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

Learn More →

Differential responses of amphibians and reptiles to land‐use change in the biodiversity hotspot of north‐eastern Madagascar

Differential responses of amphibians and reptiles to land‐use change in the biodiversity hotspot... IntroductionDemand for agricultural goods is on the rise due to increasing per‐capita consumption, changing diets, and a growing world population (Tilman et al. 2011). This situation is leading to both an expansion of croplands into natural areas and an intensification within existing production systems (Tscharntke et al. 2012). Most agricultural expansion in the tropics happens at the expense of forests and leads to an increase in forest fragmentation (Hansen et al. 2020). This expansion affects all aspects of biodiversity: research has documented losses in species richness (Scales & Marsden, 2008), shifts in species composition (Newbold et al. 2016), reductions of functional diversity (Matuoka et al. 2020), and erosion of phylogenetic diversity (Li et al. 2020). Importantly, the response of biodiversity to land‐use change differs among species groups (Newbold et al. 2014) and world regions (Williams & Newbold, 2020), but certain groups and regions are understudied compared with others. While vertebrates are comparatively well studied, we know less about the responses of amphibians and reptiles than we know about mammals and birds (Newbold et al. 2014). Similarly, we know less about what happens in the more biodiverse tropics compared with temperate regions (Gardner et al. 2009). In sum, we know that land‐use change is the main driver of biodiversity decline globally (Powers & Jetz, 2019), but how the diversity of certain understudied taxa differs between land‐use types remains an open question, especially in the tropics, where high land‐use pressure and biodiversity coincide (Newbold et al. 2020).Tropical agricultural landscapes contribute to food security but also provide opportunities for nature conservation (Perfecto & Vandermeer, 2010). The quality of the agro‐ecosystem for biodiversity conservation depends on the farming systems. Agricultural landscapes dominated by large‐scale industrial monocultures have lower conservation value than diverse mosaics of forest fragments, agroforestry systems, and more intensively farmed annual crop fields (Mendenhall et al. 2016; Murray & Nowakowski, 2021). For example, Ricciardi et al. (2021) reported that small‐scale agricultural landscapes host more biodiversity than large‐scale ones and De Palma et al. (2015) showed that low‐intensity agriculture can maintain relatively diverse bee communities. Furthermore, forest patches within tropical small‐scale agricultural landscapes increase the variety of ecological niches and benefit species with small ranges in particular (Gray et al. 2016). Besides the value for biodiversity, small‐scale land‐use mosaics can also provide essential ecosystem services and livelihoods for rural people, making such landscapes work for humans and nature (Kremen & Merenlender, 2018). Most research investigating the value of tropical agricultural landscapes for biodiversity and humans was conducted in the Neotropics (Mendenhall et al. 2016) while the conservation value of Afrotropical agricultural landscapes is less understood (Powers et al. 2011). Furthermore, it remains largely unclear how different forms of forest conversion – that is with the use of fire (slash‐and‐burn shifting cultivation) or without the use of fire (forest degradation and forest‐derived agroforestry) – affect amphibian and reptile diversity in resulting land‐use types.Here, we investigate how herpetofauna taxonomic diversity and species composition differ across prevalent burned and unburned land‐use types in Madagascar, providing insights into the vulnerability of amphibians and reptiles to various forms of land‐use change. Madagascar has lost 44% of forest cover since the 1950s, mainly due to the transformation of natural ecosystems to agricultural lands (Vieilledent et al. 2018). Along with unsustainable extraction rates (Whitehurst et al. 2009), this has resulted in more than half of evaluated Malagasy vertebrate species being at risk of extinction (IUCN, 2019). Outstanding levels of endemism (Goodman & Benstead, 2005) and ongoing threats qualify Madagascar as a global biodiversity hotspot (Myers et al. 2000). While the forests and protected areas of the island are increasingly well surveyed, the biodiversity in the agricultural landscapes is less known (Irwin et al. 2010). Focusing on amphibians and reptiles is particularly relevant since Madagascar has a diverse and highly endemic herpetofauna: amphibian species diversity is currently estimated at around 370 (AmphibiaWeb, 2020) and almost all species are endemic (Goodman & Benstead, 2005). Reptile diversity stands at around 440 species (Uetz et al. 2021) with 91% endemism (Goodman & Benstead, 2005). Due to cryptic taxonomic complexes, many species still await discovery or description, suggesting that total species richness numbers will increase further (Vieites et al. 2009). Globally, herpetofauna is sensitive to various anthropogenic threats (Vallan, 2000) including chytrid fungi, environmental pollution, collection for the pet trade, climate change, and conversion of forest habitat into agricultural lands (Hof et al. 2011). All of these factors are also threatening Malagasy herpetofauna, with deforestation being the top pressure (Cordier et al. 2021).Our study focuses on north‐eastern Madagascar. The area retains more forest cover than other parts of the country (Vieilledent et al. 2018) and is a global priority area for amphibian research (Nori et al. 2018). Besides being known for its remarkable biodiversity, north‐eastern Madagascar is also a global center for vanilla cultivation (Hänke et al. 2018). The price boom of the spice between 2012 and 2019 has triggered an expansion of vanilla agroforests (Llopis et al. 2019), and roughly 80% of rural households in the study region farm vanilla (Hänke et al. 2018). How the expansion of vanilla agroforestry is impacting biodiversity is not well known, despite that the majority of global vanilla cultivation takes place across three biodiversity hotspots (Madagascar, Indonesia, and MexicoMyers et al. 2000; FAO, 2020). Investigating how vanilla agroforestry compares with other land‐use types in terms of biodiversity conservation thus represents a research gap. Besides farming vanilla, the rural population in north‐eastern Madagascar also practices slash‐and‐burn shifting cultivation for hill rice production. Valleys and plains in the study region are commonly occupied by irrigated rice paddies, forming the backbone of staple crop supply. We thus compared the taxonomic diversity and community composition of amphibians and reptiles among plots in six land‐use types within the small‐scale agricultural landscape and compared this with continuous old‐growth forest inside Marojejy National Park. Within the small‐scale agricultural landscape, we sampled vanilla agroforests of contrasting land‐use history (forest‐ and fallow‐derived vanilla agroforests), as well as forest fragment, herbaceous fallow, woody fallow, and rice paddy. Following Cordier et al. (2021) and Gonzalez & González‐Trujillo (2021), we hypothesized that taxonomic diversity would be highest in old‐growth forest (for both amphibians and reptiles). We further expected losses in taxonomic diversity and shifts in species composition after old‐growth forest conversion, particularly in those cases where land‐use conversion involves the use of fire for slash‐and‐burn shifting cultivation. Based on previous research (Martin et al. 2020), we further expected that rehabilitation through agroforestry on fallow land would positively affect the taxonomic diversity of the herpetofauna.Materials and methodsStudy region and study designWe conducted our study in the SAVA region in north‐eastern Madagascar (Fig. 1a,b) where forests outside protected areas are now highly fragmented (Vieilledent et al. 2018) and the landscape is dominated by smallholder agriculture.1FigureOverview of the study region and the study design. (a) SAVA region in north‐eastern Madagascar; (b) study area within SAVA region; (c) study design showing the distribution of 80 plots in 10 villages and at two sites inside Marojejy National Park; and (d) overview of the studied land‐use types and the typical land‐use transformation trajectory from old‐growth forest to forest fragments and agricultural land‐uses. Panel (d) modified after Martin et al. (2021) and Panel (c) modified after Dröge et al. (2021). Rice paddy is not part of the main land‐use trajectory.We collected data on circular plots with a 25 m radius (largest possible plot size in typical smallholder parcels) at low to mid‐altitude (7–819 m above sea level; mean = 192 m ± 207 m) surrounding 10 villages and in Marojejy National Park, a UNESCO World Heritage Site. In each village, we selected seven plots: three vanilla agroforests (forest‐derived and/or fallow‐derived vanilla agroforest), one forest fragment, one herbaceous fallow, one woody fallow, and one rice paddy. In the 10 villages, we first selected 30 vanilla agroforests along a canopy cover gradient. After consultation with the agroforest owner and a visual confirmation on the plot, we found that 20 vanilla agroforests were fallow‐derived while 10 agroforests were forest‐derived. This approach allowed us to understand the response of amphibians and reptiles to the land‐use history of vanilla agroforests. Additionally, we chose 10 plots at two sites (five plots each) inside Marojejy National Park. All land‐use types were replicated 10 times, except fallow‐derived vanilla agroforest, which was replicated 20 times. In total, we surveyed 80 plots (Fig. 1c,d). The selection of 10 plot per land‐use types was a compromise between a large number of replicates and logistical constraints. The average minimum distance between one plot and the next closest plot in the same village/site was 719 m ± 438 m, while the smallest minimum distance between two plots was 260 m as a compromise between independence of plots and logistical constraints.Sampled land‐use typesWe selected 10 old‐growth forest plots at two sites with five plots each. One of the sites has experienced some selective logging in the past but is now well protected (touristic zone in Manantenina valley; site Tourist on Fig. 1), the other site suffers from ongoing occasional selective logging and trapping (Bangoabe; site East on Fig. 1), but we chose plots that did not show signs of recent disturbance. The old‐growth forest plots are a minimum of 300 m from the National Park boundary. In the study region, isolated forest fragments occur around villages and represent remnants of the continuous forest cover that existed in the region prior to the large‐scale deforestation that began in