broadleaved species. (DOCX 17 kb) occur progressively over several decades. Additional file 3: Mean tree component biomass, above-ground (AGB) The previously unavailable species-specific information and below-ground (BGB) biomass totals and BGB/AGB ratio. (DOCX 23 kb) on rates of whole-plant biomass accumulation for the Additional file 4: Mean below-ground attributes of six indigenous indigenous forest species presented in this paper will conifer and two broadleaved species. (DOCX 17 kb) improve current estimates of forest carbon stocks −1 Additional file 5: Mean root and root bole biomass (g), as a percent of (t CO ha ) during the juvenile stage of growth of total below-ground biomass (BGB). (DOCX 19 kb) mixed-species or single-species woodlots proposed for Additional file 6: Mean root biomass (g) and percent of total root new plantings on land considered marginal for pastoral biomass, by 0.5-m radial increments from the root bole. (DOCX 19 kb) use and unsustainable for short-rotation exotic forestry. Additional file 7: Mean and percent of total root length (m) of roots > 1 mm diameter, by 0.5-m radial increments from the root bole. Species-specific growth performance data for both (DOCX 23 kb) old-growth and juvenile trees of many of New Zealand’s indigenous species remain elusive. It is therefore import- Abbreviations ant that further time-series baseline data on plant AGB: Above-ground biomass; AGS: Afforestation Grant Scheme; BGB: Below- growth rates are collected for a greater variety of New ground biomass; BP: Before present; C: Carbon; CO : Carbon dioxide; DBH: Diameter at breast height; ETS: Emissions Trading Scheme; Zealand’s indigenous species across a diversity of site FCCC: Framework Convention for Climate Change; GPR: Ground-penetrating conditions known to influence plant growth and that radar; IPCC: Intergovernmental Panel on Climate Change; PFSI: Permanent data on root metrics, in particular, are collected in a for- Forest Sink Initiative; RCD: Root collar diameter; spha: Stems per hectare mat suitable for inclusion in slope stability and/or soil Acknowledgements reinforcement models (e.g. Schwarz et al. 2010), many of We acknowledge the support of the Tairawhiti Polytechnic Rural Studies which suffer from a paucity of the types of data collected Unit, Gisborne, on whose land this plant trial was located. We thank interns in this study (Phillips et al. 2012). Claire Butty (France), Sandra Viel (Germany) and Kaisa Valkonen (Finland) and Landcare Research colleagues Alex Watson and Richard Hemming for assistance with field extraction and the partitioning and measuring of plant components. Graphics were drawn by Nic Faville and Suzanne Lambie. Anne Endnotes Austin edited the script, and Anne Sutherland provided GIS support. Robyn Simcock reviewed and provided valuable comment. Thanks also to the Dale, J. M. (2013). Evaluation of methods for quantify- anonymous external reviewers for their valuable comments This research ing carbon storage of urban trees in New Zealand. Paper was supported by funding from the New Zealand Ministry of Business, as partial fulfilment of Bachelor of Engineering (Honours) Innovation and Employment (MBIE), Specific Science Investment Fund (SSIF), ‘Growing Confidence in Forestry’s Future’ research programme (contract thesis, University of Auckland, Auckland, New Zealand. CO4X1306) and ‘Clean Water Productive Land’ research programme (contract Beets, P. N., Oliver, G. R., Kimberley, M. O., Pearce, S. H. C10X1006). (2008). Allometric functions for estimating above-ground Availability of data and materials carbon in native forest trees, shrubs and ferns. Scion Report Please contact the author for data requests. 12679, prepared for the Ministry for the Environment. Allen, R. B., Wiser, S. K., Hall, G., Moore, T., Goulding, Authors’ contributions MM was the primary author. MM, SL and CJP undertook the fieldwork. SL C., Beets, P., Nordmeyer, A. (1998). A national system for compiled the data into spreadsheets and completed the statistical analyses. monitoring change in carbon storage in indigenous forests All authors read and approved the manuscript. and scrub: testing plot locations on a South Island tran- Competing interests sect. Contract Report JNT9798/166 for the Ministry for The authors declare that they have no competing interests. the Environment, Wellington, New Zealand. Peltzer, D. A., Payton, I. J. (2006). Analysis of carbon Ethics approval and consent to participate monitoring system data for indigenous forests and shrub Not applicable. Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 24 of 26 Publisher’sNote Coutts, MP. (1987). Developmental processes in tree root systems. Canadian Springer Nature remains neutral with regard to jurisdictional claims in published Journal of Forest Research, 17(8), 761–767. maps and institutional affiliations. Cranwell, LM, & Moore, LB. (1936). The occurrence of kauri in montane forest on Te Moehau. New Zealand Journal of Science and Technology, 18, 531–543. Author details Czernin, A, & Phillips, CJ. (2005). Below-ground morphology of Cordyline australis Landcare Research, NZ, Ltd, PO Box 445, Gisborne 4010, New Zealand. (New Zealand cabbage tree) and its suitability for riverbank stabilisation. New Landcare Research, NZ, Ltd, Private Bag 3127, Hamilton 3240, New Zealand. Zealand Journal of Botany, 43, 851–864. Landcare Research, NZ, Ltd, PO Box 69041, Lincoln 7640, New Zealand. Dupuy, L, Fourcaud, T, Stokes, A. (2005). A numerical investigation into factors affecting the anchorage of roots in tension. European Journal of Soil Science, Received: 15 November 2017 Accepted: 8 May 2018 56, 319–327. Easson, G, & Yarbrough, LD. (2002). The effects of riparian vegetation on bank stability. Environmental & Engineering Geoscience, 8, 247–260. Ekanayake, J, Marden, M, Watson, AJ, Rowan, D. (1997). Tree roots and slope stability: a comparison between Pinus radiata (Radiata pine) and Kunzea References ericoides (Kanuka). New Zealand Journal of Forestry Science, 27, 216–233. Abernethy, B, & Rutherfurd, ID. (2001). The distribution and strength of riparian Foweraker, C. E. (1929). The podocarp rain forests of Westland, New Zealand. tree roots in relation to riverbank reinforcement. Hydrological Processes, 15, Kahikatea and Totara forests and their relationships to silting. Te Kura 63–79. Ngahere. A forestry journal issued by the Forestry Club of the Canterbury Allan, HH. (1926). The surface roots of an individual matai. New Zealand Journal of College School, Christchurch, New Zealand. Science and Technology, 8, 233–234. Gray, DH, & Sotir, RB (1996). Biotechnical and soil bioengineering slope stabilisation. Allan Herbarium (2000). New Zealand plant names database. New Zealand: New York: Wiley. Landcare Research Available http://nzflora.landcareresearch.co.nz. Accessed Hall, G, Wiser, S, Allen, R, Beets, P, Goulding, C. (2001). Strategies to estimate o 15 Jan 2005. national forest carbon stocks from inventory data: the 1990 New Zealand Becker, P, & Castillo, A. (1990). Root architecture of shrubs and saplings in the baseline. Global Change Biology, 7, 389–403. understory of a tropical moist forest in lowland Panama. Biotropica, 22, Hall, GMJ. (2001). Mitigating an organisation’s future net carbon emissions by 242–249. native forest restoration. Ecological Applications, 11, 1622–1633. Beets, PN. (1980). Amount and distribution of dry matter in a mature beech/ Hasselmann, K. (1997). Climate change after Kyoto. Nature, 390, 225–226. podocarp community. New Zealand Journal of Forestry Science, 10, 395–418. Beets, PN, Kimberley, MO, Oliver, GR, Pearce, BH, Graham, JD, Brandon, A. (2012). Heth, D, & Donald, DGM. (1978). Root biomass of Pinus radiata D. Don. South Allometric equations for estimating carbon stocks in natural forest in New Africa Forestry Journal, 107,60–70. Zealand. Forests, 3, 818–839. Hewitt, AE (2010). New Zealand soil classification, Landcare Research Science Series Beets, PN, Kimberley, MO, Paul, TSH, Oliver, GR, Pearce, SH, Buswell, JM. (2014). No. 1. Lincoln: Manaaki Whenua Press. The inventory of carbon stocks in New Zealand’s post-1989 natural forest for Hinds, HV, & Reid, JS (1957). Forest trees and timbers of New Zealand, New Zealand reporting under the Kyoto Protocol. Forests, 5(9), 2230–2252. Forest Service Bulletin 12. Rotorua: New Zealand Forest Service. Beets, PN, Pearce, SH, Oliver, GR, Clinton, PW. (2007). Root/shoot ratios for Hruska, J., Cermak, J., Sustek, S. (1999). Mapping tree root systems with ground- deriving below-ground biomass of Pinus radiata stands. New Zealand Journal penetrating radar. Tree Physiology, 19, 125–130. (2000). of Forestry Science, 37(2), 267–288. IPCC (2000). In RT Watson, IR Noble, B Bolin, NH Ravindranth, DJ Verardo, DJ Bergin, D (2003). Totara: establishment, growth and management, New Zealand Dokken (Eds.), Land use, land-use change, and forestry. A special report of the Indigenous Tree Bulletin No. 1. Rotorua: New Zealand Forest Research Institute. IPCC. Cambridge: Cambridge University Press. Bergin, D, & Gea, L (2005). Native trees: planting and early management for Jara, MC, Henry, M, Réjou-Méchain, M, Wayson, C, Zapata-Cuartas, M, Piotto, D, production wood, New Zealand Indigenous Tree Bulletin No. 3. Rotorua: New Guier, FA, Lombis, HC, López, EC, Lara, RC, Rojas, KC, Pasquel, JD, Montoya, Zealand Forest Research Institute. AD, Vega, JF, Galo, AJ, López, OR, Marklund, LG, Fuentes, JMM, Milla, F, Bergin, D, & Gea, L (2007). Native trees: planting and early management for wood Chaidez, JJN, Malavassi, EO, Pérez, J, Zea, CR, Garcia, LR, Pons, RR, Saint-André, production, New Zealand Indigenous Tree Bulletin No. 3 (revised edition). L, Sanquetta, C, Scott, C, Westfall, J. (2014). Guidelines for documenting and Rotorua: New Zealand Forest Research Institute. reporting tree allometric equations. Annals of Forest Science, 72, 763–768. Bergin, D, & Steward, G (2004). Kauri: ecology, establishment, growth and Ketterings, QM, Coe, R, van Noordwijk, M, Ambagau, Y, Palm, CA. (2001). management, New Zealand Indigenous Tree Bulletin No, 2. Rotorua: New Reducing uncertainty in the use of allometric biomass equations for Zealand Forest Research Institute. predicting above-ground tree biomass in the mixed secondary forests. Forest Cairns, MA, Brown, S, Helmer, EH, Baumgardner, GA. (1997). Root biomass Ecology and Management, 146, 199–209. allocation in the worlds upland forests. Oecologia, 111,1–11. Kira, T, & Shidei, T. (1967). Primary production and turnover of organic matter in different forest ecosystems of the western Pacific. Japenese Journal of Cameron, RJ. (1963). A study of the rooting habits of rimu and tawa in pumice Ecology, 17,70–87. soils. New Zealand Journal of Forestry, 8, 771–785. Campbell, DJ. (1984). The vascular flora of the DSIR study area, lower Korner, C. (1998). A re-assessment of high elevation tree line positions and their Orongorongo Valley, Wellington, New Zealand. New Zealand Journal of explanation. Oecologia, 115, 445–459. Botany, 22, 223–270. Magalhães, TM. (2015). Allometric equations for estimating belowground biomass Cao, KF, & Ohkubo, T. (1998). Allometry, root/shoot ratio and root architecture in of Androstachys johnsonii Prain. Carbon Balance and Management, 10, 16. understory saplings of deciduous dicotyledonous trees in central Japan. Marden, M., Lambie, S. (2016). Plot based, growth performance of space-planted Ecological Research, 13(2), 217–227. mānuka (Leptospermum scoparium) on marginal land, and vulnerability to erosion: Carswell, FE, Burrows, LE, Hall, GMJ, Mason, NWH, Allen, RB. (2012). Carbon and final report. Ministry for Primary Industries Technical Paper No 2016/20. http:// plant diversity gain during 200 years of woody succession in lowland New www.mpi.govt.nz/news-resources/publications.aspx Accessed 30 Aug 2016. Zealand. New Zealand Journal of Ecology, 36(2), 191–202. Marden, M, Lambie, S, Rowan, D. (2018). Root system attributes of 12 juvenile Chiatante, D, Scippa, SG, Iorio, AD, Sarnataro, M. (2003). The influence of steep slopes indigenous early-colonising shrub and tree species with potential for on root system development. Journal of Plant Growth Regulation, 21,247–260. mitigating erosion in New Zealand. New Zealand Journal of Forestry Science, 48 (in press). Claesson, S, Sahlén, K, Lundmark, T. (2001). Functions for biomass estimation of young Pinus sylvestris, Picea abies and Betula spp. from stands in Northern Sweden with Marden, M, Phillips, CJ, Rowan, D (1991). Declining soil loss with increasing age high stand densities. Scandinavian Journal of Forestry Research, 16,136–146. of plantation forest in the Uawa catchment, East Coast region, North Island, Clark, LJ, Price, AH, Steel, KA, Whalley, WR. (2008). Evidence from near-isogenic New Zealand. In Proceedings of the International Conference on Sustainable lines that root penetration increases with root diameter and bending Land Management, (pp. 358–361). Napier: Simon Printing Co. Ltd. stiffness in rice. Functional Plant Biology, 35, 1163–1171. Marden, M, Rowan, D, Lambie, S. (2016). Root development and whole-tree Coomes, DA, Allen, RB, Scott, NA, Goulding, C, Beets, P. (2002). Designing systems allometry of juvenile trees of five seed lots of Pinus radiata D.Don: implications to monitor carbon stocks in forests and shrublands. Forest Ecology and for forest establishment on erosion-prone terrain, East Coast region, North Management, 164,89–108. Island, New Zealand. New Zealand Journal of Forestry Science, 46:24. Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 25 of 26 Marden, M, Rowan, D, Phillips, C. (2005). Stabilising characteristics of New Zealand Schwendenmann, L, & Mitchell, ND. (2014). Carbon accumulation by native trees and indigenous riparian colonising plants. Plant and Soil, 278,95–105. https://doi. soil in an urban park, Auckland. New Zealand Journal of Ecology, 38(2), 213–220. org/10.1007/s11104-004-7598-2. Scott, NA, White, JD, Townsend, JA, Whitehead, D, Leathwick, JR, Hall, GMJ, Mason, NWH, Beets, PN, Payton, I, Burrows, L, Holdaway, RJ, Carswell, FE. (2014). Marden, M, Rogers, GND, Watson, AJ, Whaley, PT. (2000). Carbon and Individual-based allometric equations accurately measure carbon storage nitrogen distribution and accumulation in a New Zealand scrubland and sequestration in shrubland. Forests, 5, 309–324. ecosystem. Canadian Journal of Forest Research-Revue Canadienne De Masters, S. E., Holloway, J. T., Mc Kelvey, P. J. (1957). The National Forest Survey of Recherche Forestiere, 30(8), 1246–1255. New Zealand, 1955. Vol. 1. Wellington: Government Printer. Sidle, RC, & Ochiai, H (2006). Landslides: processes, prediction, and land use, American Geophysical Union, water resources monogram 18. Washington, DC: McGlone, MS (1988). Glacial and Holocene vegetation history; 20-ky to present: New Zealand. In B Huntly, T Webb III (Eds.), Vegetation history. Handbook of AGU. vegetation science, (pp. 557–559). Dordrecht: Kluwer. Sidle, RC, Pearce, AJ, O’Loughlin, CL (1985). Hillslope stability and land use, McSweeney, GD. (1982). Matai/totara flood-plain forests in South Westland. New American Geophysical Union, water resource monogram, 11 (p. 140). Zealand Journal of Ecology, 5, 121–128. Washington, DC: American Geophysical Union. Ministry for Primary Industries (2015). A guide to the afforestation grant Steward, GA, & Beveridge, AE. (2010). A review of New Zealand kauri (Agathis scheme. Wellington: Ministry for Primary Industries http://www.mpi.govt. australis (D. Don) Lindl.): its ecology, history, growth and potential for nz/funding-and-programmes/forestry/afforestation-grant-scheme/ management for timber. New Zealand Journal of Forestry Science, 40,34–59. Accessed 21 May 2017. Stokes, A, & Guitard, D (1997). Tree root response to mechanical stress. In A Ministry for Primary Industries (2017). A guide to carbon look-up tables for forestry Altman, Y Waisel (Eds.), Biology of root formation and development, (pp. 227– in the emissions trading scheme. Wellington: Ministry for Primary Industries 236). New York: Plenum Press. http://www.mpi.govt.nz/growing-and-producing/forestry/forestry-in-the- Stokes, A, Norris, JE, van Beek, LPH, Bogaard, T, Cammeraat, E, Mickovski, SB, emmissions-trading-scheme/emissions-returns/ Accessed 11 Sept 2017. Jenner, A, Di Iorio, A, Fourcaud, T (2008). How vegetation reinforces soil on National Water and Soil Conservation Organisation [NWASCO] (1976). New slopes. In JE Norris, A Stokes, SB Mickovski, E Cammeraat, R van Beek, BC Zealand land resource inventory worksheets. Wellington: Government Printer. Nicoll, A Achim (Eds.), Slope stability and erosion control: ecotechnological New Zealand Climate Change Office (2003). Climate change. National Inventory solutions, (pp. 65–118). Dordrecht: Springer. Report for New Zealand. Greenhouse gas inventory 1990–2001. Wellington: Sun, HL, Li, SC, Xiong, WL, Yang, ZR, Cui, BS, Yang, T. (2008). Influence of slopes on Ministry for the Environment. root system anchorage of Pinus yunnanensis. Ecological Engineering, 32,60–67. Nicoll, BC, & Ray, D. (1996). Adaptive growth of tree root systems in response to Swanston, DN, & Swanson, FJ (1976). Timber harvesting, mass erosion, and wind action and site conditions. Tree Physiology, 16, 891–898. steepland forest geomorphology in the Pacific North West. In DR Coates (Ed.), Norris, JE, Stokes, A, Mickovski, SB, Cammeraat, E, van Beek, LPH, Nicol, B, Achim, Geomorphology and engineering,(pp. 199–221). Stroudsburg: Dowden, A (2008). Slope stability and erosion control: ecotechnological solutions. Hutchinson, and Ross. Dordrecht: Springer. Tate, KR, Scott, NA, Parshotam, A, Brown, L, Wilde, RH, Giltrap, DJ, Trustrum, NA, Gomez, B, Ross, DJ. (2000). A multi-scale analysis of a terrestrial carbon Page, MJ, Trustrum, NA, De Rose, RC. (1994). A high-resolution record of storm- budget: is New Zealand a source or sink of carbon? Agriculture Ecology and induced erosion from lake sediments, New Zealand. Journal of Environment, 82, 229–246. Paleolimnology, 11, 333–348. Parliamentary Commissioner for the Environment (2002). Weaving resilience into Tate, KR, Scott, NA, Saggar, S, Giltrap, DJ, Baisden, WT, Newsome, PF, Trotter, CM, our working lands: recommendations for the future roles of native plants. Wilde, RH. (2003). Land use change alters New Zealand’s terrestrial carbon Wellington: Parliamentary Commissioner for the Environment. budget: uncertainties associated with estimates of soil carbon change Phillips, CJ, Marden, M, Lambie, S. (2014). Observations of root growth of between 1990–2000. Tellus, 55B, 365–377. young poplar and willow planting types. New Zealand Journal of Forestry Trotter, C, Tate, K, Scott, N, Townsend, J, Wilde, H, Lambie, S, Marden, M, Pinkney, Science, 44:15. T. (2005). Afforestation/reforestation of New Zealand marginal pasture lands Phillips, CJ, Marden, M, Lambie, S. (2015). Observations of “coarse” root by indigenous shrub lands: the potential for Kyoto forest sinks. Annals of development in young trees of nine exotic species from a New Zealand plot Forestry Science, 62, 865–871. trial. New Zealand Journal of Forestry Science, 45: 13. Wagner, RG, & Ter-Mikaelian, MT. (1999). Comparison of biomass component equations for four species of northern coniferous tree seedlings. Annals of Phillips, CJ, Marden, M, Lambie, S, Watson, A, Ross, C, Fraser, S. (2012). Forestry Science, 56, 193–199. Observations of below-ground characteristics of young redwood trees (Sequoia sempervirens) from two sites in New Zealand- implications for Walters, MB, & Reich, PB. (1996). Are shade tolerance, survival, and growth linked? erosion control. Plant and Soil. https://doi.org/10.1007/s11104-012-1286-4. Low light and nitrogen effects on hardwood seedlings. Ecology, 77(3), 841–853. Phillips, CJ, Marden, M, Rowan, D, Ekanayake, JC (2001). Stabilising characteristics Wardle, P (2002). Vegetation of New Zealand. New Jersey: The Blackburn Press. of native riparian vegetation in New Zealand. In Proceedings of 3rd Australian Watson, AJ, Marden, M, Rowan, D (1995). Tree species performance and slope Stream Management Conference, (pp. 507–512). Brisbane: Cooperative stability. In DH Barker (Ed.), Vegetation and slopes, (pp. 161–171). London: Research centre for catchment Hydrology. Thomas Telford Press. Pollock, K. M. (1986). Plant materials handbook for soil conservation, Vol 3: Native Watson, AJ, & O’Loughlin, CL. (1990). Structural root morphology and biomass of plants. Water and soil miscellaneous publications no. 95, Palmerston North: three age classes of Pinus radiata. New Zealand Journal of Forestry Science, 20, Soil Conservation Centre, Ministry of Works and Development. 97–110. Poole, AL, & Adams, NM (1994). Trees and shrubs of New Zealand: field guide series. Watson, AJ, Phillips, CJ, Marden, M. (1999). Root strength, growth, and rates of Canterbury: Manaaki Whenua Press. decay: root reinforcement changes of two tree species and their Russo, SE, Jenkins, KL, Wiser, SR, Uriate, M, Duncan, RP, Coombs, DA. (2010). contribution to slope stability. Plant & Soil, 217,39–47. Interspecific relationships among growth, mortality and xylem traits of Watson, AJ, & Tombleson, JD. (2002). Toppling in juvenile pines: a comparison of woody species from New Zealand. Functional Ecology, 24, 253–262. the root system characteristics of direct-sown seedlings, and bare-root Salmon, JT (1986). The reed field guide to New Zealand native trees. Auckland: seedlings and cuttings. Plant and Soil, 239, 187–196. Reed Books. Watson, AJ, & Tombleson, JD. (2004). Toppling in young pines: temporal changes Schiechtl, HM (1980). Bioengineering for land reclamation and conservation. in root system characteristics of bare-root seedlings and cuttings. New Edmonton: University of Alberta Press. Zealand Journal of Forestry Science, 34(1), 39–48. Schiechtl, HM, & Stern, R (1994). Handbuch fur naturnahen Wasserbau, Eine Anleitung Will, GM (1966). Root growth and dry-matter production in a high-producing stand fur ingenieurbiologische Bauweisen. Wien: Osterreichischer Agrarverlag. of Pinus radiata. Research Notes 44. Rotorua: New Zealand Forest Service, Forest Research Institute. Schmidt, KM, Roering, JJ, Stock, JD, Dietrich, WE, Montgomery, DR, Schaub, T. (2001). Root cohesion variability and shallow landslide susceptibility in the Williams, J (1978). Carbon dioxide, climate and society: proceedings of an Oregon Coast Range. Canadian Geotechnical Journal, 38, 995–1024. International Institute for Applied Systems Analysis (IIASA) workshop. February Schwarz, M, Lehmann, P, Or, D. (2010). Quantifying lateral root reinforcement in 21–24, 1978. Oxford: Pergamon Press. steep slopes from a bundle of roots to tree stands. Earth Surface Processes Wilmshurst, JM. (1997). The impact of human settlement on vegetation and soil and Landforms, 35(3), 354–367. https://doi.org/10.1002/esp.1927. stability in Hawke’s Bay, New Zealand. Journal of Botany (NZ), 35,97–111. Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 26 of 26 Wu, TH, McKinnel, WP, Swanston, DN. (1979). Strength of tree roots and landslides on Prince of Wales Island. Alaska Canadian Geotechnical Journal, 16,19–33. Yeates, JS. (1924). The root nodules of New Zealand pines. New Zealand Journal of Science and Technology, 7, 121–124. Zianis, D, & Mencuccini, M. (2003). Aboveground biomass relationships for beech (Fagus moesiaca Cz) trees in Vermio Mountain, Northern Greece, and generalised equations for Fagus sp. Annals of Forestry Science, 60, 439–448. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png New Zealand Journal of Forestry Science Springer Journals

Biomass and root attributes of eight of New Zealand’s most common indigenous evergreen conifer and broadleaved forest species during the first 5years of establishment

Loading next page...
 
/lp/springer-journals/biomass-and-root-attributes-of-eight-of-new-zealand-s-most-common-Bq6kCta7y1
Publisher
Springer Journals
Copyright
Copyright © 2018 by The Author(s).
Subject
Life Sciences; Forestry
eISSN
1179-5395
DOI
10.1186/s40490-018-0113-y
Publisher site
See Article on Publisher Site

Abstract

Background: Tree allometric equations are critical tools for determining tree volume, biomass and carbon stocks. However, there is a lack of species-specific biomass equations for juvenile trees of many of New Zealand’sindigenous species. The aim of this study was to provide allometric equations for total above- and below-ground biomass and total root biomass and length for eight common evergreen conifer and broadleaved species. Methods: In a plot-based field trial, growth metrics of conifers Prumnopitys taxifolia (matai), Agathis australis (kauri), Prumnopitys ferruginea (miro), Podocarpus totara (totara), Dacrycarpus dacrydioides (kahikatea) and Dacrydium cupressinum (rimu) and broadleaved species Alectryon excelsus (titoki) and Vitex lucens (puriri) were measured annually. These species were selected based on their potential role as a long-term solution for mitigating erosion in areas of marginal land proposed for new afforestation/reforestation and as an important carbon (C) sink. Results: Root collar diameter (RCD) provided the best fit for tree height, total above-ground biomass (AGB) and total below-ground biomass (BGB), and all regressions were highly significant (P = 0.001). Most species showed significant increases in annual growth and, by year 5, the BGB ranged between 21 and 42% of total biomass and decreased with increasing plant age. Of the conifers, Podocarpus totara had the greatest mean maximum root spread (2.2 m) exceeded only by the broadleaved Vitex lucens (2.5m).Forallspecies,andin each year of the trial, 100% of the BGB remained confined to within 0.5 m of the ground surface. With the exception of Vitex lucens and Podocarpus totara, > 90% of the total root length remained within a 0.5-m radius of the root bole. The species-specific mean tree biomass of 5-year-old plants ranged from 0.32 to −1 4.28 kg plant . A mixed-species forest established at 1000 stems per hectare (spha), consisting of 200 of −1 each of the best performed of the trialled species, would amass ~ 2.3 t ha of biomass and a forest carbon −1 stock of 3.8 t CO ha within 5 years. Conclusions: Inter-species differences in the allocation of BGB and AGB appeared to be age dependent. The root-growth metrics of these common indigenous forest species, as candidates for erosion control, have improved our understanding of their potential usefulness for stabilising marginal land. Whole-plant biomass of juvenile trees will greatly improve the accuracy of current estimates of forest carbon stocks for proposed new areas of indigenous afforestation/reforestation. Keywords: Biomass, Root systems, Conifer and broadleaved species, 5-year plot-based trial * Correspondence: mardenm@landcareresearch.co.nz Landcare Research, NZ, Ltd, PO Box 445, Gisborne 4010, New Zealand Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 2 of 26 Background Wardle 2002). The resultant forests typically comprise Before human settlement, New Zealand was almost long-lived conifers and broadleaved species that can completely vegetated with indigenous evergreen podo- achieve large stature and an active growth phase that ex- carp (mainly conifer) and broadleaved forest in the low- tends from 150 to 500 years (Hall 2001, Hinds and Reid lands (~ 200 m), transitioning to Nothofagus forest at 1957). As suggested by Hasselmann (1997), active net elevations above 600 m to the tree line at 1050 m and carbon accumulation over such time frames is consistent alpine-subalpine shrubland and grasslands on the high- with the prolonged effort likely required to effect signifi- est parts of the axial ranges (McGlone 1988; Wilmshurst cant reductions in atmospheric carbon dioxide (CO ) 1997). However, since Maori (beginning ~ 600 years BP) levels. and then European settlement (starting in the 1820s), Given the pressing need to quantify carbon fluxes as- much of New Zealand’s indigenous forest has been sociated with terrestrial vegetation dynamics, an increas- cleared for conversion to pastoral use, for the develop- ing number of researchers have sought to improve ment of a forest industry involving non-indigenous spe- estimates of tree volume, biomass and carbon stocks. cies and/or for urban development. By the time the first Many New Zealand-based studies have attempted to National Forest Survey began in 1945, only 25% of indi- quantify the biomass of indigenous forest stands, mainly genous forest remained on a total land area of 26.6 mil- for live-tree carbon sequestration (Carswell et al. 2012; lion hectares, mostly in mountainous (difficult to farm) Scott et al. 2000; Trotter et al. 2005; Beets et al. 2014; areas (Masters et al. 1957). In more recent years, the on- Schwendenmann and Mitchell 2014; Dale 2013 ). These going clearance of vegetation has led to the progressive use allometric functions based largely on figures derived loss of biodiversity and degradation of waterways and for individual specimens of a single species such as water quality through increased erosion, sedimentation kānuka (Kunzea ericoides var. ericoides (A. Rich.) J. and nutrient pollution. Thompson), and, where available, published values for a An increase in awareness of the poor health of New limited number of other species (Russo et al. 2010). The Zealand’s remaining areas of indigenous forest (as well few available biomass relationships are for a handful of as public and government pressure to maintain and en- old-growth indigenous tree species from a subset of New hance the indigenous biodiversity of New Zealand) (Par- Zealand forest classes and from a limited number of sites 3 4 liamentary Commissioner for the Environment 2002) (Beets 1980; Allen et al. 1998 ; Peltzer and Payton 2006 ; has encouraged reversion of cleared land back to forests. Beets et al. 2008 ). Thus, where species-specific and/or The planting of indigenous trees has become an integral regional values remain unavailable, congeneric values part of indigenous-forest restoration efforts. Further- have been used instead or in their absence, the mean more, since ratifying the Kyoto Protocol (Intergovern- of all published values (Beets et al. 2008). Further- mental Panel on Climate Change (IPCC) 2000; New more, the majority of these studies were of mature Zealand Climate Change Office 2003), New Zealand pro- trees though the age of the sample trees was not posed to reduce its carbon emission target in the short always determined. to medium term through afforestation/reforestation. Tree allometric equations are critical tools for deter- Exotic forests established on former grasslands since mining tree volume, biomass and carbon stocks and 1990 are likely to provide the major carbon offsets, but have the potential to improve our understanding about it has been recognised that an important additional car- carbon sequestration in woody vegetation to support the bon sink could be created through afforestation/refores- implementation of policies and mechanisms designed to tation of steep, erosion-prone pastoral hill country that mitigate climate change (Jara et al. 2014). In New environmentally is marginal for long-term agriculture by Zealand, substantial gains have been made in improving using indigenous shrubs and trees (Tate et al. 2000; Tate the accuracy of the above-ground live tree carbon stored et al. 2003, Trotter et al. 2005). Relatively little is known in the most widely distributed forest and shrubland about carbon sequestration by indigenous shrub species, ecosystems following the establishment of a nationwide but the density and diversity of species typical of natur- network of permanent sample plots, (Carswell et al. 2012; ally reverting stands suggests that relatively high levels Scott et al. 2000; Trotter et al. 2005; Peltzer and Payton of carbon storage are achievable. A range of additional 2006; Beets et al. 2014). Nonetheless, as identified by Pelt- objectives in sustainable environmental management will zer and Payton (2006), there remains a growing need for also be achievable. These include improving indigenous further age-specific and species-specific allometric rela- biodiversity, erosion mitigation and soil conservation tionships particularly for the 20 most abundant species and consequent improvements in water quality. Import- that comprise 90% of the total live-tree carbon in antly, from a carbon-sink perspective, indigenous shrubs old-growth forests. However, the below-ground biomass provide the first step to a successional pathway to a per- component has proven difficult to quantify for old-growth manent cover of indigenous tall forest (Hall 2001; trees. In particular, little is known about root allocation Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 3 of 26 except that there are interspecies differences in root distri- damage to infrastructure and the degradation of stream bution. Each species is known to allocate