the early 20th century (Gade, 1996). The 10 forest fragments have not been burned in historic times and are all used for the extraction of timber and non‐timber forest products. Herbaceous fallows occur after slash‐and‐burn shifting hill rice cultivation (locally referred to as tavy; Styger et al. 2007) and are sometimes used for grazing cows. The herbaceous fallow plots in this study had last burned at the end of 2016, 1 year before the onset of data collection in 2017. Woody fallows represent successional stages of herbaceous fallows, containing shrubs and small trees. Woody fallows are also occasionally grazed. The woody fallows in our study had last burnt 4–16 years before the onset of data collection in 2017 according to the reports of land‐owners. The climbing vanilla orchid (Vanilla planifolia) is farmed in agroforestry systems with two distinct land‐use histories (following Martin et al. 2020): forest‐derived vanilla agroforests, where vanilla is directly planted inside the forest after removing understory trees and shrubs and managing tall trees for shade (not involving burning). In fallow‐derived vanilla agroforests, vanilla is planted on fallow land which resulted from slash‐and‐burn shifting hill rice cultivation in the past. In these fallow‐derived vanilla agroforests, farmers allow natural regeneration of trees or plant trees to provide shade or as support structures for the vanilla vines. Lastly, we studied irrigated rice paddies that occur in valley bottoms and flood plains. Rice is planted and harvested between one to three times per year. The rice paddies chosen for our study had wider‐than‐average banks to facilitate movement and sampling within the plots.Sampling, data collection, and field identificationTo collect data in the villages, we organized two sampling campaigns during the driest period of the year (October to December 2017 and late August to December 2018) and one campaign during the wettest period (Mid‐January to early April 2018). In Marojejy National Park, we also organized two sampling campaigns during the driest period (late August to early September 2018 and December 2018) and one during the wettest period (February 2019). During each campaign, we visited each plot once during the day (08:00–17:00) and once at night (18:30–23:00). Overall, we did six visits per plot, three times during the day and three times at night.We collected data on amphibian and reptile communities during time‐standardized searches (Kadlec et al. 2012). During the survey, we systematically searched each plot in a zig‐zag pattern (Kadlec et al. 2012). Each search was standardized to 45 min of searching time by two observers. In sum, we thus conducted 270 min of searching time on each plot, summing up to 408 h of searching time across all plots. To detect individuals hiding under rocks, in leaf axils, tree barks, tree holes, leaf‐litter, or deadwood, we actively inspected those microhabitats by lifting removable objects to check underneath.To identify individuals to species level during fieldwork, we used morphological characteristics and referred to keys and visualizations available in “A Field Guide to Amphibians and Reptiles of Madagascar” (Glaw & Vences, 2007) and additional literature (Rakotoarison et al. 2017; Ratsoavina et al. 2019). For those individuals who we could not identify with confidence in the field, we extracted tissue samples for follow‐up DNA analysis and/or collected the specimen. To denote those individuals, we used “cf.” (confer or compare with) and “sp. aff.” (affinis) in conjuncture with the genus name when we found slightly different morphological characters compared with the literature that we used or compared with a given identified species. We used sp. (sp1 or sp2) in conjuncture with the genus name when we could not find any given identified species to compare with the encountered individual. The sp. denotation was thus used to differentiate several unknown species belonging to the same genus and we consider them as morphospecies but used DNA barcodes to confirm presumed species status (see below). Following the field identification, we resumed the searching time; thus abundance and diversity of amphibians and reptiles were independent of the time spent on the plot. Throughout this manuscript, we refer to each encountered individual as an “encounter” rather than an individual as we cannot exclude the possibility of having encountered the same individual at more than one sampling event.Species identification with DNA barcodesWe collected muscle or toe clips as tissue samples of individuals in case of non‐reliable morphological‐characteristics‐based identification. We preserved tissue samples in Eppendorf tubes with 90% ethanol, stored, and analyzed at the Evolutionary Biology laboratory of Prof. Miguel Vences at the University of Braunschweig, Germany. We preserved amphibian specimens in 70% ethanol and deposited them in the collection of the Regional University Centre of the SAVA region (CURSA), Antalaha, Madagascar. We extracted genomic DNA following the standard single‐tube salt extraction protocol (Bruford et al. 1992). To do so, we cut small pieces from the collected tissue and proceeded to protein digestion following the protocol of Cacciali et al. (2019). We then passed the extracted DNA to a PCR (Polymerase Chain Reaction) thermocycler, in which we amplified DNA fragments of two mitochondrial genes, “16S” (Vences et al. 2005) and “COI” with the primers 16SAL (5´‐CGCCTGTTTATCAAAAACAT‐3´) and 16SBH (5´‐CCGGTCTGAACTCAGATCACGT‐3´) from Palumbi et al. (1991). Amplification was set for 45 s at 94–96°C as denaturation, for 45 s at 52°C as hybridization, and for 90 s at 72°C as elongation. The whole process was repeated for 35 cycles. We generated DNA sequences through a sequencer. We first cleaned the obtained sequences with the software CodonCode Aligner version 8.0.2 (Codon Code Corporation, Centerville, MA, USA). Then, we processed newly generated sequences for additional quality control with a Blast search (Altschul et al. 1990) which queries for similar DNA sequences in the GenBank database to provide information on closest taxa or potential contamination. Finally, we built a phylogenetic tree using MEGA‐X (Kumar et al. 2018) to reveal clusters of each newly generated sequence, allowing us to reliably cluster individuals to known or presumed species. We submitted all newly determined sequences to GenBank under accession numbers MZ019097–MZ019430. In this study, the DNA barcoding differentiated between 43 species of amphibians and 24 species of reptiles. Among them were 12 candidates for new species of which eight were documented for the first time (marked as sp. CaNEW) and four were already found by other researchers (marked as sp. Ca) but have not yet been described. The new species were based on a comparison of obtained DNA sequences with available sequences with the same extraction DNA method conducted in the same lab. In case the DNA barcoding method did not yield results due to short sequences or contaminations, we kept the field identification for those species (for 15 species, eight for amphibians, and seven for reptiles).We attributed the red list status to evaluated species following IUCN (2019). We determined endemicity according to AmphibiaWeb (2020) for amphibians and Uetz et al. (2021) for reptiles. Because most amphibian and reptile genera are completely endemic to Madagascar, we could attribute endemicity also to morphospecies.Data analysis and visualization: species diversities and encountersDuring the data analysis, we conducted all statistical analyses using R version 3.5.1 (R Core Team, 2019) and the R package “ggplot2” (Wickham, 2016) for visualization. To compare observed species richness and species encounters at the plot‐level among land‐use types, we computed a generalized linear model with observed species richness as a response, land‐use type as the explanatory variable, and village (respectively old‐growth forest site) as a random factor and with Poisson family for count data as distribution. We then ran the glht function of the R‐package “multcomp” (Hothorn et al. 2008) applying a Tukey all‐pair‐comparison with Bonferroni correction. We then used the same approach to compare observed species richness between plots from unburned land‐use types (old‐growth forest, forest fragment, forest‐derived vanilla) and plots from burned land‐use types (herbaceous fallow, woody fallow, fallow‐derived vanilla).We used the encounter data to compute sample size‐based extrapolation curves (Hsieh et al. 2016) with the “iNEXT” package to assess the diversity (species richness, Shannon, and Simpson diversity) across land‐use types using the Hill number framework (Chao et al. 2014). We calculated species richness, Shannon diversity, and Simpson diversity (Hsieh et al. 2016). Chao & Jost (2012) define q = 0 (0D) as species richness, that is the effective number of species in the community, giving equal weight to frequent and infrequent species; q = 1 (1D) as Shannon diversity, giving more weight to more frequently observed species and q = 2 (2D), as Simpson diversity interpreted as the effective number of abundant species. This approach is now well established given key advantages over traditional Shannon and Simpson diversity indices (Roswell et al. 2021).To display the total species diversity in each land‐use type, we sub‐sampled 10 plots of the 20 fallow‐derived vanilla agroforests. To do so, we randomly selected one fallow‐derived vanilla agroforest from each village. As one of the villages lacks fallow‐derived vanilla agroforests (village Andramanolotra, see Fig. 1), this resulted in nine agroforests. We proceeded to select one additional plot from all remaining fallow‐derived vanilla agroforests, enabling a fair comparison of total species diversity across 10 plots of each land‐use type.To compare the evenness of communities between land‐use types, we additionally plotted the observed and extrapolated diversity across hill numbers for each land‐use type, allowing for a direct comparison of slopes between land‐use types (Chao & Jost, 2015; Roswell et al. 2021).Data analysis and visualization: species compositionTo evaluate the differences in species community composition between land‐use types between burned and unburned plots, and between villages/sites, we used the metaMDS function of the R‐package “vegan” (with 1000 permutations, Oksanen et al. 2020). We used non‐metric dimensional scaling (NMDS) of Bray‐Curtis dissimilarities to visualize the dissimilarity of species composition in two dimensions. Furthermore, to test the differences between land‐use types, we used PERMANOVA (Permutational multivariate analysis of variance) implemented in the adonis function of the “vegan” package (Oksanen et al. 2020) and computed pairwise differences using the pairwise.adonis function with Bonferroni correction of the “pairwise.adonis” package (Martinez, 