differing propor- habitats. However, more fundamental knowledge is re- tions of their total biomass to roots at different stages of quired on the differences in relative growth rates be- growth, but overall, species-specific allocations of dry tween indigenous shrubland, conifer and broadleaved matter to roots are generally poorly documented (Korner species and more specifically on the contribution of their 1998). Other than destructive sampling, there are few root systems towards shallow soil reinforcement, espe- techniques with which to assess either the below-ground cially during their formative years. Such data will assist biomass component or the architectural differences in in evaluating the potential effectiveness of various root system attributes between species. The use of species both as plantings on marginal land degraded by non-invasive techniques such as ground-penetrating radar erosion and as a longer-term land-use option for areas (GPR) is limited to coarse-root systems (Hruska et al. retired from exotic forest production. As is the case for 1999). The fine-root components are particularly difficult old-growth indigenous forests, such data for the and expensive to measure accurately. Several New early-growth period of the majority of indigenous shrub Zealand and international studies have shown that roots and forest species have not been collected to date. Such comprise approximately 20% of the above-ground live tree data are essential for the development of allometric biomass (Watson and Tombleson 2002; Beets et al. 2007). relationships for measuring the live-tree carbon seques- The precision of these root estimates, however, tend to be tration for: (i) planted areas of juvenile shrubs and low (Hall et al. 2001) but have nonetheless been used in mixed species of indigenous conifer and broadleaved the absence of better species- and age-specific data. Fur- forests; (ii) areas of abandoned pasture currently in the thermore, studies that have attempted to extract the ma- juvenile stage of reverting to shrubland; and (iii) areas of jority of the below-ground root biomass often exclude the exotic forest with ‘high environmental risk’ and likely to stump (Will 1966; Heth and Donald 1978; Watson and be converted (planted or by natural reversion) to a O’Loughlin 1990) and therefore underestimate the longer-term, and more sustainable, mix of indigenous below-ground to above-ground biomass ratio. Previous shrubland and conifer/broadleaved forest. authors (e.g. Cairns et al. 1997;Coomes et al. 2002)have This study is the first known attempt to systematically suggested a value of 20% can be applied generally, but quantify, as a time series, both the above- and more work is needed before this figure can be accepted below-ground growth metrics for eight of New Zealand’s internationally. most common indigenous conifer and broadleaved forest Nationally, about 1.45 million hectares (Mha) of mar- species during their early growth period (years 1 to 5). ginal pastoral land are suitable for afforestation/refores- The goals of this study were to: (i) develop species-specific tation by indigenous shrubs or forests (Trotter et al. allometric equations to better estimate above- and 2005). The enactment of the Permanent Forest Sink Ini- below-ground live-plant biomass/carbon storage for areas tiative (PFSI) and Afforestation Grant Schemes (AGS) in of newly planted marginal land established as a pathway 2008 and 2015, respectively (Ministry for Primary Indus- to a permanent indigenous forest and long-term carbon tries 2015), and the introduction of forestry to the Emis- sink; and (ii) provide the first-ever description and classifi- sions Trading Scheme (ETS) have strongly incentivised cation of their root morphology and architecture and the planting in new areas of indigenous shrublands measurements of their root metrics (biomass, length and and/or successional tall forest species as a means of fur- the spatial distribution of roots with increasing distance ther reducing greenhouse gas emissions and ensuring from and below the root bole). Such data form the basis that New Zealand meets both its short- and long-term for assessing their potential effectiveness as an alternative obligations under the Framework Convention for land-use option in areas degraded by erosion and deemed Climate Change (FCCC). Given the current interest in marginally sustainable for pastoralism and/or for areas the retirement of economically and environmentally un- currently unsustainable as short-term rotation exotic sustainable pastoral steep lands, the planting of succes- forest. sional indigenous forest species with a longer rotation This paper does not address: (i) the effect of contain- and higher wood value may become an alternative and erised versus bare-rooted seedlings on plant growth; nor viable land use option for these areas. Furthermore, a (ii) the impacts of the wider range of ecological and en- species change to indigenous shrubs and/or forest trees vironmental stresses known to influence plant develop- could potentially alleviate issues currently associated ment (e.g. competition for nutrients and water, soil with harvesting non-native species (predominantly Pinus texture and fertility; water-logging; drought; elevation; radiata D.Don) that were originally established for climate regime; exposure to wind and sun; shade; and erosion control, particularly in terrain with a high risk of grass suppression) since it is a site-specific study; nor storm-induced landslides. Issues include excessive (iii) silviculture and management practices other than amounts of sediment and associated woody debris, planting density requirements needed to achieve effective Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 4 of 26 mitigation of erosion. Information on ecological and site and in block three (plants to be extracted 3 years after requirements and growth parameters of mature trees of planting, i.e. at age 5) at 1.5–2.0-m spacing to allow for the selected species are provided in Additional file 1 and increases in canopy and root growth during the course has been sourced from previous research by Foweraker of the trial. (1929), McSweeney (1982), Pollock (1986), Salmon (1986), Poole and Adams (1994) and Wardle (2002). The ecology, Species selection establishment, growth and management of the studied The species chosen (Table 1) include those that succeed species for wood production is adequately discussed by early colonising species to become a significant compo- Bergin (2003), Bergin and Steward (2004), Bergin and Gea nent of areas of mature coniferous/broadleaf forests (2005, 2007) and Steward and Beveridge (2010). found in most regions throughout the North Island of New Zealand (Allan Herbarium 2000). The conifers Methods (softwoods) selected included rimu (Dacrydium cupressi- Study site num (Lambert)), totara (Podocarpus totara (D.Don)), The trial site was located on a low-lying, even-surfaced miro (Prumnopitys ferruginea (D.Don) Laubenf.) and alluvial terrace adjacent to the Taraheru River, in kahikatea (Dacrycarpus dacrydioides (A.Rich.) Laubenf.) Gisborne City (Fig. 1). The same site has also been used because they are among the most abundant species that to measure ‘plant growth performance’ of: (i) 12 early collectively contribute 90% of New Zealand’s total (colonising) indigenous species considered typical of ri- live-plant carbon by volume (Peltzer and Payton 2006). parian margins (Marden et al. 2005); (ii) different clones Together with matai (Prumnopitys taxifolia (D.Don) of poplar and willow (Phillips et al. 2014); and (iii) a Laubenf.), kauri (Agathis australis (D.Don) Lindley) and range of exotic forest species (Phillips et al. 2015). Tem- the broadleaved species titoki (Alectryon excelsus (Gaert- peratures over summer average 23 °C and over winter ner)) and puriri (Vitex lucens (Kirk)), these are among 12 °C, and mean annual rainfall is 1000 mm. The soil is the most common species planted in landscape restor- a naturally fertile, free-draining Typic Sandy Brown Soil ation projects (Bergin and Gea 2005, 2007). These spe- of the Te Hapara soil series (Hewitt 2010). Although not cies are considered as potential alternatives to the used, a drip-irrigation was installed to ensure survival of non-native species currently used in commercial planted the plants in the event that drought could jeopardise the forests to meet a range of economic, erosion-mitigation longer-term aims of this trial. and other environmental benefits. Advantages include: The soil has no physical or chemical impediments to tolerance to a wide range of climate, soil and topo- root development to about 1.2 m depth, other than a graphic site conditions; seedlings are widely available; variable-depth water table that fluctuates between > and their longer growing cycles are considered to be 1.5 m depth and within ~ 0.2 m of the surface (Phillips more sustainable than faster-growing non-native et al. 2014). species. Trial design Data collection The site (50 × 20 m) was tilled, and weed mat was laid A group of 1- and 2-year-old plants were removed from down before planting in autumn 2006 (Fig. 1). Weed their containers and destructively partitioned into their mat reduced weed competition and minimised efforts to component parts. At the trial site, another group of extricate intertwined roots belonging to weeds from the 2-year-old seedlings were planted and a subset of plants anticipated dense and fibrous root network of the was excavated and similarly destructively partitioned on trialled species. the anniversary of planting for three successive years. An Containerised seedlings sourced from a local nursery air lance at 240 kPa was used to remove the soil sur- consisted of 1.2-L polythene bags (0.12 m deep by rounding each root system without damage or signifi- 0.10 m diameter) filled with a 50:50 mix of bark and cant loss of the fine root mass. The aim was to sample pumice. To minimise root disturbance, seedlings were ten plants per species per year, but frost killed some pur- planted with the soil attached. Root binding was minimal iri plants and stem breakage reduced the sample size of and root pruning before planting was not required as titoki plants (Table 1). tap root length did not exceed the depth of the planting Tree height, canopy spread, root collar diameter bag. The site was subdivided into three blocks and all (RCD—measured overbark at ground level) and diameter three blocks were planted in 1 day. A randomised block at breast height (DBH—measured at 1.4 m above-ground design was used with equal numbers of each species in level) (where applicable) were measured before removing each block. Trees within blocks 1 and 2 (plants to be ex- the plants from the trial site (Additional file 2). Canopy tracted 1 and 2 years after planting, i.e. at age 3 and spread was taken as the mean of the diameter mea- 4 years old, respectively) were planted at 1-m spacing, sured in two directions at right angles. Trees were Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 5 of 26 Fig. 1 Location and layout of the trial site cut at ground level and the above-ground of the maximum root length measured from root tip to components separated into branches, foliage and stem root tip in two directions at right angles to each other. (Additional file 3). Root system extraction and meas- Root length (excluding fibrous roots) within each of the urement methods followed well-established procedures root diameter size classes was measured by laying pieces (e.g. Watson et al. 1999;Mardenetal. 2005; Phillips of root end for end alongside a tape measure (Watson et al. 2014, 2015). Once removed from the ground, et al. 1999). The mass of the roots and root bole were the root system of each plant was washed to remove calculated separately (Additional file 5). Root distribution adhering soil matter then photographed (an example by mass (Additional file 6) and length (Additional file 7) is shown in Fig. 2). It was then described using the relative to the root bole was obtained by systematically classification of Hinds and Reid (1957) before being dissecting the root systems into 0.5-m radial and depth partitioned into root bole (stump) and roots. segments then sorted by diameter size classes, (< 1 mm Below-ground growth parameters included mean root fibrous, 1–2, 2.1–5.0, 5.1–10.0 and 10.1–20.0 mm over depth and mean root spread of the lateral roots, here- bark) (Watson and O’Loughlin 1990). after referred to as root depth and root spread All plant components, both above- and below-ground, (Additional file 4). The latter was taken as the mean were oven-dried at 80 °C for 24 h or until no further Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 6 of 26 Table 1 Indigenous conifers and broadleaved species trialled and number of trees extracted each year of the trial Botanical name Common Root Number of plants extracted/species/year name system Plant age (years) 1 2 3 4 5 Species total type Coniferous species Agathis australis Kauri Tap 10 10 10 10 10 50 Dacrycarpus dacrydioides Kahikatea Tap 10 10 10 10 10 50 Dacrydium cupressinum Rimu Tap 10 10 10 10 10 50 Podocarpus totara Totara Tap 10 10 10 10 10 50 Prumnopitys ferruginea Miro Plate 10 10 10 10 10 50 Prumnopitys taxifolia Matai Plate 10 10 10 10 10 50 Broadleaved species Alectryon excelsus Titoki Plate 10 10 10 10 9 49 Vitex lucens Puriri Tap 10 10 10 6 5 41 Annual total 80 80 80 76 74 390 weight loss was detectable then weighed to the nearest Statistical analyses 0.1 g. The biomass and root:shoot ratio were calculated ANOVA was used to determine the effects of year and using dry mass. A carbon content of 50% of biomass is species on above- and below-ground parameters. The considered robust (Coomes et al. 2002). A conversion Student-Newman-Keuls post hoc analysis was used to de- factor of biomass × 1.65 (Williams 1978) was used to cal- termine differences among the species within a year and culate carbon stocks (t CO ). Carbon stock values do among years when all species were averaged within each not include woody debris or fine litter. year. The normality of each analysis was determined by Fig. 2 Vertical and plan views of 5-year-old Agathis australis (a) and Alectryon excelsus (b) root systems Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 7 of 26 visual assessment of residual plots. The root biomass and curves for roots with > 1- and > 2-mm diameter (P = root length data at 0.5-m radial intervals exhibited 0.741). Therefore, regressions based on roots > 1-mm non-normality due largely to the prevalence of radial seg- diameter were used for analyses of total root length and ments in which no or little root material was recorded for total root biomass. Unless otherwise stated, data are pre- some of the root diameter size classes. The radially seg- sented as a mean with standard errors, and statistical ana- mented root biomass and length data were therefore lyses were considered significant if P ≤ 0.05. transformed using the equation log (x + c), where x was Differences between the grouped conifers versus the root biomass or root length and c was 1. broadleaved species were not examined because they dif- Type-1 linear regression was used to fit RCD (over- fer significantly in tree form; there were clear overlaps in bark) and tree height data using the equation (y = mx + above- and below-ground metrics between species c). RCD was used instead of DBH in the analysis of tree across these groupings and because of the likely bias in allometry because many of the species were not ex- analyses resulting from too few and/or uneven sample pected to reach DBH height (1.4 m) for some years after numbers. the establishment of the trial. More importantly, it may serve as a better predictor of both above-ground and Results below-ground biomass quantities as root collar is mea- Allometric relationships sured at the intersection of above- and below-ground Tree height parameters and as stem taper in seedlings is minimal Linear regression fitted the RCD and tree height data well, (Wagner and Ter-Mikaelian 1999). The relationship be- with r ranging between 0.768 (Vitex lucens) and 0.916 tween RCD and foliage, branches and stem (Table 2), (Dacrydium cupressinum), and all regressions were highly total above-ground biomass (AGB), total below-ground significant (P < 0.001; Fig. 3). There were significant biomass (BGB), total root biomass (> 1- and > 2-mm differences among species with respect to height in each diameter) and total root length (> 1- and > 2-mm diam- year (Additional file 2). With the exception of eter) (Table 3) was analysed using two-parameter expo- Prumnopitus taxifolia, which had the same average height nential growth analysis using the equation y = a exp (bx), in years 1 and 2 (P = 0.05: data not shown), the tree height where b represented the slope of the regressions. The re- of the remainder of species increased significantly from gression analyses include data for the containerised 1- year to year. There was, however, no clear pattern as to and 2-year-old seedlings partitioned at the time the trial which species had performed best until year 5 when site was planted. All regression analyses were undertaken Prumnopitys taxifolia (1.3 m), Agathis australis (1.6 m) using SigmaPlot 12.5 (Systat Software, San Jose, CA). and Prumnopitys ferruginea (1.6 m) were significantly Depending on the method of root extraction and focus shorter (P = 0.05) than the tallest two species Podocarpus of the research, regressions of total root biomass and totara (2.8 m) and Dacrycarpus dacrydioides (2.7 m), length are often based on different groupings of root while Vitex lucens (2.4 m), Dacrydium cupressinum diameter size classes. For comparative New Zealand stud- (2.2 m) and Alectryon excelsus (2.2 m) reached an ies, some researchers have based their regressions on roots intermediary height significantly different from the other with a > 1-mm diameter (Phillips et al. 2015)while others species but not each other (Additional file 2). have included only roots with a > 2-mm diameter (Watson and Tombleson 2002, 2004;Marden et al. 2016). In this Root collar diameter (RCD) study, when regressed against RCD, a two-tailed, unpaired Differences in RCD among species in years 1–3 were t test found no significant difference in the slopes of the variable, with no consistent species-specific trends other Table 2 Allometric relationships between RCD against stem, branch and foliage biomass Species Stem Branch Foliage 2 2 2 r P Equation r P Equation r P Equation Agathis australis (kauri) 0.822 < 0.001 y = 6.958 exp (0.114x) 0.744 < 0.001 y = 2.793 exp (0.111x) 0.799 < 0.001 y = 5.661 exp (0.106x) Dacrycarpus dacrydioide (kahikatea) 0.942 < 0.001 y = 25.937 exp (0.058x) 0.908 < 0.001 y = 16.919 exp (0.064x) 0.941 < 0.001 y = 12.900 exp (0.062x) Dacrydium cupressinum (rimu) 0.928 < 0.001 y = 17.665 exp (0.076x) 0.798 < 0.001 y = 13.086 exp (0.074x) 0.887 < 0.001 y = 23.053 exp (0.083x) Podocarpus totara (totara) 0.930 < 0.001 y = 39.461 exp (0.052x) 0.699 < 0.001 y = 35.461 exp (0.058x) 0.788 < 0.001 y = 51.688 exp (0.049x) Prumnopitys ferruginea (miro) 0.899 < 0.001 y = 10.667 exp (0.104x) 0.868 < 0.001 y = 8.871 exp (0.115x) 0.880 < 0.001 y = 9.919 exp (0.108x) Prumnopitus taxifolia (matai) 0.775 < 0.001 y = 7.987 exp (0.098x) 0.790 < 0.001 y = 26.804 exp (0.106x) 0.760 < 0.001 y = 5.446 exp (0.106x) Alectryon excelsus (titoki) 0.840 < 0.001 y = 31.875 exp (0.065x) 0.840 < 0.001 y = 14.277 exp (0.0812x) 0.734 < 0.001 y = 26.199 exp (0.052x) Vitex lucens (puriri) 0.786 < 0.001 y = 48.929 exp (0.035x) 0.948 < 0.001 y = 15.611 exp (0.049x) 0.950 < 0.001 y = 25.903 exp (0.042x) Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 8 of 26 Table 3 Allometric relationships between RCD and total root length, and total root biomass for roots > 1 mm and > 2 mm diameter Species Root size class Root length Root biomass 2 2 r P Equation r P Equation Agathis australis (kauri) > 1 mm 0.780 < 0.001 y = 0.542 exp (0.113x) 0.879 < 0.001 y = 1.173 exp (0.137x) > 2 mm 0.758 < 0.001 y = 0.161 exp (0.127x) 0.860 < 0.001 y = 0.879 exp (0.144x) Alectryon excelsus (titoki) > 1 mm 0.797 < 0.001 y = 8.179 exp (0.057x) 0.851 < 0.001 y = 17.629 exp (0.062x) > 2 mm 0.763 < 0.001 y = 1.842 exp (0.057x) 0.851 < 0.001 y = 11.206 exp (0.066x) Dacrycarpus dacrydioide (kahikatea) > 1 mm 0.783 < 0.001 y = 12.159 exp (0.045x) 0.891 < 0.001 y = 12.816 exp (0.053x) > 2 mm 0.704 < 0.001 y = 3.090 exp (0.047x) 0.897 < 0.001 y = 8.136 exp (0.057x) Dacrydium cupressinum (rimu) > 1 mm 0.816 < 0.001 y = 2.743 exp (0.080x) 0.925 < 0.001 y = 3.514 exp (0.098x) > 2 mm 0.829 < 0.001 y = 0.527 exp (0.088x) 0.946 < 0.001 y = 1.793 exp (0.106x) Podocarpus totara (totara) > 1 mm 0.461 < 0.001 y = 11.481 exp (0.040x) 0.730 < 0.001 y = 15.073 exp (0.049x) > 2 mm 0.659 < 0.001 y = 1.921 exp (0.046x) 0.780 < 0.001 y = 8.786 exp (0.053x) Prumnopitys ferruginea (miro) > 1 mm 0.831 < 0.001 y = 2.731 exp (0.095x) 0.909 < 0.001 y = 3.627 exp (0.114x) > 2 mm 0.855 < 0.001 y = 0.357 exp (0.112x) 0.931 < 0.001 y = 1.673 exp (0.127x) Prumnopitus taxifolia (matai) > 1 mm 0.779 < 0.001 y = 2.699 exp (0.101x) 0.789 < 0.001 y = 4.321 exp (0.102x) > 2 mm 0.515 < 0.001 y = 0.359 exp (0.124x) 0.808 < 0.001 y = 2.204 exp (0.111x) Vitex lucens (puriri) > 1 mm 0.812 < 0.001 y = 13.037 exp (0.037x) 0.886 < 0.001 y = 25.990 exp (0.038x) > 2 mm 0.723 < 0.001 y = 4.002 exp (0.033x) 0.878 < 0.001 y = 21.548 exp (0.038x) than Vitex lucens, which, with the exception of year 4, ranging between 0.803 (Agathis australis) and 0.944 had a larger RCD than the remaining species. In years 4 (Dacrycarpus dacrydioides), and all regressions were and 5, Agathis australis, Prumnopitys taxifolia and highly significant (P < 0.001) (Fig. 4). There was, Prumnopitys ferruginea had a significantly smaller (P = however, no significant difference in mean total AGB 0.05) RCD (~ 20–24 mm) than Dacrydium cupressinum among species until years 4 and 5, the exception being and Alectryon excelsus (~ 37–39 mm) (Additional file 2). Alectryon excelsus which showed no further increase in Significantly larger still were Podocarpus totara and mean total AGB (P = 0.05; data not shown). By year 5, Dacrycarpus dacrydioides, with a similar RCD as each the mean total AGB amassed by Agathis australis other (~ 48–50 mm), while Vitex lucens was the best (251.1 g) and Prumnopitys ferruginea (381.4 g) were not performed of the trialled species with an RCD of significantly different from each other nor from 77.6 mm, almost twice that of the next best performed Prumnopitys taxifolia (581.0 g). However, their biomass species and more than three times that of the slowest was significantly less (P = 0.05) than that of Alectryon growing Prumnopitys ferruginea (Additional file 2). excelsus (1092.3 g), Dacrydium cupressinum (1165.2 g) and Dacrycarpus dacrydioides (1231.0 g) and was Diameter at breast height (DBH) exceeded only by Podocarpus totara (2142.9 g) and Vitex There were significant differences among species with lucens (3026.0 g) (Fig. 5, Additional file 3). respect to DBH in year 5 (Additional file 2). The DBH of Prumnopitys taxifolia and Prumnopitys ferruginea were Biomass allocation to foliage, branches and stem not significantly different from each other (2–3.9 mm) Two-parameter exponential regression analysis was a but Prumnopitys taxifolia had a significantly smaller (P good fit for RCD assessed against stem (r =0.775–0.944), 2 2 = 0.05) DBH than did Agathis australis (8.0 mm), Dacry- branches (r =0.699–0.948) and foliage (r =0.734–0.950) dium cupressinum (9.8 mm) and Alectryon excelsus biomass, and all regressions were highly significant (P < (11.2 mm). Dacrycarpus dacrydioides (16.8 mm), Podo- 0.001) (Table 2). There were significant differences (P = carpus totara (21.5 mm) and Vitex lucens (23.5 mm) 0.05) among species in the allocation of biomass between each had a significantly larger (P = 0.05) DBH than did foliage, branches and stem in each year of the trial the other species but not compared with each other (Additional file 3). At 5 years of age, Agathis australis, (Additional file 2). Dacrycarpus dacrydioides, Prumnopitys ferruginea, Alectryon excelsus and Vitex lucens had the highest Above-ground biomass (AGB) proportion of their total AGB allocated to the stem (35– Two-parameter exponential regression analysis was a 49%), branches (19–40%) and least to foliage (20–32%). good fit for the RCD and total AGB data, with r Exceptions to this general trend were Dacrydium Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 9 of 26 Agathis australis Alectryon excelsus y=0.061x+0.085 y=0.053x+0.150 2 2 r =0.845; P<0.001 r =0.901; P<0.001 Dacrycarpus dacrydioide Dacrydium cupressinum y=0.052x+0.312 y=0.051x+0.366 2 2 r =0.906; P<0.001 r =0.916; P<0.001 Podocarpus totara Prumnopitys ferruginea y=0.048x+0.538 y=0.061x+0353 2 2 r =0.863; P<0.001 r =0.793; P<0.001 Prumnopitus taxifolia Vitex lucens y=0.041x+0.290 y=0.024x+0.368 2 2 r =0.849; P<0.001 r =0.768; P<0.001 Root collar diameter (mm) Fig. 3 Linear regression analysis of RCD and tree height. Note that the scale on the y and x axes differs between species cupressinum with a disproportionate 51% of its total AGB The differences in mean total BGB among species in 1- allocated to foliage comprising scale leaves closely and 2-year-old seedlings were initially small, but it be- appressed to the stem (Wardle 2002). Also, for came more significant in years 4 and 5 (P = 0.05; data Prumnopitys taxifolia and Podocarpus totara with deep, not shown), when the majority of species showed a sig- narrow crowns and short lateral branches, 69 and 37% of nificant increase in mean total BGB. Exceptions to this their respective total AGB was allocated to branches general pattern include Agathis australis, with no in- (Additional file 3). crease in mean total BGB in years 1 and 2, a modest in- crease in year 3 and no increase in year 4, and in year 5, Below-ground biomass (BGB) the increase in mean total BGB was significantly greater Similarly, two-parameter exponential regression analysis than in any of the previous years. By year 5, Agathis aus- was a good fit for the RCD and mean total BGB data, tralis (72.0 g) had a significantly less (P = 0.05) mean with r ranging between 0.757 (Podocarpus totara) and total BGB than both Prumnopitys ferruginea (134.9 g) 0.946 (Dacrydium cupressinum), and all regressions and Prumnopitys taxifolia (134.6 g). For the remaining were highly significant (P < 0.001) (Fig. 6). Additionally, species, the mean total BGB of Dacrycarpus dacry- there were significant differences (P = 0.05) between dioides (371.2 g), Dacrydium cupressinum (397.4 g), species in the allocation of biomass between root mass Alectryon excelsus (402.2 g) and Podocarpus totara and root bole in each year of the trial (Additional file 5). (450.0 g) was significantly different again but not to each Tree height (m) Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 10 of 26 Fig. 4 Exponential growth analysis of RCD and total above-ground biomass. Note that the scale on the y and x axes differs between species other. Vitex lucens (1257.4 g) had the highest mean total Podocarpus totara (103.0 g) and Dacrydium cupressinum BGB and was significantly different (P = 0.05) from all (102.6 g). In contrast, the mean total root bole biomass the other species (Fig. 5, Additional file 3). of the conifers Agathis australis (17.2 g), Prumnopitys ferruginea (22.2 g) and Prumnopitys taxifolia (22.9 g), Biomass allocation to root and root bole though not significantly different from each other, are For the majority of species, there were no significant each an order of magnitude less than for the remainder interspecies differences in the allocation of mean root of the conifers and broadleaved species (Additional file 5). and root bole biomass until years 4 and 5, at which time Vitex lucens also had the greatest root biomass (970.1 g, both root and root bole biomass increased significantly 79.3% of total BGB), while Agathis australis had the (P = 0.05; data not shown), similar to the pattern for the least (54.8 g, 75.2% of total BGB). For all species, the allocation of mean total BGB and AGB. For Agathis proportion of the total BGB allocated to the roots australis, the increase in root and root bole mass oc- (69–84%) was greater than that allocated to the root curred in year 3, whereas for Prumnopitys ferruginea bole (16–31%). and Vitex lucens, it did not increase until year 5. By year 5, Vitex lucens had the largest root bole biomass Root biomass distribution by depth and radius (287.3 g) but was not significantly different to Dacrycar- For all species in each year of the trial, 100% of the total pus dacrydioides (112.7 g), Alectryon excelsus (105.8 g), BGB (P < 0.001) was confined to the 0–0.5-m depth Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 11 of 26 and Alectryon excelsus occurred within a 0.5-m radius of the root bole, Podocarpus totara 76.8% and Vitex lucens 65.9% (Fig. 7, Additional file 6). Root biomass allocation by root diameter size class Exponential regression analysis of RCD and total root biomass for roots > 1- and > 2-mm diameter size class showed very similar regressions (Fig. 8), with RCD explaining between 73 and 92% of the variation in root biomass for roots > 1 mm and between 78 and 94% for roots > 2 mm (Table 3). In year 5, fibrous roots accounted for the highest pro- portion of the total root mass of each of the plate-rooted species Alectryon excelsus (23.5%), Prumnopitys ferrugi- nea (56.4%) and Prumnopitys taxifolia (49.7%) (Fig. 9). Similarly, for the tap-rooted species, fibrous roots accounted for between 28.1% (Dacrycarpus dacry- dioides) and 49.3% (Dacrydium cupressinum) of their total root mass. For all species, the percentage of fibrous Fig. 5 Plant biomass above- (dark bars) and below-ground (light bars) root biomass decreased as root sizes increased with in- in year 5. Error bars represent one standard error of the mean. Bars creasing plant age. The largest root diameter size class within each parameter with a different letter were significantly different (P= 0.05) (50–100 mm) was recorded for Vitex lucens only and comprised < 2% of its total root biomass. For the re- mainder of the species, the largest root diameter size increment, and inter-species differences (P = 0.05) in the class was 20–50 mm and comprised just 0.1% of Prum- mean maximum rooting depth were minimal. For most nopitys ferruginea and 6.8% of the total root mass species, this had increased from ~ 0.1 m in year 1 to ~ (largely tap root) of Agathis australis. 0.3 m by year 5 (Additional file 4). For any individual seed- ling, the tap-rooted Agathis australis developed the BGB/AGB ratio deepest (0.7 m) root, while the plate-rooted systems of In year 1, although the BGB/AGB ratio was highest for Prumnopitys taxifolia, Prumnopitys ferruginea and Alectr- the tap-rooted Vitex lucens (0.89) and lowest for the yon excelsus had the shallowest root systems (~ 0.15 m). plate-rooted Prumnopitys taxifolia (0.25), there was no By year 5, the mean maximum root depth of each of the consistent relationship between this ratio and root type trialled species was similar, whether conifers or broad- from year to year (Additional file 3). By year 5, Podocar- leaved species and tap-rooted or plate-rooted species pus totara (0.21) and Prumnopitys taxifolia (0.23) had (Additional file 4). the lowest BGB/AGB ratio, and although for Vitex Similarly, for the first 3 years, 100% of the total BGB (P < lucens the BGB/AGB ratio remained the highest of the 0.001) of all species was confined to within a 0.5-m radius trialled species (Fig. 5), it had halved to 0.42 by year 5. of the root bole. By year 4, only Dacrycarpus dacrydioides, Similarly, a decline in BGB allocation with increasing Alectryon excelsus and Vitex lucens hadextendedtheir age was apparent in Podocarpus totara (0.47 in year 1, roots to a radius of between 1.0 and 1.5 m from the root decreasing to 0.21 in year 5) and Prumnopitys ferruginea bole (Additional file 7). The root mass and percentage of (0.5 in year 1, decreasing to 0.35 in year 5). Averaged total root mass of Dacrycarpus dacrydioides (0.8 g, 0.4%) across all eight species, the BGB and AGB was 0.32 and and Alectryon excelsus (0.9 g, 0.2%) at this distance were 0.68, respectively. not significantly different from each other but were signifi- cantly different from Vitex lucens (3.4 g, 1.7%). Vitex lucens Root length and root system dimensions (diameter) was the only species by year 5 to have extended its root net- Exponential growth analyses of RCD and total root work to a radius of between 1.5 and 2.0 m from the root length fitted least well for Podocarpus totara with RCD bole; however, the roots comprised only 8.3 g, 0.8% of its explaining 46 and 66% of the variation in root length for total root mass. Between years 1 and 5, Agathis australis roots > 1 and > 2 mm, respectively. In contrast, this rela- was the only species to have 100% of its root mass confined tionship fitted best for Prumnopitys ferruginea with RCD to within 0.5 m of the root bole. At year 5, > 90% of the explaining between 83 and 85% of the variation in root total root mass of Dacrycarpus dacrydioides, Dacrydium length for roots > 1 mm and > 2 mm, respectively. cupressinum, Prumnopitys taxifolia, Prumnopitys ferruginea (Table 3, Fig. 10). Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 12 of 26 Fig. 6 Exponential growth analysis of RCD and total below-ground biomass. Note that the scale of the y and x axes differs between species At year 1, none of the species had produced roots > excelsus and Vitex lucens hadextendedtheir roots intothe 1-mm diameter. By year 2, Alectryon excelsus and Vitex 0.5–1.0-m radial sector where Vitex lucens had a signifi- lucens each produced > 2.5 m of root, significantly more cantly greater proportion (P = 0.05) of its total root length (P = 0.05) than Dacrycarpus dacrydioides (~ 2 m), which (34%)atthis radialdistancethandid Dacrycarpus in turn was significantly more than that of each of the dacrydioides (19%) and Alectryon excelsus (17%). remaining species at < 1 m of root length (Additional file 4). Similarly, at 1.0–1.5-m radial distance from the root By year 5, the total root length for each species increased bole, Vitex lucens had significantly more (P = 0.05) of substantially and differences between species became its total root length (~ 5%) than did Dacrycarpus more pronounced. For example, Agathis australis pro- dacrydioides (1%) and Alectryon excelsus (0.4%). duced the least mean total root length (9.01 m) while For all species, in years 1 and 2, there was little Vitex lucens (347.45 m) at the same age produced two or- inter-species difference in mean root system dimensions ders of magnitude more root length (Additional file 4). (i.e. mean maximum root spread). However, by year 5, of In years 2 and 3 of the trial, and for each of the species the broadleaf species, Vitex lucens had developed a sig- trialled, 100% of the total root length occurred within a nificantly greater (P = 0.05) mean maximum root spread 0.5-m radius of the root bole, and for Agathis australis,this (2.5 m) than had Alectryon excelsus (1.6 m), and its root was the case through to year 5 (Fig. 11, Additional file 7). spread was also significantly greater than that of the lar- By year 4, only Dacrycarpus dacrydioides, Alectryon gest of the conifer root systems, i.e. that of Podocarpus Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 13 of 26 Agathis australis Alectryon excelsus Dacrycarpus dacrydioide Dacrydium cupressinum Prumnopitys ferruginea Podocarpus totara Vitex lucens Prumnopitus taxifolia Distance from the root bole (m) Fig. 7 Distribution of total root biomass with increasing radial increments from the root bole. Note that the scale on the y axis differs between species totara (2.2 m) (Additional file 4). The mean maximum comprised roots in the 2–5-mm size class, 20% greater root spread of both Alectryon excelsus and Dacrycarpus than for the remainder of species (Fig. 12). The percent- dacrydioides (1.7 m), while not different from each other, age of total root length decreased with increasing root was significantly larger (P = 0.05) than the more compact size class, and for the 20–50-mm size class, where root systems of Prumnopitys taxifolia (1.2 m), Prumnop- present, roots comprised ≤ 1.5% of the total root length. itys ferruginea (1.0 m) and Dacrydium cupressinum None of the trialled species had roots in the > 50-mm (1.1 m), which in turn were not significantly different from diameter size class. each other. The mean maximum root spread of Agathis australis (0.5 m) was significantly smaller than that of the Plant biomass and carbon accumulation remainder of species trialled (Additional file 4). For each of the species trialled, the mean annual total biomass of individual trees increased significantly Root length by root diameter size class (Additional file 3). By year 5, the slowest growing of the For all the trialled species, the highest proportion of the trialled species including Agathis australis, Prumnopitys total root length of 5-year-old trees was in the 1– ferruginea and Prumnopitys taxifolia each accumulating −1 2-mm-diameter size class and with the exception of the least total biomass of ~ 0.32 kg tree (0.16 kg of C), −1 −1 Agathis australis comprised > 70% of their total root ~ 0.56 kg tree (0.26 kg of C) and ~ 0.72 kg tree length. For Agathis australis, 39% of its total root length (0.36 kg of C), respectively. Conversely, Dacrycarpus Total root biomass (g) Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 14 of 26 Agathis australis Alectryon excelsus Dacrycarpus dacrydioide Dacrydium cupressinum Podocarpus totara Prumnopitys ferruginea Prumnopitus taxifolia Vitex lucens Root collar diameter (mm) Fig. 8 Exponential growth analysis of RCD and total root biomass. Note that the scale of the y and x axes differs between species dacrydioides, Dacrydium cupressinum and Alectryon produced the longest laterals extending a distance of −1 excelsus each accumulated ~ 1.56 kg tree (0.78 kg of 2.2 m, significantly longer (P = 0.05) than all the C), while the highest biomass accumulation was by other trialled species, and of the broadleaved species, −1 Podocarpus totara ~ 2.59 kg tree (1.30 kg of C) and Vitex lucens attained the greatest mean root spread −1 Vitex lucens ~ 4.28 kg tree (2.14 kg of C) (Table 4). (2.5 m) (Additional file 4). Plate-rooted species include the conifers Prumnop- Root system types itys taxifolia and Prumnopitys ferruginea and the Two types of root systems (tap-rooted and plate-rooted) broadleaved Alectryon excelsus. Their juvenile root were observed based on the descriptive classification of systems are compact and consist largely of abundant Hinds and Reid (1957) we recognised. fibrous and numerous short, gnarly, lateral roots of Of the trialled species that developed a tap root small diameter size (< 1 mm diameter) somewhat re- (Agathis australis, Dacrycarpus dacrydioides, Vitex sembling that of Agathis australis but lacking a tap lucens, Podocarpus totara and Dacrydium cupressi- root (Fig. 2b). Of the plate-rooted species, the broad- num), the seedling radicle of Agathis australis leaved Alectryon excelsus developed the longest lateral (Fig. 2a) was prominent earliest (3-year-old plants) roots by year 5 with a spread of 1.6 m, significantly and extended deeper (0.3 m) than the remainder of longer (P = 0.05) than the other plate-rooted species the trialled species. By year 5, Podocarpus totara (Additional file 4). Total root biomass (g) Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 15 of 26 DBH is preferential to RCD as RCD can be affected by root buttress. However, RCD was used in the current study as DBH height (1.4 m) of the juvenile species trialled was not attained until year 4. Indeed, as suggested by Wagner and Ter-Mikaelian (1999), RCD may serve as a better predictor of both the above- and below-ground biomass as it is measured at the intersec- tion of the base of the stem with the root bole. In fact, RCD provided the best fit for tree height and total AGB, and all regressions were highly significant in each of the 5 years of the current trial and for all the trialled species. To date, generic rather than species-specific allometric equations have been used to estimate the AGB of ma- ture indigenous forest and shrub biomass. Extrapolating these equations to juvenile plants with smaller diameters far outside the range of data used to develop them is questionable however (Claesson et al. 2001; Zianis and Mencuccini 2003). The allometric biomass equations presented in this paper indicate that there are clear dif- ferences in growth performance among species, at least at their juvenile stage, and that different fitted coeffi- cients are required. Furthermore, there is a dearth of allometric equations for estimating the below-ground biomass of indigenous species. The absence of empirical data means that a figure of 20% is widely accepted, both locally (Beets et al. 2007) and internationally (Coomes et al. 2002) despite the precision of these root estimates tending to be low (Hall et al. 2001). Studies that include measurements of both the above- and below-ground growth attributes of the juvenile form of different species have also found RCD to be a good fit for estimating height, AGB and also BGB. For example, Marden et al. (2018; in press) found RCD to be a good fit for estimating all three attributes of 12 of New Zealand’s indigenous riparian species with each species requiring a specific fitted coefficient. Furthermore, Fig. 9 Percent root biomass (%) by diameter size class (mm) of Marden et al. (2016) found that two-parameter exponen- 5-year-old trees tial regressions using RCD was successful in predicting the BGB of five seed lots of Pinus radiata between 1 Discussion and 4 years old with RCD explaining between 93 and Plant allometry 98% of the variability in BGB. Similarly, Phillips et al. The physiological age of all the species was similar at (2015) found it possible to develop simple relationships the time of planting, but there were initial and between RCD and growth attributes with reasonable r significant differences in the above-ground metrics values for below-ground biomass and total root length. among species. Once established, interspecies differences However, year-on-year growth performance data for in year-on-year growth remained but became increas- the juvenile stage of most of New Zealand’s indigenous ingly variable. Payton et al. (2009) suggested that AGB species is particularly rare, thus limiting comparisons should be estimated using height and percentage cover with other published works. Nonetheless, there is for juvenile trees with a DBH < 50 mm. More recently, evidence to show that the proportion allocated to BGB Mason et al. (2014) used tree height and basal area to decreases with increasing age despite significant determine AGB in New Zealand shrublands while Beets year-on-year gains in total plant biomass in many spe- et al. (2012) used a two-factor exponential model which cies, both indigenous and exotic (Marden et al. 2005). uses DBH or DBH × height to estimate AGB biomass in This concurs with other local (Watson et al. 1995, 1999) mature native forests. Magalhães (2015) suggested that and international (Abernethy and Rutherfurd 2001; Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 16 of 26 Agathis australis Alectryon excelsus Dacrycarpus dacrydioide Dacrydium cupressinum Podocarpus totara Prumnopitys ferruginea Prumnopitus taxifolia Vitex lucens Root collar diameter (mm) Fig. 10 Exponential growth analysis of RCD and total root length. Note that the scale of the y and x axes differs between species Easson and Yarbrough 2002) studies where a decrease in 2005) or for 3-year-old Eucalyptus fastigata (33%), the root/shoot ratio for small saplings has been related to Sequoia sempervirons (29%), Acacia melanoxylon an increase in plant size (Kira and Shidei 1967; Walters (37%) and Pinus radiata (36%)trialledatthe same and Reich 1996) and to an increase in plant height (Cao site (Phillips et al. 2015). Although the below-ground and Ohkubo 1998). Conversely, Marden et al. (2016) biomass of each of five different 5-year-old clones of found no clear indication that the root/shoot ratio of 1- to Pinus radiata (~ 22%) established on steep hill coun- 4-year-old Pinus radiata decreased over this period but try (Marden et al. 2016)isatthe lowerend of this instead remained relatively constant. Biomass allocation to range, it is nonetheless consistent with that estab- components of woody plants is also affected by factors in- lished for similar-aged pines planted at a range of cluding plant architecture and morphology and climatic sites in New Zealand (Watson and Tombleson 2002, and edaphic factors (Ketterings et al. 2001). 2004) and likely reflects harsher site and edaphic In this study, the range of biomass apportioned to and climatic factors. below-ground components by year 5 was 23–41%, There was no consistent year-to-year difference in the not dissimilar to that of 12 early colonising indigen- root/shoot ratio between the tap- and plate-rooted ous species at the same age with a below-ground species or between conifers and broadleaved species biomass of between 24 and 44% (Marden et al. (Additional file 3). Total root length (m) Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 17 of 26 Fig. 11 Distribution of total root length with increasing radial increments from the root bole. Note that for the species with the greatest total root length there is a break in the y axis scale Root metrics and architecture 0.001) would likely remain confined to the 0–0.5-m Published accounts on the architecture of root systems depth increment, to within a 0.5-m radius of the root of New Zealand’s commonest and longest-lived native bole even in the absence of barriers to root penetration. conifer and broadleaved forest species are scarce, de- Also expected was that root biomass and root length scriptive and generally based on observations of partially would decrease with increasing depth and horizontal exposed roots undermined along river banks (Cameron distance from the root bole. Similarly, fibrous roots (as a 1963) or of root boles of windthrown trees (Bergin and percentage of total root biomass) were expected to Steward 2004). At the time of establishing this trial, decrease as root sizes increased. there were no root-related data (photographic or de- Conversely, while excavations of the root systems of scriptive) for juvenile specimens of the trialled species. juvenile Agathis australis revealed the known presence While some of the general findings on root distribution of a well-developed tap root (Bergin and Steward (2004), (depth and spread) reported here were anticipated from the presence of a tap root in Dacrycarpus dacrydioides, the results of a previous trial at the same site (Marden Vitex lucens, Podocarpus totara and Dacrydium et al. 2005), other findings were not. For example, it was cupressinum was unexpected and has not to our know- anticipated that a high proportion of the total BGB (P < ledge been previously reported. Additionally, their many Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 18 of 26 remaining devoid of root mass. This was also observed during the excavation of root systems of non-native for- est species (Phillips et al. 2015) and non-native Populus and Salix species at the same trial site (Phillips et al. 2014). These results suggest that interspecies differences in the distribution of roots and their overall root system architecture are likely to be inherent rather than due to seedling propagation, planting practices or site factors given that there were no observable differences in the physical attributes of the soil across this essentially flat terrace surface and competing influences from adjacent well-spaced plants were unlikely. Conversely, the highly asymmetric root architecture of juvenile Pinus radiata reported by Marden et al. (2016) is considered typical of tree root systems developed on steep slopes where the greatest proportion of the total root length trends in diagonally opposite directions pre- viously referred to as bilateral fan-shaped architecture (Chiatante et al. 2003). Here, the larger and stronger lat- eral roots develop in the upslope (Schiechtl 1980) under compression (Stokes and Guitard 1997) and provide the greatest contribution to tree stability/anchorage (Sun et al. 2008). Roots located in the downslope direction develop under tension, and are therefore smaller and weaker, and likely contribute less to tree anchorage. Nicoll and Ray (1996) reported that prevailing winds can influence the distribution of the structural roots of Sitka spruce (Picea sitchensis (Bong.) Carr.) where, relative to the prevailing wind, there were more roots located on the leeward side than on the windward side of trees. Marden et al. (2016) considered the development of the bilateral fan-shaped distribution of structural roots of Pinus radiata to have been influenced by the predomin- ance of northerly winds as an adaptation to improve the rigidity of the soil-root plate and to counteract their vul- nerability to windthrow. Fig. 12 Percent root length (%) by diameter size class (mm) of The openness and exposure of the conifer and broad- 5-year-old trees