2020). Since PERMANOVA may confound location and dispersion, we used PERMDISP (Permutational analysis of multivariate dispersions) to see if PERMANOVA results may have been influenced by dispersion differences between groups (Anderson & Walsh, 2013). To do so, we used the function betadisper and permutest and performed pairwise comparisons using the TukeyHSD function. To analyze the degree of forest dependency (Rembold et al. 2017), we plotted the proportion of encounters for each species across seven land‐use types (10 replicates per land‐use type). To visualize the forest dependency, we used the barplot function of R graphics.ResultsEncounters and observed species richnessIn total, we made 6215 encounters and observed 119 species of amphibians and reptiles. The 3694 amphibian encounters belong to 58 species, 15 genera, and four families (see Table S1). The most species‐rich genera of amphibians are Boophis (11 species), Stumpffia (10 species), and Gephyromantis (eight species). We found all but one species (Ptychadena mascareniensis) to be endemic according to AmphibiaWeb (2020). Among the observed amphibian species, 22 (37%) could not be identified to species level in the field. Twelve of them were recognized as candidates for new species based on genetic barcoding, another eight of them were counted as morphospecies based on the criteria mentioned in the methods section, and two were conferred to previously identified species. Among the encountered amphibian species, seven are listed in the “threatened” category (IUCN 2019). Among the threatened species, we recorded six vulnerable and one endangered species.The 2521 reptile encounters represent 61 species, 28 genera, and five families (see Table S1). The most species‐rich reptile genera were Phelsuma (11 species) and Uroplatus (six species). We found 83% of reptile species to be endemic (Uetz et al. 2021). Amongst observed reptile species, 15 could not be identified to species level in the field; of these, seven were counted as morphospecies and eight were conferred to previously identified species. Based on the IUCN red list (IUCN 2019), seven encountered reptile species are listed as “threatened.” Among the threatened species, we recorded four endangered, one critically endangered, and two vulnerable species.Plot‐level observed amphibian species richness differed significantly among land‐use types (F6,73 = 19.59, P < 0.001) and when comparing all burned to all unburned land‐use types (F1,67 = 15.89, P < 0.001; Figure S6 and Table S7). A Tukey post‐hoc test revealed significant pairwise differences between pairs of land‐use types (Fig. 2a and Table S2): old‐growth forest had significantly higher observed amphibian species richness while rice paddy had significantly lower observed species richness compared with other land‐use types (Fig. 2a). The other land‐use types had similarly observed amphibian species richness with no significant differences (Table S2). Plot‐level amphibian encounters also differed significantly between land‐use types (F6,72 = 13.01, P < 0.001; Figure S9, Tables S10 and S12).2FigurePlot‐level observed amphibian (a) and reptile (b) species richness across seven land‐use types (replicated 10 times each except fallow‐derived vanilla, which is replicated 20 times) in north‐eastern Madagascar. Each dot is the observed species richness in a plot. The black horizontal line in the box shows the median. Land‐use types with letters in common did not differ significantly based on pairwise comparisons that controlled for inflated false‐positive errors using the Tukey HSD approach (Numeric results in SI: amphibians & reptiles Table S2). See icon reference in supplementary information (Free Icons Library).Plot‐level observed reptile species richness also varied significantly among land‐use types (F6,73 = 18.55, P < 0.001) and when comparing all burned to all unburned land‐use types (F1,67 = 25.60, P < 0.001; Figure S6 and Table S8). Tukey post‐hoc tests revealed no differences in observed species richness between old‐growth forest and forest fragment (P = 0.88) and between old‐growth forest and forest‐derived vanilla agroforest (P = 0.77). These three land‐use types are recorded with the highest observed species richness, with 10 reptile species on average per plot. Rice paddy had the lowest observed reptile species richness, but there were no significant differences (P = 0.19) compared with herbaceous fallow. Observed reptile species richness was significantly higher in forest‐derived vanilla agroforest than in fallow‐derived vanilla agroforest (P = 0.01; Fig. 2b and Table S2). Plot‐level reptile encounters also differed significantly between land‐use types (F6,73 = 14, P < 0.001; Figure S9, Tables S11 and S12).Species accumulation curves, diversity, evenness, and estimated species richnessEncounter‐based accumulation curves for amphibians revealed the highest species richness, Shannon diversity, and Simpson diversity in old‐growth forest and the lowest in rice paddy. In most land‐use types, observed amphibian species richness curves flattened off, except for old‐growth forest and forest fragment. The overlap of the 95% confidence interval of extrapolated amphibian richness for old‐growth forest and forest fragment indicates no differences in species richness. Regarding the accumulated amphibian species diversity for Shannon and Simpson diversity, old‐growth forest varied significantly compared with all other land‐use types. Shannon and Simpson diversity also did not differ significantly between forest fragment and forest‐derived vanilla agroforest, respectively, compared with burned land uses, except to rice paddy (Fig. 3 and Table 1).3FigureEncounter‐based accumulation curves showing interpolated (solid line) and extrapolated (dotted line) diversities for amphibians (first and second columns of panel from left to right) and reptiles (third and fourth panels) in north‐eastern Madagascar. Unburned (first and third panels) and burned (second and fourth panels) land‐use types are separated. The species richness represented by q = 0 (first row of panel from above), Shannon diversity, q = 1 (second row of panel) and Simpson diversity q = 2 (third row of panel) with 95% confidence intervals (shaded areas) for the amphibian and the reptile data of seven land‐use types. The solid dots, triangles, and diamonds represent the reference samples, that is the observed number of encounters and species richness. See icon reference in supplementary information (Free Icons Library).1TableEstimated and observed amphibian and reptile species richness (q = 0) for all land‐use types and separated per land‐use type showing the observed and extrapolated total species richnessSpecies Group (measure)Land‐use typeAll 80 plotsOld‐growth forestForest fragmentForest‐derived vanilla agroforestFallow‐derived vanilla agroforestWoody fallowHerbaceous fallowRice paddyAmphibian (observed)5832261416864Amphibian (extrapolated)NA60.0 (20.4–99.6)46.1 (12.4–79.8)18.0 (9.1–26.9)25.0 (7.0–42. 9)9.0 (7.4–10.6)6.5 (3.9–9.1)5.0 (4.3–5.6)Reptile (observed)613430301819811Reptile (extrapolated)NA42.0 (20.4–63.5)40.1 (10.2–69.9)35.1 (18.2–52.4)26.9 (8.2–45.7)46.6 (11.5–81.7)8.0 (6.8–9.2)18.9 (4.3–33.4)Total (observed)11966564434271415Each land‐use type is represented by 10 plots; for fallow‐derived vanilla, the 10 plots are down‐sampled from 20 plots. Extrapolated species richness is based on 5000 encounters and includes the lower and upper 95% confidence interval in italics in the brackets. See SI for results of q = 1 and q = 2 (amphibians and reptiles: Table S3).Encounter‐based accumulation curves for reptiles revealed the highest species richness and Shannon diversity in old‐growth forest and the lowest in herbaceous fallow. Observed species richness curves flattened off in all land‐use types except woody fallow and fallow‐derived vanilla agroforest. We also found an overlap of the 95% confidence intervals of extrapolated species richness as well as Shannon and Simpson diversity within unburned and burned land‐use types except for Shannon and Simpson diversity in burned land‐use types (Fig. 3 and Table 1).The species diversity dropped more strongly across hill number orders (q = 0 to q = 2) in amphibians than in reptiles after old‐growth forest transformation (Fig. 3, Table S3 and Figure S5), highlighting that amphibian communities were more uneven than reptile communities (Figure S5). Furthermore, the comparison shows that old‐growth forest communities of both amphibians and reptiles were most uneven (Figure S5).Species composition and forest dependencyThe composition of amphibian communities differed significantly across land‐use types (PERMANOVA: R2 = 0.50, P < 0.001, Df = 6), partly driven by differences in group dispersion (PERMDISP: F = 4.400, P = 0.004, Df = 6, Tables S18 and S19). Pairwise comparisons showed that amphibian communities in old‐growth forest and rice paddy were significantly different from those of other land‐use types. Forest fragment differed significantly from fallow‐derived vanilla agroforest, woody fallow, and herbaceous fallow, but no significant differences were observed between forest fragment and forest‐derived vanilla agroforest. We found no significant differences between forest‐derived vanilla agroforest, fallow‐derived vanilla agroforest, woody fallow, and herbaceous fallow for amphibian communities (Fig. 4a, see Table S4). We compared also the amphibian communities between unburned and burned land‐use types and found significant differences (PERMANOVA: R2 = 0.193, P < 0.001, Df = 1, Figure S13 and Table S14), again partly drive by dispersion (PERMDISP: F = 34.718, P = 0.001, Df = 1, Table S16). Similarly, amphibian communities differed between villages and Marojejy National Park communities (PERMANOVA: R2 = 0.416, P < 0.001, Df = 10, Figure S21, Tables S22 and S24), in this case not driven by differences in group dispersion (PERMDISP: F = 1.254, P = 0.283, Df = 11, Table S25). Here, communities were very similar between villages (Figure S21).4FigureSpecies composition across seven land‐use types in north‐eastern Madagascar. Non‐metric dimensional scaling (NMDS) showing community dissimilarity of amphibian (a) and reptile (b) communities. Old‐growth forest shows unique species composition for both amphibians and reptiles. See icon reference in supplementary information (Free Icons Library).Reptile species composition showed significant differences among land‐use types (PERMANOVA: R2 = 0.40, P < 0.001, Df = 6) not driven by differences in group dispersion (PERMDISP: F = 1.870, P = 0.088, Df = 6, Table S20) indicating the homogeneity of group dispersion. Pairwise comparisons showed that the reptile communities in old‐growth forest and forest fragment were significantly different from all other land‐use types. The communities in other land‐use types showed no significant differences (Fig. 4b and Table S4). Old‐growth forest showed unique