leaved species trial site to strong winds resulted in stem breakage of Alectryon excelsus, and Vitex lucens was leading lateral roots, though initially of small diameter, defoliated during periods of heavy frost. However, site were highly symmetric; thus, few areas of soil were left factors are unlikely to have contributed to inter-species totally devoid of any roots. These species typically had a differences in their overall growth performance as all the high number of roots per unit volume of soil and, as trialled species tolerate a wide range of environmental expected for juvenile plants, the bulk of their root mass conditions. Furthermore, with weed control, unlimited and total root length was confined to a shallow depth water availability, a fertile alluvial soil and the absence of and within close proximity to the root bole. By compari- barriers to root development, the resultant growth son, the root systems of the juvenile plate-rooted species performance could be considered as optimal for these (including the conifers Prumnopitys taxifolia and species. However, as is common with nursery-raised Prumnopitys ferruginea and the broadleaved Alectryon juvenile plants, factors such as wrenching, root excelsus) typically had fewer laterals roots but were of trimming and planter bag constriction likely stunt the larger diameter and with limited branching tended to rate of root biomass accumulation, root spread and developed in preferential directions. Thus, the number depth within the first 3 years following planting (Marden of roots per unit volume of soil was lower than for the et al. 2005). To what extent minor root binding affected tap-rooted species with significant areas of soil the eventual distribution of root diameter size classes Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 19 of 26 Table 4 Mean tree biomass (dry weight), carbon content (kg) and carbon stock (t CO /1000 stems per hectare) for 5-year-old trees Botanical name Common name Mean tree biomass (kg) Carbon content (kg) Carbon stock (t CO /1000 spha) Coniferous species a a a Agathis australis Kauri 0.32 (0.03) 0.16 (0.02) 0.53 (0.05) bc bc bc Dacrycarpus dacrydioides Kahikatea 1.60 (0.24) 0.80 (0.12) 2.64 (0.40) c c c Dacrydium cupressinum Rimu 1.56 (0.13) 0.78 (0.06) 2.58 (0.21) d d d Podocarpus totara Totara 2.59 (0.32) 1.30 (0.16) 4.28 (0.53) a a a Prumnopitys ferruginea Miro 0.52 (0.09) 0.26 (0.04) 0.85 (0.15) ab ab ab Prumnopitys taxifolia Matai 0.72 (0.07) 0.36 (0.03) 1.18 (0.12) Broadleaved species c c c Alectryon excelsus Titoki 1.49 (0.23) 0.75 (0.12) 2.47 (0.39) e e e Vitex lucens Puriri 4.28 (0.92) 2.14 (0.46) 7.07 (1.52) Values in parentheses represent standard error of the mean. Values with different letters, within each parameter, were significantly different (Student-Newman-Keuls P =0.05). The carbon stock values do not include woody debris or fine litter and spatial pattern of root development of the species re- The high concentration of roots (mass and length) of ported in this paper remains uncertain. Nonetheless, it both the tap- and plate-rooted species at shallow depth was clear that the timing and proportion of total plant bio- was also anticipated. General observations of the root mass allocated to roots and root bole did not increase sig- systems of mature plate-rooted indigenous species nificantly until years 4 and 5, and this appears to be confirm that their root systems are largely confined to related to the period (years after planting) that plants re- the uppermost metre or two of the soil with the preva- quire to establish their root systems sufficiently to sustain lence of shallow rooting in forest soils likely reflecting an increase in growth rate. With the exception of the dependence on nutrient cycling through litter fall broadleaved and plate-rooted Alectryon excelsus,the (Yeates 1924; Cranwell and Moore 1936; Wardle 2002). tap-rooted and predominantly coniferous species allocated Furthermore, observations of the root systems of the more of their total BGB to the root bole than did the mature, plate-rooted Dacrycarpus dacrydioides indicate remaining plate-rooted species or the tap-rooted Agathis that their lateral structural roots are shallow and often australis. That said, there is also likely to be a relationship sub-aerial as an adaptation to imperfectly drained soils between plant size and the rate of root and root bole bio- and/or wetter flatland sites including alluvial floodplains mass accumulation with the smallest and slowest growing and terraces with a periodic high water table dominated of the conifers Agathis australis, Prumnopitys taxifolia by pasture grasses (Wardle 2002). Alternatively, as sug- and Prumnopitys ferruginea amassing the least BGB. gested by Nicoll and Ray (1996), a shallow and extensive Between years 1 and 3, there was no consistent lateral root network may also be an adaptation to im- relationship between the diameter of the canopy and prove the rigidity of the soil-root plate and counteract root system dimensions (diameter) of tap- versus the increasing vulnerability to windthrow as trees grow plate-rooted species or between conifer and broadleaved taller. Similarly, the development of an extensive species. However, by year 4 and thereafter, the diameter network of lateral structural roots in Populus and Salix of the root systems of the broadleaved Alectryon excelsus species makes them the preferred species for soil conser- and Vitex lucens and the conifer Dacrycarpus dacry- vation and slope reinforcement on poorly drained sites dioides exceeded the width of their respective canopies such as earthflows (Phillips et al. 2014). suggesting that the ratio of root spread to canopy dri- The maximum rooting depth recorded for an indigen- pline may be different for different species at the same ous forest species is for a mature tap-rooted Agathis age. From the few available descriptions of root system australis. Unless impeded by rock or a hard soil pan, tap diameters of mature trees, it is evident that the lateral roots have been recorded at depths in excess of 2 m roots of conifers Dacrydium cupressinum and Podocar- (Bergin and Steward 2004), and ‘sinker roots’ have been pus totara may extend > 9 m from the root bole (Bergin observed to descend from large, lateral roots to a depth and Gea 2005, 2007). Similarly, the lateral roots of ma- of 4 m and terminate in a network of smaller roots (R. ture Prumnopitys taxifolia are often seen exposed on the LLoyd, pers. comm. cited in; Bergin and Steward 2004). ground surface and extending to > 19.5 m from the root Agathis australis is predominately found growing in the bole before disappearing into the soil (Allan 1926), and more sub-tropical regions of New Zealand so the early thus, for the more conical-shaped conifers, their lateral development of a deep tap root and relatively large bio- roots likely extend beyond the canopy dripline. mass allocation to roots should be beneficial to surviving Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 20 of 26 droughts as suggested by Becker and Castillo (1990). It species, and for areas actively planted and managed for has also been suggested that a larger (and early) alloca- mānuka honey production, 1200–1300 spha would be tion to root biomass is beneficial for tree establishment required. (T Poulson, pers. comm. cited in Coutts 1987). For the best performed of the trialled species, a mono- culture of broadleaved Vitex lucens planted at 1000 spha Biomass and carbon accumulation could potentially achieve a level of forest carbon stock of −1 Nationally, there is 1.45 Mha of marginal pastoral land 7.07 t CO ha by year 5 (Table 4), which approximated −1 suitable for reforestation by indigenous trees and shrubs to the 7.8 t CO ha estimated by the Ministry for (Trotter et al. 2005). The industry standard for establish- Primary Industries (2017) for new areas of planted and ment plantings of non-native species (principally Pinus reverting shrubland. However, irrespective of the species radiata) on marginal land varies considerably depending mix and planting regime adopted for new areas of on the class of land and its erosion severity (National indigenous forest establishment, the potentially Water and Soil Conservation Organisation 1976). In the achievable forest carbon stock by year 5 is likely to be an −1 highly erodible East Coast region of the North Island, order of magnitude less than the ~ 91.6 t CO ha of the planting density is generally around 1000–1250 forest carbon stock recorded for 4-year-old Pinus −1 stems per hectare (spha). The establishment of a mixed radiata (from Marden et al. 2016) and the 77 t CO ha indigenous forest comprising 125 of each of the eight estimated by the Ministry for Primary Industries (2017) species trialled here (1000 spha) would potentially amass for 5-year-old post-1989 plantings in the East Coast −1 ~ 1.5 t ha within 5 years of planting, which would region. −1 generate a forest carbon stock of 2.5 t CO ha . With Notwithstanding, the planting of indigenous conifers the exception of Agathis australis, the remainder of the and broadleaved species on erodible marginal land will conifers (mainly tap-rooted species) individually ultimately afford a more continuous and permanent for- accumulated significantly more total plant biomass than est sink for carbon over an active growth phase that ex- did any of the plate-rooted species, while the broad- tends from 150 to 500 years (Hall 2001; Hinds and Reid leaved Vitex lucens (also tap-rooted) accumulated the 1957) that is consistent with the prolonged effort likely largest total biomass of any of the trialled species required to effect significant reductions in atmospheric (Table 4, Additional file 3). Selection of a species mix CO levels. comprising 200 of each of the best performing of the tap-rooted conifers (Dacrycarpus dacrydioides, Dacry- Towards implementation dium cupressinum, Podocarpus totara) and the broad- The second objective of this trial was to assess the po- leaved Vitex lucens, together with the best performing of tential effectiveness of the trialled species for erosion the plate-rooted broadleaved Alectryon excelsa would control on newly afforested/reforested marginal pastoral −1 amass ~ 2.3 t ha , a forest carbon stock of land and/or as replacement species best suited to pro- −1 3.8 t CO ha . Furthermore, an early colonising viding long-term stability to areas of exotic forest indigenous shrubland species mānuka (Leptospermum deemed at greatest risk to storm-related erosion and scoparium (Forster et Forster f.)) established at the same under consideration for retirement in the future. −1 density would amass ~ 3.7 t ha resulting in a forest The well-developed literature on the geomechanical −1 carbon stock of 6.1 t CO ha over the same 5-year effects of vegetation on hill slope stability clearly indicate period (Marden and Lambie 2016). Thus, the planting of that tree roots reinforce soils, contribute to soil strength Leptospermum scoparium on marginal land would store and stabilise eroding hillslopes. The mechanical interac- −1 an additional ~ 2.3 t CO ha over an equivalent period tions of woody roots with the soil medium play an of time than that gained from planting an equivalent important role in tree anchorage and thus in the preven- density of the best performed of the trialled broadleaved tion of slope failure. Thus, one of the most important and conifer species. Interestingly, the carbon stock for traits is rooting depth, which, when soil conditions are areas of indigenous forest 5 years after establishment not limiting, is species dependent. The most effective (planting), or from the date when land-use change to re- slope stabilisation results when root development occurs −1 version commences, is estimated at 7.8 t CO ha at different depths in the soil profile (Schiechtl and Stern (Ministry for Primary industries 2017). By implication, 1994; Czernin and Phillips 2005). This has been ob- to achieve a similar level of carbon stock within this served in exposed root boles of mature Dacrydium time frame would for areas of new plantings require ~ cupressinum and Agathis australis. 2000 spha of the best performed of the trialled In New Zealand, shallow landslides on typically steep broadleaved and conifer species. Similarly, for areas set slopes underlain by Tertiary-aged bedrock are generally aside to permit passive reversion to occur would require ≤ 1 m deep (Marden et al. 1991; Page et al. 1994)so an equivalent density of mixed broadleaved and conifer roots must cross the basal shear plane in order to anchor Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 21 of 26 the soil into more stable substrate and stabilise a slope Larger diameter roots at greater soil depths would be against the initiation of shallow landslides (Wu et al. desirable to stabilise the deeper forms of mass move- 1979). This study shows that, for most of the tree species ment; thus, root diameter is another important consider- examined, the root systems are concentrated predomin- ation on landslide-prone slopes as thick roots reinforce antly in the upper soil profile with mean maximum root soil in the same way that concrete is reinforced with depths by year 5 ranging between 0.3 and 0.4 m. Further- steel rods, and the thicker the root, the greater its ability more, at this stage of their development, there was no sig- to penetrate soil (Clark et al. 2008). The most common nificant difference in this metric among individual conifer barriers limiting the development of a deep rooting species, between conifer and broadleaved species or system include a permanent high water table, bedrock between plate-rooted and tap-rooted species. The results interface or a cemented iron pan and stoniness and also indicate that there is a rapid decline in roots with depth of colluvium. In general terms, the root:shoot increasing depth and distance from the stem, that there ratios of seedlings are depressed on such sites. As previ- are interspecies differences in root distribution and that ously found for other New Zealand woody species, root each species allocates differing proportions of their total depth was not necessarily correlated with tree age but biomass to roots at different stages of growth, a finding rather to the stoniness and depth of slope colluvium similar to other local (Watson et al. 1995, 1999) and inter- (Watson et al. 1995). national (Abernethy and Rutherfurd 2001;Eassonand In terms of root anchorage, tap-rooted species will, in Yarbrough 2002) studies on plant root distribution. This, time, likely provide a higher level of reinforcement together with the few published reports on the root depth directly under a tree and to a greater depth than for of some of New Zealand’s tallest native conifer and hard- plate-rooted species. Thus, for tap-rooted species, root wood forest species (Cameron 1963), with the exception reinforcement quickly tapers off laterally as root density of Agathis australis, indicate that the rooting depth of ma- declines with increasing distance from the tree. In con- ture trees rarely exceeded 2–3 m. By implication, their trast, plate-rooted species with their widespread devel- comparatively shallow rooting depth will likely be a limit- opment of a lateral root network will likely provide a ing factor in their ability to contribute to deep-soil high level of near-surface soil-root reinforcement but reinforcement. Similarly, the ability of widely spaced nonetheless remain more susceptible to uprooting than plantings to prevent the occurrence of shallow landslide will tap-rooted species because as the number of roots diminishes as root distribution decreases with increasing decline with increased depth (Gray and Sotir 1996)so distance from the root bole. does root anchorage rapidly decline (Dupuy et al. 2005; Interspecies differences in root distribution (depth and Norris et al. 2008). lateral spread) have implications for the planting As a consequence of their shallow-rooted habit, many densities required to provide full, near-surface root of New Zealand’s conifer and broadleaved trees will have occupancy of the soil and/or to maximise the root dens- limited effectiveness in floodplain reaches where channel ity to the depth requirement at specific sites (Phillips hydraulic conditions are likely to result in the undercut- et al. 2012). For example, a dense lateral network of ting of stream banks to a depth greater than that of the woody roots