communities of reptiles and amphibians. The reptile species composition between unburned and burned land‐use types showed significant differences (PERMANOVA: R2 = 0.210, P = .001, Df = 1, Table S15) not driven by differences in group dispersion (PERMDISP: F = 3.918, P = 0.057, Df = 1, Table S17). The reptile communities between villages and Marojejy National Park showed significant differences (PERMANOVA: R2 = 0.389, P < 0.001, Df = 10, Table S23 and S24) not driven by differences in group dispersion (PERMDISP: F = 0.835, P = 0.626, Df = 11, Table S26).Finally, the analyses of forest dependency revealed that 22 amphibian species (38%) were found exclusively in old‐growth forest, 11 amphibian species (19%) exclusively in forest fragment, three amphibian species (5%) exclusively in forest‐derived vanilla agroforest, and four amphibian species (7%) exclusively in fallow‐derived vanilla agroforest (Fig. 5a).5FigureProportion of species encounters across seven land‐use types (with 10 replicates each) in north‐eastern Madagascar. All 58 amphibian species with 3694 encounters (a) and all 61 reptile species (b) with 2521 encounters are displayed by forest dependency rank: 38% of amphibian species and 26% of reptile species exclusively occurred in old‐growth forest, despite that old‐growth forest is only accounting for 12.5% of the total plots. See icon reference in supplementary information (Free Icons Library).We observed 16 reptile species (26%) exclusively in old‐growth forest, three reptile species (5%) exclusively in forest fragment, four reptile species (7%) exclusively in forest‐derived vanilla agroforest, and one reptile species (2%) exclusively in fallow‐derived vanilla agroforest (Fig. 5b). No species were exclusively found in woody fallow, herbaceous fallow, and rice paddy for both groups.DiscussionIn this study, we provide a comprehensive assessment of the response of amphibian and reptile species diversity to land‐use change in the biodiversity hotspot of north‐eastern Madagascar. Overall, we document a highly diverse herpetofauna with 58 amphibian and 61 reptile species and presumed species, and a high proportion of endemic species (98% for amphibians and 83% for reptiles).At a plot‐level, observed species richness for both amphibians and reptiles was very high, with up to 12 and 14 species, respectively. For both groups, old‐growth forest was significantly different from all other land‐use types in terms of total estimated species richness and community composition. Rice paddy and herbaceous fallow harbored the lowest observed species richness for amphibians and reptiles, respectively. Reptile species diversity varied significantly between forest‐derived and fallow‐derived vanilla agroforests whereas amphibian species richness showed no differences. We found no unique species in woody fallow, herbaceous fallow, or rice paddy plots for either taxon. Importantly, we found that amphibians and reptiles responded differently to land‐use history. After any kind of old‐growth forest conversion, whether through burning or not, amphibian species communities both declined in richness and shifted in structure away from old‐growth forest. Deforestation for shifting cultivation through slash‐and‐burn led to species losses in reptiles as well, but changes were less pronounced if old‐growth forest was transformed to forest fragment or forest‐derived vanilla agroforest since these transformations refrain from using fire.Outstanding diversity of amphibians and reptiles in north‐eastern MadagascarThe diversity of amphibians and reptiles documented here was exceptional, both within Madagascar and compared with other tropical biodiversity hotspots. The few available studies on the response of amphibian communities to land‐use change in Madagascar reported lower overall values, with 32 and 28 species, respectively (Vallan, 2000; Ndriantsoa et al. 2017). Andreone et al. (2000) found 42 amphibian and 23 reptile species in north‐eastern Malagasy rainforest. Despite differences in methodology and study sites, our study provides additional evidence that the north‐east of Madagascar is one of the most species‐rich regions for herpetofauna in the country, as previously suggested (Brown et al. 2016). The documented diversity also exceeds values found in other tropical biodiversity hotspots (Cordier et al. 2021; Murray & Nowakowski, 2021) – for example, Urbina‐Cardona et al. (2006) found 21 amphibian and 33 reptile species in the region of Veracruz, Mexico, and Kurz et al. (2014) found 25 amphibian and 20 reptile species in a region of north‐eastern Costa Rica (versus 58 amphibian and 61 reptile species in this study). The high species diversity may partly be driven by our extensive sampling effort, in terms of number of plots (total n = 80), diversity of different land‐use types (seven), and search effort (total of 270 min searching time by two observers per plot), and is likely also influenced by the use of genetic samples for species identification.Response of amphibian diversity to land‐use changeWe found a strong negative response of amphibian species richness to any anthropogenic land use. Old‐growth forest had a significantly higher observed amphibian species richness with 38% species exclusively occurring there. The amphibian species composition is unique compared with all other land‐use types, driven by a high share belonging to the Microhylidae family and the Spinomantis, Guibemantis, and Mantella genera (Fig. 5a). Only forest fragments harbored comparable accumulated amphibian species richness. Subsequently, we infer that Malagasy amphibians are very sensitive to habitat change. This result is in line with findings from Ndriantsoa et al. (2017) in disturbed habitats of eastern Madagascar which documented strongly impoverished frog communities in secondary vegetation and rice fields compared with forests. Forest fragments are thus valuable for Malagasy amphibians to maintain diversity within the agricultural landscape but cannot substitute large and continuous old‐growth forest (Vallan, 2000).The other land‐use types (fallows that form part of the slash‐and‐burn cycle, vanilla agroforests, and rice paddies) are of minor importance for amphibian conservation, given the low species diversity that consists mainly of common and widely ranging species. Nonetheless, amphibians could play an important functional role in these habitats: the abundance of amphibians is high throughout, reflected by the high number of encounters, particularly in rice paddies (see Fig. 3b and Figure S9). As such, they may be an important food source for other taxa or could provide pest control services in crops (Hocking et al. 2014).The strong negative response of amphibians to deforestation could be explained by the fact that numerous species rely on moist environments to avoid dehydration (Watling & Braga, 2015), especially species living in evergreen forest habitat (Hof et al. 2011). Given that selective logging, forest fragmentation, and deforestation severely change microclimate conditions (Ewers & Banks‐Leite, 2013), many species may struggle to cope. Furthermore, amphibians have different habitat requirements throughout their life cycle stages (Hof et al. 2011; Hocking et al. 2014). Such heterogeneity is often lost in concert with the simplification in forest structure (Wanger et al. 2009), making new habitats unsuitable for many species.Response of reptile diversity to land‐use changeWe found a strong effect of land‐use history on reptile diversity. Unburned land‐use types had a significantly higher average species richness, more uneven communities, more unique species, and a distinct species composition compared with burned land‐use types and rice paddy. The distinct species composition in unburned land‐use types was caused by an increased number of species belonging to the Chamaeleonidae family and the Uroplatus genus (Fig. 5b) compared with other land‐use types. This is, to our knowledge, the first study shedding light on the response of reptiles to land‐use change in Madagascar’s humid eastern escarpment. Studies from the drier southern region (Scott et al. 2006; Gardner et al. 2016; Nopper et al. 2018) show that Malagasy reptile species react less strongly to habitat change than other taxa, especially if diurnal (Nopper et al. 2018).Interestingly, the less pronounced effects of land‐use change on reptiles compared with amphibians in tropical landscapes has also been demonstrated by others (Wanger et al. 2009; Kurz et al. 2014) and was found in a review by Palacios et al. (2013). This may be due to the high thermotolerance in reptiles (Thompson & Donnelly, 2018), thanks to a fairly impermeable skin covered with scales. This structure prevents excessive evaporative water loss and adapts reptiles to various microclimates (Watling & Braga, 2015). Furthermore, egg development is not heavily impacted by temperature rise, which reduces the size of hatchlings but accelerates incubation (Packard et al. 1982). Reptile traits associated with more open, hotter, and drier environments (Morin, 2005) may make them less vulnerable to land‐use change.Land‐use history of vanilla agroforests affects reptiles but not amphibiansIn our study, we have separated forest‐derived vanilla from fallow‐derived vanilla agroforests, thereby explicitly accounting for land‐use history (Martin et al. 2020). For amphibians, we found no differences between the two agroforestry systems across metrics. Reptile communities, on the other hand, were significantly more diverse in forest‐derived vanilla agroforests on plot‐level, more species‐rich overall, more uneven, and compositionally different compared with fallow‐derived vanilla agroforest. Reptile communities observed in forest‐derived vanilla agroforests are thus more similar to old‐growth forest and forest fragment. The reptile communities recorded in fallow‐derived vanilla agroforests were comparable with fallow land. Both kinds of agroforest thus resembled the land‐use types they were derived from. However, we observed an increase in accumulated amphibian species richness from fallow land to fallow‐derived vanilla agroforest, which suggests rehabilitation through agroforestry (Osen et al. 2021). These findings are in line with the prediction from a recent study by Martin et al. (2020) that found that the land‐use history of agroforestry systems matters for biodiversity.We further hypothesize that the strong importance of forest‐derived vanilla agroforests for reptiles may be in part driven by leaf‐litter depth, which is known to positively influence reptile diversity and abundance (Urbina‐Cardona et al. 2006), and which is typically deeper in forest‐derived vanilla agroforests. According to the same study, these microenvironmental variables also showed positive effects on amphibians and thus cannot explain why there are no differences between the two kinds of agroforests for amphibians. Elsewhere, the loss of canopy cover had negative impacts on amphibian and reptile species richness (Scott et al. 2006). Since canopy cover differed significantly between forest‐derived vanilla agroforest (higher) and fallow‐derived vanilla agroforest (lower) in the same study site (Osen et al. 2021), this calls for further investigation combining habitat characteristics with species traits (Oliveira et al. 2017) to elucidate the drivers of change in both species groups.Conservation implicationsThe strong negative response of amphibians to old‐growth forest modification and the high old‐growth forest dependency of amphibians and reptiles calls for the effective protection of the last remaining old‐growth forests. Additionally, conserving forest fragments within the agricultural landscape will be important to many reptile and amphibian species that are absent from other land‐use types. This is particularly important given that numerous species are micro‐endemics (Brown et al. 2014), meaning that they only occur very locally (Vieites et al. 2009), possibly in only a handful of forest fragments. Management strategies are thus needed to safeguard the long‐term existence of forest fragments and to improve the connectivity between fragments. The protection of large continuous forests throughout the region is also important under a changing climate (Hof et al. 2011) and in light of emerging threats, such as the chytrid fungi (Kolby & Skerratt, 2015) and the recent spread of the invasive Asian common toad (Duttaphrynus melanostictus, Soorae et al. 2020). These conservation needs are underscored by the exceptional diversity of reptiles and amphibians in north‐eastern Madagascar as well as by the high proportion of endemic species.Our findings from north‐eastern Madagascar also have broader relevance to herpetofauna conservation in other tropical biodiversity hotspots. The differential response of amphibians and reptiles to land‐use change shows that conservationists should not treat herpetofauna as a homogenous group when devising conservation programs. Furthermore, the distinct differences in reptile communities between plots of contrasting land‐use history shows that past land‐use must also be considered, and suggests that the maintenance of remnant forest fragments and forest‐derived agroforests is important. Lastly, we highlight the importance of unburned compared with burned land‐use types in the agricultural matrix for the conservation of reptiles.AcknowledgmentsThe authors are grateful to all chef de fokontany, landowners, and Madagascar National Parks for granting them access to sites and information. They thank Prof. Miguel Vences and his laboratory staff for invaluable support with the DNA barcoding, Saskia Dröge for preparing panel C of Fig. 1, James Herrera for feedback, and Prof. Miguel Acevedo, Prof. Arne Mooers, Philippe Fernandez‐Fournier, and an anonymous reviewer for their constructive reviews and feedback. They also thank “CodonCode grants” for offering them a free license under the CodonCode Aligner License Grant Program. They collected data under research permits N100/17/MEEF/SG/DGF/DSAP/SCB. Re, N°163/17/MEEF/SG/DGF/DSAP/SCB. Re, N°18/18/MEEF/SG/DGF/DSAP/SCB. Re and N°254/18/MEEF/SG/DGF/DSAP/SCB. Re granted by the Ministry of Environment and Sustainable Development, Antananarivo. They transported DNA samples domestically under the transport permit N° 34/17‐MEEF/SG/DREEF/SAVA/SRF granted by the Regional Office of the Ministry of Environment and Sustainable Development, Sambava. They exported the samples from Madagascar under CITES permits N°323C‐EA05/MG18 and N°180C‐EA03/MG19 granted by the Ministry of Environment and Sustainable Development, Antananarivo, and imported the samples to Germany under CITES permits DE‐E‐03377/18, DE‐E‐03378/18, DE‐E‐02422/19, DE‐E‐02423/19, DE‐E‐02424/19, and DE‐E‐02425/19 granted by the German Federal Agency for Nature Conservation, Bonn. This study was financially supported by the Niedersächsisches Vorab of Volkswagen Foundation as part of the research project “Diversity Turn in Land Use Science” (Grant number 11‐76251‐99‐35/13 (ZN3119)) and by the German Academic Exchange Service (DAAD) within the “Partnerships for Supporting Biodiversity in Developing Countries” initiative (Project Nr. 57449386). Open Access funding enabled and organized by Projekt DEAL. WOA Institution: GEORG‐AUGUST‐UNIVERSITAET GOTTINGEN. Blended DEAL: Projekt DEALConflict of interestNo conflict of interest to declare.Author’s contributionsTRF, DAM, KO, AW, HK, AA, and FMR designed the study. TRF, EB, RB, and RR collected amphibian and reptile data under the lead of TRF. TRF and DAM analyzed the data. TRF and DAM wrote the first manuscript draft. All authors revised the manuscript.Data Availability StatementData are available from Zenodo: http://doi.org/10.5281/zenodo.4548955ReferencesAltschul, S.F., Gish, W., Miller, W., Myers, E.W. & Lipman, D.J. (1990). Basic local alignement search tool. J. Mol. Biol. 215, 403–410.AmphibiaWeb. (2020). List of amphibians of Madagascar. Retrieve from: AmphibiaWeb., org/cgi/amphib_query?rel‐isocc=like&orderbyaw=Order&where‐isocc=Madagascar. Accessed December 14, 2020.Anderson, M.J. & Walsh, D.C.I. (2013). PERMANOVA, ANOSIM, and the Mantel test in the face of heterogeneous dispersions: what null hypothesis are you testing? Ecol. Monogr. 83, 557–574.Andreone, F., Randrianirina, J.E., Jenkins, P.D. & Aprea, G. (2000). Species diversity of Amphibia, Reptilia and Lipotyphla (Mammalia) at Ambolokopatrika, a rainforest between the Anjanaharibe‐Sud and Marojejy massifs, NE Madagascar. Biodivers. Conserv. 9, 1587–1622.Brown, J.L., Cameron, A., Yoder, A.D. & Vences, M. (2014). A necessarily complex model to explain the biogeography of the amphibians and reptiles of Madagascar. Nat. Commun. 5, 1–10.Brown, J.L., Sillero, N., Glaw, F., Bora, P., Vieites, D.R. & Vences, M. (2016). Spatial biodiversity patterns of Madagascar’s amphibians and reptiles. PLoS One 11, 1–26.Bruford, M.W., Hanotte, O., Brookfield, J.F.Y. & Burke, T. (1992). Single‐locus and multilocus DNA fingerprinting. In: Molecular genetic analysis of populations: A practical approach: 225–269. Hoelzel, A.R. (Ed.). Oxford: IRL Press.Cacciali, P., Buongermini, E. & Köhler, G. (2019). Barcoding analysis of Paraguayan squamata. Diversity 11, 152.Chao, A., Gotelli, N.J., Hsieh, T.C., Sander, E.L., Ma, K.H., Colwell, R.K. & Ellison, A.M. (2014). Rarefaction and extrapolation with Hill numbers: a framework for sampling and estimation in species diversity studies. Ecol. Monogr. 84, 45–67.Chao, A. & Jost, L. (2012). Coverage‐based rarefaction and extrapolation: standardizing samples by completeness rather than size. Ecology 93, 2533–2547.Chao, A. & Jost, L. (2015). Estimating diversity and entropy profiles via discovery rates of new species. Methods Ecol. Evol. 6, 873–882.Cordier, J.M., Aguilar, R., Lescano, J.N., Leynaud, G.C., Bonino, A., Miloch, D., Loyola, R. & Nori, J. (2021). A global assessment of amphibian and reptile responses to land‐use changes. Biol. Conserv. 253, 108863.De Palma, A., Kuhlmann, M., Roberts, S.P.M., Potts, S.G., Börger, L., Hudson, L.N., Lysenko, I., Newbold, T. & Purvis, A. (2015). Ecological traits affect the sensitivity of bees to land‐use pressures in European agricultural landscapes. J. Appl. Ecol. 52, 1567–1577.Dröge, S., Martin, D.A., Andriafanomezantsoa, R., Burivalova, Z., Fulgence, T.R., Osen, K., Rakotomalala, E., Schwab, D., Wurz, A., Richter, T. & Kreft, H. (2021). Listening to a changing landscape: acoustic indices reflect bird species richness and plot‐scale vegetation structure across different land‐use types in north‐eastern Madagascar. Ecol. Indic. 120, 106929.Ewers, R.M. & Banks‐Leite, C. (2013). Fragmentation impairs the microclimate buffering effect of tropical forests. PLoS One 8, e58093.FAO. (2020). FAOSTAT. Food and Agriculture Organisation of the United Nations. http://www.FAO.org/faostat/en/#home.Gade, D.W. (1996). Deforestation and its effects in Highland Madagascar. Mt. Res. Dev. 16, 101–116.Gardner, C.J., Jasper, L.D., Eonintsoa, C., Duchene, J.J. & Davies, Z.G. (2016). The impact of natural resource use on bird and reptile communities within multiple‐use protected areas: evidence from sub‐arid Southern Madagascar. Biodivers. Conserv. 25, 1773–1793.Gardner, T.A., Barlow, J., Chazdon, R., Ewers, R.M., Harvey, C.A., Peres, C.A. & Sodhi, N.S. (2009). Prospects for tropical forest biodiversity in a human‐modified world. Ecol. Lett. 12, 561–582.Glaw, F. & Vences, M. (2007). In Ny Toro‐hay momba ny Amphibia sy ny Reptilia an’i Madagasikara: 3rd edn: 529. Vences, M. & Glaw, V. (Eds.). Cologne: Vences and Glaw Verlag.González, T.M., González‐Trujillo, J.D., Muñoz, A. & Armenteras, D. (2021). Differential effects of fire on the occupancy of small mammals in neotropical savanna‐gallery forests. Perspect. Ecol. Conserv. 19, 179–188.Goodman, S.M. & Benstead, J.P. (2005). Updated estimates of biotic diversity and endemism for Madagascar. Oryx 39, 73–77.Gray, C.L., Hill, S.L.L., Newbold, T., Hudson, L.N., Börger, L., Contu, S., Hoskins, A.J., Ferrier, S., Purvis, A. & Scharlemann, J.P.W. (2016). Local biodiversity is higher inside than outside terrestrial protected areas worldwide. Nat. Commun. 7, 1–7.Hänke, H., Barkmann, J., Blum, L., Franke, Y., Martin, D.A., Niens, J., et al. (2018). Socio‐ economic, land use and value chain perspectives on vanilla farming in the SAVA Region (north‐ eastern Madagascar): The Diversity Turn Baseline Study (DTBS). Discussion Paper, 1806. https://doi.org/10.13140/RG.2.2.22059.80163Hansen, M.C., Wang, L., Song, X.P., Tyukavina, A., Turubanova, S., Potapov, P.V. & Stehman, S.V. (2020). The fate of tropical forest fragments. Sci. Adv. 6, 1–10.Hocking, D.J., Babbitt, K.J. & Hocking, D.J. (2014). Amphibian contributions to ecosystem services. Herpetol. Conserv. Biol. 9, 1–17.Hof, C., Araújo, M.B., Jetz, W. & Rahbek, C. (2011). Additive threats from pathogens, climate and land‐use change for global amphibian diversity. Nature 480, 516–519.Hothorn, T., Bretz, F. & Westfall, P. (2008). Simultaneous inference in general parametric models. Biom. J. 50, 346–363.Hsieh, T.C., Ma, K.H. & Chao, A. (2016). iNEXT: an R package for rarefaction and extrapolation of species diversity (Hill numbers). Methods Ecol. Evol. 7, 1451–1456.Irwin, M.T., Wright, P.C., Birkinshaw, C., Fisher, B.L., Gardner, C.J., Glos, J., Goodman, S.M., Loiselle, P., Rabeson, P., Raharison, J.‐L., Raherilalao, M.J., Rakotondravony, D., Raselimanana, A., Ratsimbazafy, J., Sparks, J.S., Wilmé, L. & Ganzhorn, J.U. (2010). Patterns of species change in anthropogenically disturbed forests of Madagascar. Biol. Conserv. 143, 2351–2362.IUCN. (2019). The IUCN. red list of threatened species. Version 2019‐1. IUCN., Gland, Switzerland. Available from http://www.IUCN.redlist.org. Accessed March 2019.Kadlec, T., Tropek, R. & Konvicka, M. (2012). Timed surveys and transect walks as comparable methods for monitoring butterflies in small plots. J. Insect. Conserv. 