predominantly confined to the upper soil maximum penetration depth of root systems (Marden horizons often forms a membrane that stabilises shallow et al. 2005). If the potential for streambed degradation soils (Schmidt et al. 2001) and significantly reduces exists, additional protection in the form of structural shallow landslide potential on steep slopes (Ekanayake materials (e.g. gabion baskets and rip rap) will be re- et al. 1997), while the larger tree roots can provide quired along the toe of banks and to some depth below reinforcement across lateral planes of weakness bound- normal streambed level. The limitations of root depth ing potential slope failures (Sidle et al. 1985). However, aside, some of the trialled conifer and broadleaved deeper soil mantles (> 5 m) benefit little from such species have adaptations suitable for floodplain restor- reinforcement as root density decreases dramatically ation. Alluvial communities of some of New Zealand’s with depth and few large roots are able to anchor across oldest podocarps (Podocarpus totara, Dacrycarpus dacry- the basal shear plane (Stokes et al. 2008). The only bene- dioides) are tolerant of wet soils and swampy sites subjected fit of root strength to the stabilisation of deep-seated to periodic flooding and seasonally high water tables. They landslides, e.g. earthflows and slumps, is when very large have the ability to produce a new root system after flooding lateral woody roots cross planes of weakness along the and inundation with river silt, and Podocarpus totara is able flanks of potential failures (Sidle and Ochiai 2006). to produce tiers of adventitious roots from the buried part However, such reinforcement would only benefit of the stem should the tree topple or break. Dacrycarpus deep-seated landslides of small area extent whereas dacrydioides, however, is intolerant of siltation if greater lateral roots also offer protection against most shallow than ~ 60 cm (Foweraker 1929; McSweeney 1982; landslides (Swanston and Swanson 1976). Campbell 1984). Many species, including Dacrydium Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 22 of 26 cupressinum, Podocarpus totara and Agathis australis,are species (e.g. Agathis australis) have the potential to equally tolerant of harsh growing conditions on steep develop thicker and stronger roots able to penetrate slopes with skeletal soils deficient in nutrients that are across the basal shear plane to provide a deeper level of prone to drying during seasonal droughts. reinforcement, albeit only for the smaller deep-seated Although native conifers and broadleaved trees have failures, than would plate-rooted species. Nonetheless, not traditionally been used specifically for soil conserva- as was evident before the mature indigenous forest cover tion, their successional role meets most of the require- was removed from hill-country areas, given time to ma- ments for use in revegetation where long-term erosion ture, these same species will slow down but may not control is required (Pollock 1986). It is also important to completely arrest many of the deeper-seated forms of consider that, in addition to its erosion-control function, mass failure. a species mix may have greater associated ecological, For many restoration sites, the strategy adopted should cultural, aesthetic and economic benefits than a single ideally include a mix of early colonising and successional species stand. In addition, most are sufficiently diverse species with plate- and tap-rooted characteristics. On in rooting form to stabilise slopes prone to shallow land- less exposed sites, a mixture of shrub and tree species slides and dry ravelling, while those with the limitations can be planted concurrently with the longer-lived conifer of root depth will be less effective in stabilising the dee- and broadleaved species interplanted at near final spa- per forms of mass movements such as earthflows and cing (Bergin and Gea 2005, 2007). Another option that slumps as their failure plane tends to be deeper than the has found favour with many community groups is to re- vertical limit of root penetration. Plant density and spe- strict the number of colonising species (three or four) cies mix are also key factors in determining plant effect- and then carry out follow-up planting with the late suc- iveness; thus, options that promote the quickest canopy cession, longer-lived species after a few years to fill in closure and root development at all levels of the soil any gaps created by failures from the initial planting profile are likely to be the most effective in promoting (Fred Lichtwark pers. comm.). slope stability (Phillips et al. 2001). The faster the soil is During selection of suitable plant materials, the cap- occupied by roots, the greater the reinforcement of the acity of species to match specific site conditions, e.g. soil and the greater the chance of limiting landslide initi- those that are long lived and can overcome potential ation on erosion-prone slopes and dry ravelling of ex- limitations to successful establishment such as over- posed soil and rock. As an example, soil occupancy by crowding resulting in suppressed growth (less is often the larger root systems of the broadleaved Vitex lucens best in the long term), frost, fungal and insect attack, and Alectryon excelsus would occur sooner (~ 5 years browsing, wind throw and stem breakage, must also be after planting) than the best performed of the conifers taken into account. Potential limitations to a successful (Podocarpus totara) tested under optimal growth condi- establishment outcome of the trialled species, particu- tions and using the mean maximum root spread at year larly during the early growth period, include susceptibil- 5 as an indicator of the rate at which soil-root ity to frost (e.g. Alectryon excelsus, Vitex lucens, reinforcement might be considered as being effective Dacrydium cupressinum, Agathis australis), insect dam- against erosion. Conversely, the more compact root sys- age (e.g. Podocarpus totara, Dacrydium cupressinum, tems of Prumnopitys taxifolia, Prumnopitys ferruginea Dacrycarpus dacrydioides, Agathis australis), fungi (e.g. and Dacrydium cupressinum would take longer (~ 10– Agathis australis, Podocarpus totara), browsing feral 15 years) to be effective and the significantly smaller mammals (e.g. Dacrydium cupressinum, Podocarpus to- root system of Agathis australis, the longest. Phillips tara, Agathis australis) and stem snap (e.g. Alectryon et al. (2015) compared the lateral root spread of exotic excelsus) (Bergin and Gea 2005, 2007). species including alder (Alnus rubra), cherry (Prunus serrulatus), blackwood (Acacia melanoxylon) and red- Conclusions wood (Sequoia sempervirons) with reported values for This study has provided much-needed new information indigenous riparian species (Marden et al. 2005), for mā- on the overall growth performance of eight of New nuka (Marden and Lambie 2016) and for Pinus radiata Zealand’s more common conifer and broadleaved forest at a similar age (Watson and Tombleson 2002, 2004; species. However, the conclusions drawn are primarily Marden et al. 2016). The conclusion drawn was that the premised on their early growth period (first 5 years) at species with the larger root system dimensions would one trial site; thus, it must be acknowledged that provide earlier soil-root reinforcement and thus be more allometric relationships will likely change with increasing effective in mitigating the initiation of shallow landslides. tree age. The excavation of whole-plants has provided Similarly, using root depth as an indicator of species po- rare and valuable insights into inter-species differences tential for stabilising deeper-seated forms of erosion (e.g. in root architecture (dimensions), root sizes and biomass rotational slumps and larger earthflows), tap-rooted and their distribution relative to soil depth and distance Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 23 of 26 from the stem. Such data are integral to the selection of lands collected in 2002/03. Landcare Research Contract an appropriate species mix, planting density and as a Report LC0506/099. means of assessing the period (years after planting) Payton, I., Forrester, G., Lambie, S., Berben, P., Pin- juvenile plantings are likely to remain vulnerable to the kney, T. (2009). Development and validation of allomet- initiation of shallow landslides. Although the juvenile ric equations for carbon inventory of indigenous forests plants of the trialled species typically develop very shallow and shrublands. Landcare Research Contract Report root systems, observations and anecdotal descriptions of LC0910/004. Prepared for the Ministry of Agriculture the root systems of older trees show they have potential and Forestry. for mitigating shallow landslides, earthflows of small ex- tent and dry ravelling within a few years of planting. Con- Additional files versely, as the basal failure plane of the deeper forms of Additional file 1: Ecological and site requirements of the trialled mass movements such as earthflows and slumps tends to species. (DOCX 16 kb) be deeper than the vertical limit of root penetration of the Additional file 2: Mean above-ground attributes of six indigenous coni- majority of the trialled species, their stabilisation will fer and two broadleaved species. (DOCX 17 kb) occur progressively over several decades. Additional file 3: Mean tree component biomass, above-ground (AGB) The previously unavailable species-specific information and below-ground (BGB) biomass totals and BGB/AGB ratio. (DOCX 23 kb) on rates of whole-plant biomass accumulation for the Additional file 4: Mean below-ground attributes of six indigenous indigenous forest species presented in this paper will conifer and two broadleaved species. (DOCX 17 kb) improve current estimates of forest carbon stocks −1 Additional file 5: Mean root and root bole biomass (g), as a percent of (t CO ha ) during the juvenile stage of growth of total below-ground biomass (BGB). (DOCX 19 kb) mixed-species or single-species woodlots proposed for Additional file 6: Mean root biomass (g) and percent of total root new plantings on land considered marginal for pastoral biomass, by 0.5-m radial increments from the root bole. (DOCX 19 kb) use and unsustainable for short-rotation exotic forestry. Additional file 7: Mean and percent of total root length (m) of roots > 1 mm diameter, by 0.5-m radial increments from the root bole. Species-specific growth performance data for both (DOCX 23 kb) old-growth and juvenile trees of many of New Zealand’s indigenous species remain elusive. It is therefore import- Abbreviations ant that further time-series baseline data on plant AGB: Above-ground biomass; AGS: Afforestation Grant Scheme; BGB: Below- growth rates are collected for a greater variety of New ground biomass; BP: Before present; C: Carbon; CO : Carbon dioxide; DBH: Diameter at breast height; ETS: Emissions Trading Scheme; Zealand’s indigenous species across a diversity of site FCCC: Framework Convention for Climate Change; GPR: Ground-penetrating conditions known to influence plant growth and that radar; IPCC: Intergovernmental Panel on Climate Change; PFSI: Permanent data on root metrics, in particular, are collected in a for- Forest Sink Initiative; RCD: Root collar diameter; spha: Stems per hectare mat suitable for inclusion in slope stability and/or soil Acknowledgements reinforcement models (e.g. Schwarz et al. 2010), many of We acknowledge the support of the Tairawhiti Polytechnic Rural Studies which suffer from a paucity of the types of data collected Unit, Gisborne, on whose land this plant trial was located. We thank interns in this study (Phillips et al. 2012). Claire Butty (France), Sandra Viel (Germany) and Kaisa Valkonen (Finland) and Landcare Research colleagues Alex Watson and Richard Hemming for assistance with field extraction and the partitioning and measuring of plant components. Graphics were drawn by Nic Faville and Suzanne Lambie. Anne Endnotes Austin edited the script, and Anne Sutherland provided GIS support. Robyn Simcock reviewed and provided valuable comment. Thanks also to the Dale, J. M. (2013). Evaluation of methods for quantify- anonymous external reviewers for their valuable comments This research ing carbon storage of urban trees in New Zealand. Paper was supported by funding from the New Zealand Ministry of Business, as partial fulfilment of Bachelor of Engineering (Honours) Innovation and Employment (MBIE), Specific Science Investment Fund (SSIF), ‘Growing Confidence in Forestry’s Future’ research programme (contract thesis, University of Auckland, Auckland, New Zealand. CO4X1306) and ‘Clean Water Productive Land’ research programme (contract Beets, P. N., Oliver, G. R., Kimberley, M. O., Pearce, S. H. C10X1006). (2008). Allometric functions for estimating above-ground Availability of data and materials carbon in native forest trees, shrubs and ferns. Scion Report Please contact the author for data requests. 12679, prepared for the Ministry for the Environment. Allen, R. B., Wiser, S. K., Hall, G., Moore, T., Goulding, Authors’ contributions MM was the primary author. MM, SL and CJP undertook the fieldwork. SL C., Beets, P., Nordmeyer, A. (1998). A national system for compiled the data into spreadsheets and completed the statistical analyses. monitoring change in carbon storage in indigenous forests All authors read and approved the manuscript. and scrub: testing plot locations on a South Island tran- Competing interests sect. Contract Report JNT9798/166 for the Ministry for The authors declare that they have no competing interests. the Environment, Wellington, New Zealand. Peltzer, D. A., Payton, I. J. (2006). Analysis of carbon Ethics approval and consent to participate monitoring system data for indigenous forests and shrub Not applicable. Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 24 of 26 Publisher’sNote Coutts, MP. (1987). Developmental processes in tree root systems. Canadian Springer Nature remains neutral with regard to jurisdictional claims in published Journal of Forest Research, 17(8), 761–767. maps and institutional affiliations. Cranwell, LM, & Moore, LB. (1936). The occurrence of kauri in montane forest on Te Moehau. New Zealand Journal of Science and Technology, 18, 531–543. Author details Czernin, A, & Phillips, CJ. (2005). Below-ground morphology of Cordyline australis Landcare Research, NZ, Ltd, PO Box 445, Gisborne 4010, New Zealand. (New Zealand cabbage tree) and its suitability for riverbank stabilisation. New Landcare Research, NZ, Ltd, Private Bag 3127, Hamilton 3240, New Zealand. Zealand Journal of Botany, 43, 851–864. Landcare Research, NZ, Ltd, PO Box 69041, Lincoln 7640, New Zealand. Dupuy, L, Fourcaud, T, Stokes, A. (2005). A numerical investigation into factors affecting the anchorage of roots in tension. European Journal of Soil Science, Received: 15 November 2017 Accepted: 8 May 2018 56, 319–327. Easson, G, & Yarbrough, LD. (2002). The effects of riparian vegetation on bank stability. Environmental & Engineering Geoscience, 8, 247–260. Ekanayake, J, Marden, M, Watson, AJ, Rowan, D. (1997). Tree roots and slope stability: a comparison between Pinus radiata (Radiata pine) and Kunzea References ericoides (Kanuka). New Zealand Journal of Forestry Science, 27, 216–233. Abernethy, B, & Rutherfurd, ID. (2001). The distribution and strength of riparian Foweraker, C. E. (1929). The podocarp rain forests of Westland, New Zealand. tree roots in relation to riverbank reinforcement. Hydrological Processes, 15, Kahikatea and Totara forests and their relationships to silting. Te Kura 63–79. Ngahere. A forestry journal issued by the Forestry Club of the Canterbury Allan, HH. (1926). The surface roots of an individual matai. New Zealand Journal of College School, Christchurch, New Zealand. Science and Technology, 8, 233–234. Gray, DH, & Sotir, RB (1996). Biotechnical and soil bioengineering slope stabilisation. Allan Herbarium (2000). New Zealand plant names database. New Zealand: New York: Wiley. Landcare Research Available http://nzflora.landcareresearch.co.nz. Accessed Hall, G, Wiser, S, Allen, R, Beets, P, Goulding, C. (2001). Strategies to estimate o 15 Jan 2005. national forest carbon stocks from inventory data: the 1990 New Zealand Becker, P, & Castillo, A. (1990). Root architecture of shrubs and saplings in the baseline. Global Change Biology, 7, 389–403. understory of a tropical moist forest