16, 275–280.Kolby, J.E. & Skerratt, L.F. (2015). Amphibian chytrid fungus in Madagascar neither shows widespread presence nor signs of certain establishment. PLoS One 10, 1–6.Kremen, C. & Merenlender, A.M. (2018). Landscapes that work for biodiversity and people. Science 362. https://doi.org/10.1126/science.aau6020Kumar, S., Stecher, G., Li, M., Knyaz, C. & Tamura, K. (2018). MEGA X: molecular evolutionary genetics analysis across computing platforms. Mol. Biol. Evol. 35, 1547–1549.Kurz, D.J., Nowakowski, A.J., Tingley, M.W., Donnelly, M.A. & Wilcove, D.S. (2014). Forest‐land use complementarity modifies community structure of a tropical herpetofauna. Biol. Conserv. 170, 246–255.Li, D., Olden, J.D., Lockwood, J.L., Record, S., McKinney, M.L. & Baiser, B. (2020). Changes in taxonomic and phylogenetic diversity in the Anthropocene: changes in biodiversity. Proc. R. Soc. B Biol. Sci. 287, 20200777.Llopis, J.C., Harimalala, P.C., Bär, R., Heinimann, A., Rabemananjara, Z.H. & Zaehringer, J.G. (2019). Effects of protected area establishment and cash crop price dynamics on land use transitions 1990–2017 in north‐eastern Madagascar. J. Land Use Sci. 14, 52–80.Martin, D.A., Andriafanomezantsoa, R., Dröge, S., Osen, K., Rakotomalala, E., Wurz, A., Andrianarimisa, A. & Kreft, H. (2021). Bird diversity and endemism along a land‐use gradient in Madagascar: the conservation value of vanilla agroforests. Biotropica 53, 179–190.Martin, D.A., Osen, K., Grass, I., Hölscher, D., Tscharntke, T., Wurz, A. & Kreft, H. (2020). Land‐use history determines ecosystem services and conservation value in tropical agroforestry. Conserv. Lett. 13, 1–12.Martinez, A.P. pairwise.Adonis: Pairwise Multilevel Comparison Using Adonis. R Package Version 0.0.1. Available online: https://github.com/pmartinezarbizu/pairwiseAdonis. Accessed on 6 December 2020.Matuoka, M.A., Benchimol, M., de Almeida‐Rocha, J.M. & Morante‐Filho, J.C. (2020). Effects of anthropogenic disturbances on bird functional diversity: a global meta‐analysis. Ecol. Indic 116, 106471.Mendenhall, C.D., Shields‐Estrada, A., Krishnaswami, A.J. & Daily, G.C. (2016). Quantifying and sustaining biodiversity in tropical agricultural landscapes. Proc. Natl Acad. Sci. USA 113, 14544–14551.Morin, K.C. (2005). Herpetofaunal responses to prescibed fire in upland pine communities of Northeast Florida: 77 pp. PhD Thesis. University of Florida.Murray, A.H., Nowakowski, A.J. & Frishkoff, L.O. (2021). Climate and land‐use change severity alter trait‐based responses to habitat conversion. Global Ecol. Biogeogr. 30, 598–610.Myers, N., Mittermeler, R.A., Mittermeler, C.G., Da Fonseca, G.A.B. & Kent, J. (2000). Biodiversity hotspots for conservation priorities. Nature 403, 853–858.Ndriantsoa, S.H., Riemann, J.C., Raminosoa, N., Rödel, M.O. & Glos, J.S. (2017). Amphibian diversity in the matrix of a fragmented landscape around Ranomafana in Madagascar depends on matrix quality. Trop. Conserv. Sci. 10, 194008291668606.Newbold, T., Hudson, L.N., Hill, S.L.L., Contu, S., Gray, C.L., Scharlemann, J.P.W., Börger, L., Phillips, H.R.P., Sheil, D., Lysenko, I. & Purvis, A. (2016). Global patterns of terrestrial assemblage turnover within and among land uses. Ecography 39, 1151–1163.Newbold, T., Hudson, L.N., Phillips, H.R.P., Hill, S.L.L., Contu, S., Lysenko, I., Blandon, A., Butchart, S.H.M., Booth, H.L., Day, J., De Palma, A., Harrison, M.L.K., Kirkpatrick, L., Pynegar, E., Robinson, A., Simpson, J., Mace, G.M., Scharlemann, J.P.W. & Purvis, A. (2014). A global model of the response of tropical and sub‐tropical forest biodiversity to anthropogenic pressures. Proc. R. Soc. B Biol. Sci. 281, 20141371.Newbold, T., Oppenheimer, P., Etard, A. & Williams, J.J. (2020). Tropical and Mediterranean biodiversity is disproportionately sensitive to land‐use and climate change. Nat. Ecol. Evol. 4, 1630–1638.Nopper, J., Riemann, J.C., Brinkmann, K., Rödel, M.O. & Ganzhorn, J.U. (2018). Differences in land cover – biodiversity relationships complicate the assignment of conservation values in human‐used landscapes. Ecol. Indic. 90, 112–119.Nori, J., Villalobos, F. & Loyola, R. (2018). Global priority areas for amphibian research. J. Biogeogr. 45, 2588–2594.Oksanen, J., Blanchet, F.G., Friendly, M., Kindt, R., Legendre, P., Mcglinn, D., Minchin, P.R., Hara, R.B.O., Simpson, G.L., Solymos, P., Stevens, M.H.H. & Szoecs, E. (2020). Community Ecology Package. 1–299.Oliveira, B.F., São‐Pedro, V.A., Santos‐Barrera, G., Penone, C. & Costa, G.C. (2017). AmphiBIO, a global database for amphibian ecological traits. Sci. Data 4, 1–7.Osen, K., Soazafy, M.R., Martin, D.A., Wurz, A., März, A., Ranarijaona, H.L.T. & Hölscher, D. (2021). Land‐use history determines stand structure and tree diversity in vanilla agroforests of northeastern Madagascar. Appl. Veg. Sci. 24, 1–14.Packard, M.J., Packard, G.C. & Boardman, T.J. (1982). Structure of eggshells and water relations of reptilian eggs. Herpetologica 38, 136–155.Palacios, C.P., Agüero, B. & Simonetti, J.A. (2013). Agroforestry systems as habitat for herpetofauna: Is there supporting evidence? Agrofor. Syst. 87, 517–523.Palumbi, S., Romano, S., Mcmillan, W.O. & Grabowski, G. (1991). the Simple Fool ’ S Guide To PCR. October 96822:1–45.Perfecto, I. & Vandermeer, J. (2010). The agroecological matrix as alternative to the land‐sparing/agriculture intensification model. Proc. Natl Acad. Sci. USA 107, 5786–5791.Powers, J.S., Corre, M.D., Twine, T.E. & Veldkamp, E. (2011). Geographic bias of field observations of soil carbon stocks with tropical land‐Use changes precludes spatial extrapolation. Proc. Natl Acad. Sci. USA 108, 6318–6322.Powers, R.P. & Jetz, W. (2019). Global habitat loss and extinction risk of terrestrial vertebrates under future land‐use‐change scenarios. Nat. Clim. Chang. 9, 323–329.R Core Team (2019). R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing.Rakotoarison, A., Scherz, M.D., Glaw, F., Köhler, J., Andreone, F., Franzen, M., Glos, J., Hawlitschek, O., Jono, T., Mori, A., Ndriantsoa, S.H., Raminosoa, N.R., Riemann, J.C., Rödel, M.O., Rosa, G.M., Vieites, D.R., Crottini, A. & Vences, M. (2017). Describing the smaller majority: Integrative taxonomy reveals twenty‐six new species of tiny microhylid frogs (genus Stumpffia) from Madagascar. Vertebr. Zool. 67, 271–398.Ratsoavina, F.M., Raselimanana, A.P., Scherz, M.D., Rakotoarison, A., Razafindraibe, J.H., Glaw, F. & Vences, M. (2019). Finaritra! A splendid new leaf‐tailed gecko (Uroplatus) species from Marojejy National Park in north‐eastern Madagascar. Zootaxa 4545, 563–577.Rembold, K., Mangopo, H., Tjitrosoedirdjo, S.S. & Kreft, H. (2017). Plant diversity, forest dependency, and alien plant invasions in tropical agricultural landscapes. Biol. Conserv. 213, 234–242.Ricciardi, V., Mehrabi, Z., Wittman, H., James, D. & Ramankutty, N. (2021). Higher yields and more biodiversity on smaller farms. Nat. Sustain. 4, 651–657.Roswell, M., Dushoff, J. & Winfree, R. (2021). A conceptual guide to measuring species diversity. Oikos 130, 321–338.Scales, B.R. & Marsden, S.J. (2008). Biodiversity in small‐scale tropical agroforests: a review of species richness and abundance shifts and the factors influencing them. Environ. Conserv. 35, 160–172.Scott, D.M., Brown, D., Mahood, S., Denton, B., Silburn, A. & Rakotondraparany, F. (2006). The impacts of forest clearance on lizard, small mammal and bird communities in the arid spiny forest, southern Madagascar. Biol. Conserv. 127, 72–87.Soorae, P.S., Frankham, G.J. & Mohamed, A.A. (2020). The first record of the asian common toad Duttaphrynus melanostictus schneider, 1799 in Abu Dhabi. United Arab Emirates. Bioinvasions Rec. 9, 434–443.Styger, E., Rakotondramasy, H.M., Pfeffer, M.J., Fernandes, E.C.M. & Bates, D.M. (2007). Influence of slash‐and‐burn farming practices on fallow succession and land degradation in the rainforest region of Madagascar. Agric. Ecosyst. Environ. 119, 257–269.Thompson, M.E. & Donnelly, M.A. (2018). Effects of secondary forest succession on amphibians and reptiles: a review and meta‐analysis. Copeia 106, 10–19.Tilman, D., Balzer, C., Hill, J. & Befort, B.L. (2011). Global food demand and the sustainable intensification of agriculture. Proc. Natl Acad. Sci. USA 108, 20260–20264.Tscharntke, T., Clough, Y., Wanger, T.C., Jackson, L., Motzke, I., Perfecto, I., Vandermeer, J. & Whitbread, A. (2012). Global food security, biodiversity conservation and the future of agricultural intensification. Biol. Conserv. 151, 53–59.Uetz, P., Freed, P., Aguilar, R. & Hošek, J. (eds.) (2021) The Reptile Database, http://www.reptile‐database.org. Accessed August 25, 2021.Urbina‐Cardona, J.N., Olivares‐Pérez, M. & Reynoso, V.H. (2006). Herpetofauna diversity and microenvironment correlates across a pasture‐edge‐interior ecotone in tropical rainforest fragments in the Los Tuxtlas Biosphere Reserve of Veracruz, Mexico. Biol. Conserv. 132, 61–75.Vallan, D. (2000). Influence of forest fragmentation on amphibian diversity in the nature reserve of Ambohitantely, highland Madagascar. Biol. Conserv. 96, 31–43.Vences, M., Thomas, M., Bonett, R.M. & Vieites, D.R. (2005). Deciphering amphibian diversity through DNA barcoding: chances and challenges. Philos. Trans. R. Soc. B Biol. Sci. 360, 1859–1868.Vieilledent, G., Grinand, C., Rakotomalala, F.A., Ranaivosoa, R., Rakotoarijaona, J.R., Allnutt, T.F. & Achard, F. (2018). Combining global tree cover loss data with historical national forest cover maps to look at six decades of deforestation and forest fragmentation in Madagascar. Biol. Conserv. 222, 189–197.Vieites, D.R., Wollenberg, K.C., Andreone, F., Köhler, J., Glaw, F. & Vences, M. (2009). Vast underestimation of Madagascar’s biodiversity evidenced by an integrative amphibian inventory. Proc. Natl Acad. Sci. USA 106, 8267–8272.Wanger, T.C., Saro, A., Iskandar, D.T., Brook, B.W., Sodhi, N.S., Clough, Y. & Tscharntke, T. (2009). Conservation value of cacao agroforestry for amphibians and reptiles in South‐East Asia: combining correlative models with follow‐up field experiments. J. Appl. Ecol. 46, 823–832.Watling, J.I. & Braga, L. (2015). Desiccation resistance explains amphibian distributions in a fragmented tropical forest landscape. Landsc. Ecol. 30, 1449–1459.Whitehurst, A.S., Sexton, J.O. & Dollar, L. (2009). Land cover change in western Madagascar’s dry deciduous forests: a comparison of forest changes in and around Kirindy Mite National Park. Oryx 43, 275.Wickham, H. (2016). Programming with ggplot2. 241–253. https://doi.org/10.1007/978‐3‐319‐24277‐4_12Williams, J.J. & Newbold, T. (2020). Local climatic changes affect biodiversity responses to land use: a review. Divers. Distrib. 26, 76–92. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Animal Conservation Wiley

Differential responses of amphibians and reptiles to land‐use change in the biodiversity hotspot of north‐eastern Madagascar

Loading next page...