in lowland Panama. Biotropica, 22, Hall, GMJ. (2001). Mitigating an organisation’s future net carbon emissions by 242–249. native forest restoration. Ecological Applications, 11, 1622–1633. Beets, PN. (1980). Amount and distribution of dry matter in a mature beech/ Hasselmann, K. (1997). Climate change after Kyoto. Nature, 390, 225–226. podocarp community. New Zealand Journal of Forestry Science, 10, 395–418. Beets, PN, Kimberley, MO, Oliver, GR, Pearce, BH, Graham, JD, Brandon, A. (2012). Heth, D, & Donald, DGM. (1978). Root biomass of Pinus radiata D. Don. South Allometric equations for estimating carbon stocks in natural forest in New Africa Forestry Journal, 107,60–70. Zealand. Forests, 3, 818–839. Hewitt, AE (2010). New Zealand soil classification, Landcare Research Science Series Beets, PN, Kimberley, MO, Paul, TSH, Oliver, GR, Pearce, SH, Buswell, JM. (2014). No. 1. Lincoln: Manaaki Whenua Press. The inventory of carbon stocks in New Zealand’s post-1989 natural forest for Hinds, HV, & Reid, JS (1957). Forest trees and timbers of New Zealand, New Zealand reporting under the Kyoto Protocol. Forests, 5(9), 2230–2252. Forest Service Bulletin 12. Rotorua: New Zealand Forest Service. Beets, PN, Pearce, SH, Oliver, GR, Clinton, PW. (2007). Root/shoot ratios for Hruska, J., Cermak, J., Sustek, S. (1999). Mapping tree root systems with ground- deriving below-ground biomass of Pinus radiata stands. New Zealand Journal penetrating radar. Tree Physiology, 19, 125–130. (2000). of Forestry Science, 37(2), 267–288. IPCC (2000). In RT Watson, IR Noble, B Bolin, NH Ravindranth, DJ Verardo, DJ Bergin, D (2003). Totara: establishment, growth and management, New Zealand Dokken (Eds.), Land use, land-use change, and forestry. A special report of the Indigenous Tree Bulletin No. 1. Rotorua: New Zealand Forest Research Institute. IPCC. Cambridge: Cambridge University Press. Bergin, D, & Gea, L (2005). Native trees: planting and early management for Jara, MC, Henry, M, Réjou-Méchain, M, Wayson, C, Zapata-Cuartas, M, Piotto, D, production wood, New Zealand Indigenous Tree Bulletin No. 3. Rotorua: New Guier, FA, Lombis, HC, López, EC, Lara, RC, Rojas, KC, Pasquel, JD, Montoya, Zealand Forest Research Institute. AD, Vega, JF, Galo, AJ, López, OR, Marklund, LG, Fuentes, JMM, Milla, F, Bergin, D, & Gea, L (2007). Native trees: planting and early management for wood Chaidez, JJN, Malavassi, EO, Pérez, J, Zea, CR, Garcia, LR, Pons, RR, Saint-André, production, New Zealand Indigenous Tree Bulletin No. 3 (revised edition). L, Sanquetta, C, Scott, C, Westfall, J. (2014). Guidelines for documenting and Rotorua: New Zealand Forest Research Institute. reporting tree allometric equations. Annals of Forest Science, 72, 763–768. Bergin, D, & Steward, G (2004). Kauri: ecology, establishment, growth and Ketterings, QM, Coe, R, van Noordwijk, M, Ambagau, Y, Palm, CA. (2001). management, New Zealand Indigenous Tree Bulletin No, 2. Rotorua: New Reducing uncertainty in the use of allometric biomass equations for Zealand Forest Research Institute. predicting above-ground tree biomass in the mixed secondary forests. Forest Cairns, MA, Brown, S, Helmer, EH, Baumgardner, GA. (1997). Root biomass Ecology and Management, 146, 199–209. allocation in the worlds upland forests. Oecologia, 111,1–11. Kira, T, & Shidei, T. (1967). Primary production and turnover of organic matter in different forest ecosystems of the western Pacific. Japenese Journal of Cameron, RJ. (1963). A study of the rooting habits of rimu and tawa in pumice Ecology, 17,70–87. soils. New Zealand Journal of Forestry, 8, 771–785. Campbell, DJ. (1984). The vascular flora of the DSIR study area, lower Korner, C. (1998). A re-assessment of high elevation tree line positions and their Orongorongo Valley, Wellington, New Zealand. New Zealand Journal of explanation. Oecologia, 115, 445–459. Botany, 22, 223–270. Magalhães, TM. (2015). Allometric equations for estimating belowground biomass Cao, KF, & Ohkubo, T. (1998). Allometry, root/shoot ratio and root architecture in of Androstachys johnsonii Prain. Carbon Balance and Management, 10, 16. understory saplings of deciduous dicotyledonous trees in central Japan. Marden, M., Lambie, S. (2016). Plot based, growth performance of space-planted Ecological Research, 13(2), 217–227. mānuka (Leptospermum scoparium) on marginal land, and vulnerability to erosion: Carswell, FE, Burrows, LE, Hall, GMJ, Mason, NWH, Allen, RB. (2012). Carbon and final report. Ministry for Primary Industries Technical Paper No 2016/20. http:// plant diversity gain during 200 years of woody succession in lowland New www.mpi.govt.nz/news-resources/publications.aspx Accessed 30 Aug 2016. Zealand. New Zealand Journal of Ecology, 36(2), 191–202. Marden, M, Lambie, S, Rowan, D. (2018). Root system attributes of 12 juvenile Chiatante, D, Scippa, SG, Iorio, AD, Sarnataro, M. (2003). The influence of steep slopes indigenous early-colonising shrub and tree species with potential for on root system development. Journal of Plant Growth Regulation, 21,247–260. mitigating erosion in New Zealand. New Zealand Journal of Forestry Science, 48 (in press). Claesson, S, Sahlén, K, Lundmark, T. (2001). Functions for biomass estimation of young Pinus sylvestris, Picea abies and Betula spp. from stands in Northern Sweden with Marden, M, Phillips, CJ, Rowan, D (1991). Declining soil loss with increasing age high stand densities. Scandinavian Journal of Forestry Research, 16,136–146. of plantation forest in the Uawa catchment, East Coast region, North Island, Clark, LJ, Price, AH, Steel, KA, Whalley, WR. (2008). Evidence from near-isogenic New Zealand. In Proceedings of the International Conference on Sustainable lines that root penetration increases with root diameter and bending Land Management, (pp. 358–361). Napier: Simon Printing Co. Ltd. stiffness in rice. Functional Plant Biology, 35, 1163–1171. Marden, M, Rowan, D, Lambie, S. (2016). Root development and whole-tree Coomes, DA, Allen, RB, Scott, NA, Goulding, C, Beets, P. (2002). Designing systems allometry of juvenile trees of five seed lots of Pinus radiata D.Don: implications to monitor carbon stocks in forests and shrublands. Forest Ecology and for forest establishment on erosion-prone terrain, East Coast region, North Management, 164,89–108. Island, New Zealand. New Zealand Journal of Forestry Science, 46:24. Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 25 of 26 Marden, M, Rowan, D, Phillips, C. (2005). Stabilising characteristics of New Zealand Schwendenmann, L, & Mitchell, ND. (2014). Carbon accumulation by native trees and indigenous riparian colonising plants. Plant and Soil, 278,95–105. https://doi. soil in an urban park, Auckland. New Zealand Journal of Ecology, 38(2), 213–220. org/10.1007/s11104-004-7598-2. Scott, NA, White, JD, Townsend, JA, Whitehead, D, Leathwick, JR, Hall, GMJ, Mason, NWH, Beets, PN, Payton, I, Burrows, L, Holdaway, RJ, Carswell, FE. (2014). Marden, M, Rogers, GND, Watson, AJ, Whaley, PT. (2000). Carbon and Individual-based allometric equations accurately measure carbon storage nitrogen distribution and accumulation in a New Zealand scrubland and sequestration in shrubland. Forests, 5, 309–324. ecosystem. Canadian Journal of Forest Research-Revue Canadienne De Masters, S. E., Holloway, J. T., Mc Kelvey, P. J. (1957). The National Forest Survey of Recherche Forestiere, 30(8), 1246–1255. New Zealand, 1955. Vol. 1. Wellington: Government Printer. Sidle, RC, & Ochiai, H (2006). Landslides: processes, prediction, and land use, American Geophysical Union, water resources monogram 18. Washington, DC: McGlone, MS (1988). Glacial and Holocene vegetation history; 20-ky to present: New Zealand. In B Huntly, T Webb III (Eds.), Vegetation history. Handbook of AGU. vegetation science, (pp. 557–559). Dordrecht: Kluwer. Sidle, RC, Pearce, AJ, O’Loughlin, CL (1985). Hillslope stability and land use, McSweeney, GD. (1982). Matai/totara flood-plain forests in South Westland. New American Geophysical Union, water resource monogram, 11 (p. 140). Zealand Journal of Ecology, 5, 121–128. Washington, DC: American Geophysical Union. Ministry for Primary Industries (2015). A guide to the afforestation grant Steward, GA, & Beveridge, AE. (2010). A review of New Zealand kauri (Agathis scheme. Wellington: Ministry for Primary Industries http://www.mpi.govt. australis (D. Don) Lindl.): its ecology, history, growth and potential for nz/funding-and-programmes/forestry/afforestation-grant-scheme/ management for timber. New Zealand Journal of Forestry Science, 40,34–59. Accessed 21 May 2017. Stokes, A, & Guitard, D (1997). Tree root response to mechanical stress. In A Ministry for Primary Industries (2017). A guide to carbon look-up tables for forestry Altman, Y Waisel (Eds.), Biology of root formation and development, (pp. 227– in the emissions trading scheme. Wellington: Ministry for Primary Industries 236). New York: Plenum Press. http://www.mpi.govt.nz/growing-and-producing/forestry/forestry-in-the- Stokes, A, Norris, JE, van Beek, LPH, Bogaard, T, Cammeraat, E, Mickovski, SB, emmissions-trading-scheme/emissions-returns/ Accessed 11 Sept 2017. Jenner, A, Di Iorio, A, Fourcaud, T (2008). How vegetation reinforces soil on National Water and Soil Conservation Organisation [NWASCO] (1976). New slopes. In JE Norris, A Stokes, SB Mickovski, E Cammeraat, R van Beek, BC Zealand land resource inventory worksheets. Wellington: Government Printer. Nicoll, A Achim (Eds.), Slope stability and erosion control: ecotechnological New Zealand Climate Change Office (2003). Climate change. National Inventory solutions, (pp. 65–118). Dordrecht: Springer. Report for New Zealand. Greenhouse gas inventory 1990–2001. Wellington: Sun, HL, Li, SC, Xiong, WL, Yang, ZR, Cui, BS, Yang, T. (2008). Influence of slopes on Ministry for the Environment. root system anchorage of Pinus yunnanensis. Ecological Engineering, 32,60–67. Nicoll, BC, & Ray, D. (1996). Adaptive growth of tree root systems in response to Swanston, DN, & Swanson, FJ (1976). Timber harvesting, mass erosion, and wind action and site conditions. Tree Physiology, 16, 891–898. steepland forest geomorphology in the Pacific North West. In DR Coates (Ed.), Norris, JE, Stokes, A, Mickovski, SB, Cammeraat, E, van Beek, LPH, Nicol, B, Achim, Geomorphology and engineering,(pp. 199–221). Stroudsburg: Dowden, A (2008). Slope stability and erosion control: ecotechnological solutions. Hutchinson, and Ross. Dordrecht: Springer. Tate, KR, Scott, NA, Parshotam, A, Brown, L, Wilde, RH, Giltrap, DJ, Trustrum, NA, Gomez, B, Ross, DJ. (2000). A multi-scale analysis of a terrestrial carbon Page, MJ, Trustrum, NA, De Rose, RC. (1994). A high-resolution record of storm- budget: is New Zealand a source or sink of carbon? Agriculture Ecology and induced erosion from lake sediments, New Zealand. Journal of Environment, 82, 229–246. Paleolimnology, 11, 333–348. Parliamentary Commissioner for the Environment (2002). Weaving resilience into Tate, KR, Scott, NA, Saggar, S, Giltrap, DJ, Baisden, WT, Newsome, PF, Trotter, CM, our working lands: recommendations for the future roles of native plants. Wilde, RH. (2003). Land use change alters New Zealand’s terrestrial carbon Wellington: Parliamentary Commissioner for the Environment. budget: uncertainties associated with estimates of soil carbon change Phillips, CJ, Marden, M, Lambie, S. (2014). Observations of root growth of between 1990–2000. Tellus, 55B, 365–377. young poplar and willow planting types. New Zealand Journal of Forestry Trotter, C, Tate, K, Scott, N, Townsend, J, Wilde, H, Lambie, S, Marden, M, Pinkney, Science, 44:15. T. (2005). Afforestation/reforestation of New Zealand marginal pasture lands Phillips, CJ, Marden, M, Lambie, S. (2015). Observations of “coarse” root by indigenous shrub lands: the potential for Kyoto forest sinks. Annals of development in young trees of nine exotic species from a New Zealand plot Forestry Science, 62, 865–871. trial. New Zealand Journal of Forestry Science, 45: 13. Wagner, RG, & Ter-Mikaelian, MT. (1999). Comparison of biomass component equations for four species of northern coniferous tree seedlings. Annals of Phillips, CJ, Marden, M, Lambie, S, Watson, A, Ross, C, Fraser, S. (2012). Forestry Science, 56, 193–199. Observations of below-ground characteristics of young redwood trees (Sequoia sempervirens) from two sites in New Zealand- implications for Walters, MB, & Reich, PB. (1996). Are shade tolerance, survival, and growth linked? erosion control. Plant and Soil. https://doi.org/10.1007/s11104-012-1286-4. Low light and nitrogen effects on hardwood seedlings. Ecology, 77(3), 841–853. Phillips, CJ, Marden, M, Rowan, D, Ekanayake, JC (2001). Stabilising characteristics Wardle, P (2002). Vegetation of New Zealand. New Jersey: The Blackburn Press. of native riparian vegetation in New Zealand. In Proceedings of 3rd Australian Watson, AJ, Marden, M, Rowan, D (1995). Tree species performance and slope Stream Management Conference, (pp. 507–512). Brisbane: Cooperative stability. In DH Barker (Ed.), Vegetation and slopes, (pp. 161–171). London: Research centre for catchment Hydrology. Thomas Telford Press. Pollock, K. M. (1986). Plant materials handbook for soil conservation, Vol 3: Native Watson, AJ, & O’Loughlin, CL. (1990). Structural root morphology and biomass of plants. Water and soil miscellaneous publications no. 95, Palmerston North: three age classes of Pinus radiata. New Zealand Journal of Forestry Science, 20, Soil Conservation Centre, Ministry of Works and Development. 97–110. Poole, AL, & Adams, NM (1994). Trees and shrubs of New Zealand: field guide series. Watson, AJ, Phillips, CJ, Marden, M. (1999). Root strength, growth, and rates of Canterbury: Manaaki Whenua Press. decay: root reinforcement changes of two tree species and their Russo, SE, Jenkins, KL, Wiser, SR, Uriate, M, Duncan, RP, Coombs, DA. (2010). contribution to slope stability. Plant & Soil, 217,39–47. Interspecific relationships among growth, mortality and xylem traits of Watson, AJ, & Tombleson, JD. (2002). Toppling in juvenile pines: a comparison of woody species from New Zealand. Functional Ecology, 24, 253–262. the root system characteristics of direct-sown seedlings, and bare-root Salmon, JT (1986). The reed field guide to New Zealand native trees. Auckland: seedlings and cuttings. Plant and Soil, 239, 187–196. Reed Books. Watson, AJ, & Tombleson, JD. (2004). Toppling in young pines: temporal changes Schiechtl, HM (1980). Bioengineering for land reclamation and conservation. in root system characteristics of bare-root seedlings and cuttings. New Edmonton: University of Alberta Press. Zealand Journal of Forestry Science, 34(1), 39–48. Schiechtl, HM, & Stern, R (1994). Handbuch fur naturnahen Wasserbau, Eine Anleitung Will, GM (1966). Root growth and dry-matter production in a high-producing stand fur ingenieurbiologische Bauweisen. Wien: Osterreichischer Agrarverlag. of Pinus radiata. Research Notes 44. Rotorua: New Zealand Forest Service, Forest Research Institute. Schmidt, KM, Roering, JJ, Stock, JD, Dietrich, WE, Montgomery, DR, Schaub, T. (2001). Root cohesion variability and shallow landslide susceptibility in the Williams, J (1978). Carbon dioxide, climate and society: proceedings of an Oregon Coast Range. Canadian Geotechnical Journal, 38, 995–1024. International Institute for Applied Systems Analysis (IIASA) workshop. February Schwarz, M, Lehmann, P, Or, D. (2010). Quantifying lateral root reinforcement in 21–24, 1978. Oxford: Pergamon Press. steep slopes from a bundle of roots to tree stands. Earth Surface Processes Wilmshurst, JM. (1997). The impact of human settlement on vegetation and soil and Landforms, 35(3), 354–367. https://doi.org/10.1002/esp.1927. stability in Hawke’s Bay, New Zealand. Journal of Botany (NZ), 35,97–111. Marden et al. New Zealand Journal of Forestry Science (2018) 48:9 Page 26 of 26 Wu, TH, McKinnel, WP, Swanston, DN. (1979). Strength of tree roots and landslides on Prince of Wales Island. Alaska Canadian Geotechnical Journal, 16,19–33. Yeates, JS. (1924). The root nodules of New Zealand pines. New Zealand Journal of Science and Technology, 7, 121–124. Zianis, D, & Mencuccini, M. (2003). Aboveground biomass relationships for beech (Fagus moesiaca Cz) trees in Vermio Mountain, Northern Greece, and generalised equations for Fagus sp. Annals of Forestry Science, 60, 439–448.

Journal

New Zealand Journal of Forestry ScienceSpringer Journals

Published: Jun 18, 2018

References