 
/lp/wiley/differential-responses-of-amphibians-and-reptiles-to-land-use-change-RuyHWoH6s0

References (105)

Publisher
Wiley
Copyright
Copyright © 2022 The Zoological Society of London
ISSN
1367-9430
eISSN
1469-1795
DOI
10.1111/acv.12760
Publisher site
See Article on Publisher Site

Abstract

IntroductionDemand for agricultural goods is on the rise due to increasing per‐capita consumption, changing diets, and a growing world population (Tilman et al. 2011). This situation is leading to both an expansion of croplands into natural areas and an intensification within existing production systems (Tscharntke et al. 2012). Most agricultural expansion in the tropics happens at the expense of forests and leads to an increase in forest fragmentation (Hansen et al. 2020). This expansion affects all aspects of biodiversity: research has documented losses in species richness (Scales & Marsden, 2008), shifts in species composition (Newbold et al. 2016), reductions of functional diversity (Matuoka et al. 2020), and erosion of phylogenetic diversity (Li et al. 2020). Importantly, the response of biodiversity to land‐use change differs among species groups (Newbold et al. 2014) and world regions (Williams & Newbold, 2020), but certain groups and regions are understudied compared with others. While vertebrates are comparatively well studied, we know less about the responses of amphibians and reptiles than we know about mammals and birds (Newbold et al. 2014). Similarly, we know less about what happens in the more biodiverse tropics compared with temperate regions (Gardner et al. 2009). In sum, we know that land‐use change is the main driver of biodiversity decline globally (Powers & Jetz, 2019), but how the diversity of certain understudied taxa differs between land‐use types remains an open question, especially in the tropics, where high land‐use pressure and biodiversity coincide (Newbold et al. 2020).Tropical agricultural landscapes contribute to food security but also provide opportunities for nature conservation (Perfecto & Vandermeer, 2010). The quality of the agro‐ecosystem for biodiversity conservation depends on the farming systems. Agricultural landscapes dominated by large‐scale industrial monocultures have lower conservation value than diverse mosaics of forest fragments, agroforestry systems, and more intensively farmed annual crop fields (Mendenhall et al. 2016; Murray & Nowakowski, 2021). For example, Ricciardi et al. (2021) reported that small‐scale agricultural landscapes host more biodiversity than large‐scale ones and De Palma et al. (2015) showed that low‐intensity agriculture can maintain relatively diverse bee communities. Furthermore, forest patches within tropical small‐scale agricultural landscapes increase the variety of ecological niches and benefit species with small ranges in particular (Gray et al. 2016). Besides the value for biodiversity, small‐scale land‐use mosaics can also provide essential ecosystem services and livelihoods for rural people, making such landscapes work for humans and nature (Kremen & Merenlender, 2018). Most research investigating the value of tropical agricultural landscapes for biodiversity and humans was conducted in the Neotropics (Mendenhall et al. 2016) while the conservation value of Afrotropical agricultural landscapes is less understood (Powers et al. 2011). Furthermore, it remains largely unclear how different forms of forest conversion – that is with the use of fire (slash‐and‐burn shifting cultivation) or without the use of fire (forest degradation and forest‐derived agroforestry) – affect amphibian and reptile diversity in resulting land‐use types.Here, we investigate how herpetofauna taxonomic diversity and species composition differ across prevalent burned and unburned land‐use types in Madagascar, providing insights into the vulnerability of amphibians and reptiles to various forms of land‐use change. Madagascar has lost 44% of forest cover since the 1950s, mainly due to the transformation of natural ecosystems to agricultural lands (Vieilledent et al. 2018). Along with unsustainable extraction rates (Whitehurst et al. 2009), this has resulted in more than half of evaluated Malagasy vertebrate species being at risk of extinction (IUCN, 2019). Outstanding levels of endemism (Goodman & Benstead, 2005) and ongoing threats qualify Madagascar as a global biodiversity hotspot (Myers et al. 2000). While the forests and protected areas of the island are increasingly well surveyed, the biodiversity in the agricultural landscapes is less known (Irwin et al. 2010). Focusing on amphibians and reptiles is particularly relevant since Madagascar has a diverse and highly endemic herpetofauna: amphibian species diversity is currently estimated at around 370 (AmphibiaWeb, 2020) and almost all species are endemic (Goodman & Benstead, 2005). Reptile diversity stands at around 440 species (Uetz et al. 2021) with 91% endemism (Goodman & Benstead, 2005). Due to cryptic taxonomic complexes, many species still await discovery or description, suggesting that total species richness numbers will increase further (Vieites et al. 2009). Globally, herpetofauna is sensitive to various anthropogenic threats (Vallan, 2000) including chytrid fungi, environmental pollution, collection for the pet trade, climate change, and conversion of forest habitat into agricultural lands (Hof et al. 2011). All of these factors are also threatening Malagasy herpetofauna, with deforestation being the top pressure (Cordier et al. 2021).Our study focuses on north‐eastern Madagascar. The area retains more forest cover than other parts of the country (Vieilledent et al. 2018) and is a global priority area for amphibian research (Nori et al. 2018). Besides being known for its remarkable biodiversity, north‐eastern Madagascar is also a global center for vanilla cultivation (Hänke et al. 2018). The price boom of the spice between 2012 and 2019 has triggered an expansion of vanilla agroforests (Llopis et al. 2019), and roughly 80% of rural households in the study region farm vanilla (Hänke et al. 2018). How the expansion of vanilla agroforestry is impacting biodiversity is not well known, despite that the majority of global vanilla cultivation takes place across three biodiversity hotspots (Madagascar, Indonesia, and MexicoMyers et al. 2000; FAO, 2020). Investigating how vanilla agroforestry compares with other land‐use types in terms of biodiversity conservation thus represents a research gap. Besides farming vanilla, the rural population in north‐eastern Madagascar also practices slash‐and‐burn shifting cultivation for hill rice production. Valleys and plains in the study region are commonly occupied by irrigated rice paddies, forming the backbone of staple crop supply. We thus compared the taxonomic diversity and community composition of amphibians and reptiles among plots in six land‐use types within the small‐scale agricultural landscape and compared this with continuous old‐growth forest inside Marojejy National Park. Within the small‐scale agricultural landscape, we sampled vanilla agroforests of contrasting land‐use history (forest‐ and fallow‐derived vanilla agroforests), as well as forest fragment, herbaceous fallow, woody fallow, and rice paddy. Following Cordier et al. (2021) and Gonzalez & González‐Trujillo (2021), we hypothesized that taxonomic diversity would be highest in old‐growth forest (for both amphibians and reptiles). We further expected losses in taxonomic diversity and shifts in species composition after old‐growth forest conversion, particularly in those cases where land‐use conversion involves the use of fire for slash‐and‐burn shifting cultivation. Based on previous research (Martin et al. 2020), we further expected that rehabilitation through agroforestry on fallow land would positively affect the taxonomic diversity of the herpetofauna.Materials and methodsStudy region and study designWe conducted our study in the SAVA region in north‐eastern Madagascar (Fig. 1a,b) where forests outside protected areas are now highly fragmented (Vieilledent et al. 2018) and the landscape is dominated by smallholder agriculture.1FigureOverview of the study region and the study design. (a) SAVA region in north‐eastern Madagascar; (b) study area within SAVA region; (c) study design showing the distribution of 80 plots in 10 villages and at two sites inside Marojejy National Park; and (d) overview of the studied land‐use types and the typical land‐use transformation trajectory from old‐growth forest to forest fragments and agricultural land‐uses. Panel (d) modified after Martin et al. (2021) and Panel (c) modified after Dröge et al. (2021). Rice paddy is not part of the main land‐use trajectory.We collected data on circular plots with a 25 m radius (largest possible plot size in typical smallholder parcels) at low to mid‐altitude (7–819 m above sea level; mean = 192 m ± 207 m) surrounding 10 villages and in Marojejy National Park, a UNESCO World Heritage Site. In each village, we selected seven plots: three vanilla agroforests (forest‐derived and/or fallow‐derived vanilla agroforest), one forest fragment, one herbaceous fallow, one woody fallow, and one rice paddy. In the 10 villages, we first selected 30 vanilla agroforests along a canopy cover gradient. After consultation with the agroforest owner and a visual confirmation on the plot, we found that 20 vanilla agroforests were fallow‐derived while 10 agroforests were forest‐derived. This approach allowed us to understand the response of amphibians and reptiles to the land‐use history of vanilla agroforests. Additionally, we chose 10 plots at two sites (five plots each) inside Marojejy National Park. All land‐use types were replicated 10 times, except fallow‐derived vanilla agroforest, which was replicated 20 times. In total, we surveyed 80 plots (Fig. 1c,d). The selection of 10 plot per land‐use types was a compromise between a large number of replicates and logistical constraints. The average minimum distance between one plot and the next closest plot in the same village/site was 719 m ± 438 m, while the smallest minimum distance between two plots was 260 m as a compromise between independence of plots and logistical constraints.Sampled land‐use typesWe selected 10 old‐growth forest plots at two sites with five plots each. One of the sites has experienced some selective logging in the past but is now well protected (touristic zone in Manantenina valley; site Tourist on Fig. 1), the other site suffers from ongoing occasional selective logging and trapping (Bangoabe; site East on Fig. 1), but we chose plots that did not show signs of recent disturbance. The old‐growth forest plots are a minimum of 300 m from the National Park boundary. In the study region, isolated forest fragments occur around villages and represent remnants of the continuous forest cover that existed in the region prior to the large‐scale deforestation that began in the early 20th century (Gade, 1996). The 10 forest fragments have not been burned in historic times and are all used for the extraction of timber and non‐timber forest products. Herbaceous fallows occur after slash‐and‐burn shifting hill rice cultivation (locally referred to as tavy; Styger et al. 2007) and are sometimes used for grazing cows. The herbaceous fallow plots in this study had last burned at the end of 2016, 1 year before the onset of data collection in 2017. Woody fallows represent successional stages of herbaceous fallows, containing shrubs and small trees. Woody fallows are also occasionally grazed. The woody fallows in our study had last burnt 4–16 years before the onset of data collection in 2017 according to the reports of land‐owners. The climbing vanilla orchid (Vanilla planifolia) is farmed in agroforestry systems with two distinct land‐use histories (following Martin et al. 2020): forest‐derived vanilla agroforests, where vanilla is directly planted inside the forest after removing understory trees and shrubs and managing tall trees for shade (not involving burning). In fallow‐derived vanilla agroforests, vanilla is planted on fallow land which resulted from slash‐and‐burn shifting hill rice cultivation in the past. In these fallow‐derived vanilla agroforests, farmers allow natural regeneration of trees or plant trees to provide shade or as support structures for the vanilla vines. Lastly, we studied irrigated rice paddies that occur in valley bottoms and flood plains. Rice is planted and harvested between one to three times per year. The rice paddies chosen for our study had wider‐than‐average banks to facilitate movement and sampling within the plots.Sampling, data collection, and field identificationTo collect data in the villages, we organized two sampling campaigns during the driest period of the year (October to December 2017 and late August to December 2018) and one campaign during the wettest period (Mid‐January to early April 2018). In Marojejy National Park, we also organized two sampling campaigns during the driest period (late August to early September 2018 and December 2018) and one during the wettest period (February 2019). During each campaign, we visited each plot once during the day (08:00–17:00) and once at night (18:30–23:00). Overall, we did six visits per plot, three times during the day and three times at night.We collected data on amphibian and reptile communities during time‐standardized searches (Kadlec et al. 2012). During the survey, we systematically searched each plot in a zig‐zag pattern (Kadlec et al. 2012). Each search was standardized to 45 min of searching time by two observers. In sum, we thus conducted 270 min of searching time on each plot, summing up to 408 h of searching time across all plots. To detect individuals hiding under rocks, in leaf axils, tree barks, tree holes, leaf‐litter, or deadwood, we actively inspected those microhabitats by lifting removable objects to check underneath.To identify individuals to species level during fieldwork, we used morphological characteristics and referred to keys and visualizations available in “A Field Guide to Amphibians and Reptiles of Madagascar” (Glaw & Vences, 2007) and additional literature (Rakotoarison et al. 2017; Ratsoavina et al. 2019). For those individuals who we could not identify with confidence in the field, we extracted tissue samples for follow‐up DNA analysis and/or collected the specimen. To denote those individuals, we used “cf.” (confer or compare with) and “sp. aff.” (affinis) in conjuncture with the genus name when we found slightly different morphological characters compared with the literature that we used or compared with a given identified species. We used sp. (sp1 or sp2) in conjuncture with the genus name when we could not find any given identified species to compare with the encountered individual. The sp. denotation was thus used to differentiate several unknown species belonging to the same genus and we consider them as morphospecies but used DNA barcodes to confirm presumed species status (see below). Following the field identification, we resumed the searching time; thus abundance and diversity of amphibians and reptiles were independent of the time spent on the plot. Throughout this manuscript, we refer to each encountered individual as an “encounter” rather than an individual as we cannot exclude the possibility of having encountered the same individual at more than one sampling event.Species identification with DNA barcodesWe collected muscle or toe clips as tissue samples of individuals in case of non‐reliable morphological‐characteristics‐based identification. We preserved tissue samples in Eppendorf tubes with 90% ethanol, stored, and analyzed at the Evolutionary Biology laboratory of Prof. Miguel Vences at the University of Braunschweig, Germany. We preserved amphibian specimens in 70% ethanol and deposited them in the collection of the Regional University Centre of the SAVA region (CURSA), Antalaha, Madagascar. We extracted genomic DNA following the standard single‐tube salt extraction protocol (Bruford et al. 1992). To do so, we cut small pieces from the collected tissue and proceeded to protein digestion following the protocol of Cacciali et al. (2019). We then passed the extracted DNA to a PCR (Polymerase Chain Reaction) thermocycler, in which we amplified DNA fragments of two mitochondrial genes, “16S” (Vences et al. 2005) and “COI” with the primers 16SAL (5´‐CGCCTGTTTATCAAAAACAT‐3´) and 16SBH (5´‐CCGGTCTGAACTCAGATCACGT‐3´) from Palumbi et al. (1991). Amplification was set for 45 s at 94–96°C as denaturation, for 45 s at 52°C as hybridization, and for 90 s at 72°C as elongation. The whole process was repeated for 35 cycles. We generated DNA sequences through a sequencer. We first cleaned the obtained sequences with the software CodonCode Aligner version 8.0.2 (Codon Code Corporation, Centerville, MA, USA). Then, we processed newly generated sequences for additional quality control with a Blast search (Altschul et al. 1990) which queries for similar DNA sequences in the GenBank database to provide information on closest taxa or potential contamination. Finally, we built a phylogenetic tree using MEGA‐X (Kumar et al. 2018) to reveal clusters of each newly generated sequence, allowing us to reliably cluster individuals to known or presumed species. We submitted all newly determined sequences to GenBank under accession numbers MZ019097–MZ019430. In this study, the DNA barcoding differentiated between 43 species of amphibians and 24 species of reptiles. Among them were 12 candidates for new species of which eight were documented for the first time (marked as sp. CaNEW) and four were already found by other researchers (marked as sp. Ca) but have not yet been described. The new species were based on a comparison of obtained DNA sequences with available sequences with the same extraction DNA method conducted in the same lab. In case the DNA barcoding method did not yield results due to short sequences or contaminations, we kept the field identification for those species (for 15 species, eight for amphibians, and seven for reptiles).We attributed the red list status to evaluated species following IUCN (2019). We determined endemicity according to AmphibiaWeb (2020) for amphibians and Uetz et al. (2021) for reptiles. Because most amphibian and reptile genera are completely endemic to Madagascar, we could attribute endemicity also to morphospecies.Data analysis and visualization: species diversities and encountersDuring the data analysis, we conducted all statistical analyses using R version 3.5.1 (R Core Team, 2019) and the R package “ggplot2” (Wickham, 2016) for visualization. To compare observed species richness and species encounters at the plot‐level among land‐use types, we computed a generalized linear model with observed species richness as a response, land‐use type as the explanatory variable, and village (respectively old‐growth forest site) as a random factor and with Poisson family for count data as distribution. We then ran the glht function of the R‐package “multcomp” (Hothorn et al. 2008) applying a Tukey all‐pair‐comparison with Bonferroni correction. We then used the same approach to compare observed species richness between plots from unburned land‐use types (old‐growth forest, forest fragment, forest‐derived vanilla) and plots from burned land‐use types (herbaceous fallow, woody fallow, fallow‐derived vanilla).We used the encounter data to compute sample size‐based extrapolation curves (Hsieh et al. 2016) with the “iNEXT” package to assess the diversity (species richness, Shannon, and Simpson diversity) across land‐use types using the Hill number framework (Chao et al. 2014). We calculated species richness, Shannon diversity, and Simpson diversity (Hsieh et al. 2016). Chao & Jost (2012) define q = 0 (0D) as species richness, that is the effective number of species in the community, giving equal weight to frequent and infrequent species; q = 1 (1D) as Shannon diversity, giving more weight to more frequently observed species and q = 2 (2D), as Simpson diversity interpreted as the effective number of abundant species. This approach is now well established given key advantages over traditional Shannon and Simpson diversity indices (Roswell et al. 2021).To display the total species diversity in each land‐use type, we sub‐sampled 10 plots of the 20 fallow‐derived vanilla agroforests. To do so, we randomly selected one fallow‐derived vanilla agroforest from each village. As one of the villages lacks fallow‐derived vanilla agroforests (village Andramanolotra, see Fig. 1), this resulted in nine agroforests. We proceeded to select one additional plot from all remaining fallow‐derived vanilla agroforests, enabling a fair comparison of total species diversity across 10 plots of each land‐use type.To compare the evenness of communities between land‐use types, we additionally plotted the observed and extrapolated diversity across hill numbers for each land‐use type, allowing for a direct comparison of slopes between land‐use types (Chao & Jost, 2015; Roswell et al. 2021).Data analysis and visualization: species compositionTo evaluate the differences in species community composition between land‐use types between burned and unburned plots, and between villages/sites, we used the metaMDS function of the R‐package “vegan” (with 1000 permutations, Oksanen et al. 2020). We used non‐metric dimensional scaling (NMDS) of Bray‐Curtis dissimilarities to visualize the dissimilarity of species composition in two dimensions. Furthermore, to test the differences between land‐use types, we used PERMANOVA (Permutational multivariate analysis of variance) implemented in the adonis function of the “vegan” package (Oksanen et al. 2020) and computed pairwise differences using the pairwise.adonis function with Bonferroni correction of the “pairwise.adonis” package (Martinez, 2020). Since PERMANOVA may confound location and dispersion, we used PERMDISP (Permutational analysis of multivariate dispersions) to see if PERMANOVA results may have been influenced by dispersion differences between groups (Anderson & Walsh, 2013). To do so, we used the function betadisper and permutest and performed pairwise comparisons using the TukeyHSD function. To analyze the degree of forest dependency (Rembold et al. 2017), we plotted the proportion of encounters for each species across seven land‐use types (10 replicates per land‐use type). To visualize the forest dependency, we used the barplot function of R graphics.ResultsEncounters and observed species richnessIn total, we made 6215 encounters and observed 119 species of amphibians and reptiles. The 3694 amphibian encounters belong to 58 species, 15 genera, and four families (see Table S1). The most species‐rich genera of amphibians are Boophis (11 species), Stumpffia (10 species), and Gephyromantis (eight species). We found all but one species (Ptychadena mascareniensis) to be endemic according to AmphibiaWeb (2020). Among the observed amphibian species, 22 (37%) could not be identified to species level in the field. Twelve of them were recognized as candidates for new species based on genetic barcoding, another eight of them were counted as morphospecies based on the criteria mentioned in the methods section, and two were conferred to previously identified species. Among the encountered amphibian species, seven are listed in the “threatened” category (IUCN 2019). Among the threatened species, we recorded six vulnerable and one endangered species.The 2521 reptile encounters represent 61 species, 28 genera, and five families (see Table S1). The most species‐rich reptile genera were Phelsuma (11 species) and Uroplatus (six species). We found 83% of reptile species to be endemic (Uetz et al. 2021). Amongst observed reptile species, 15 could not be identified to species level in the field; of these, seven were counted as morphospecies and eight were conferred to previously identified species. Based on the IUCN red list (IUCN 2019), seven encountered reptile species are listed as “threatened.” Among the threatened species, we recorded four endangered, one critically endangered, and two vulnerable species.Plot‐level observed amphibian species richness differed significantly among land‐use types (F6,73 = 19.59, P < 0.001) and when comparing all burned to all unburned land‐use types (F1,67 = 15.89, P < 0.001; Figure S6 and Table S7). A Tukey post‐hoc test revealed significant pairwise differences between pairs of land‐use types (Fig. 2a and Table S2): old‐growth forest had significantly higher observed amphibian species richness while rice paddy had significantly lower observed species richness compared with other land‐use types (Fig. 2a). The other land‐use types had similarly observed amphibian species richness with no significant differences (Table S2). Plot‐level amphibian encounters also differed significantly between land‐use types (F6,72 = 13.01, P < 0.001; Figure S9, Tables S10 and S12).2FigurePlot‐level observed amphibian (a) and reptile (b) species richness across seven land‐use types (replicated 10 times each except fallow‐derived vanilla, which is replicated 20 times) in north‐eastern Madagascar. Each dot is the observed species richness in a plot. The black horizontal line in the box shows the median. Land‐use types with letters in common did not differ significantly based on pairwise comparisons that controlled for inflated false‐positive errors using the Tukey HSD approach (Numeric results in SI: amphibians & reptiles Table S2). See icon reference in supplementary information (Free Icons Library).Plot‐level observed reptile species richness also varied significantly among land‐use types (F6,73 = 18.55, P < 0.001) and when comparing all burned to all unburned land‐use types (F1,67 = 25.60, P < 0.001; Figure S6 and Table S8). Tukey post‐hoc tests revealed no differences in observed species richness between old‐growth forest and forest fragment (P = 0.88) and between old‐growth forest and forest‐derived vanilla agroforest (P = 0.77). These three land‐use types are recorded with the highest observed species richness, with 10 reptile species on average per plot. Rice paddy had the lowest observed reptile species richness, but there were no significant differences (P = 0.19) compared with herbaceous fallow. Observed reptile species richness was significantly higher in forest‐derived vanilla agroforest than in fallow‐derived vanilla agroforest (P = 0.01; Fig. 2b and Table S2). Plot‐level reptile encounters also differed significantly between land‐use types (F6,73 = 14, P < 0.001; Figure S9, Tables S11 and S12).Species accumulation curves, diversity, evenness, and estimated species richnessEncounter‐based accumulation curves for amphibians revealed the highest species richness, Shannon diversity, and Simpson diversity in old‐growth forest and the lowest in rice paddy. In most land‐use types, observed amphibian species richness curves flattened off, except for old‐growth forest and forest fragment. The overlap of the 95% confidence interval of extrapolated amphibian richness for old‐growth forest and forest fragment indicates no differences in species richness. Regarding the accumulated amphibian species diversity for Shannon and Simpson diversity, old‐growth forest varied significantly compared with all other land‐use types. Shannon and Simpson diversity also did not differ significantly between forest fragment and forest‐derived vanilla agroforest, respectively, compared with burned land uses, except to rice paddy (Fig. 3 and Table 1).3FigureEncounter‐based accumulation curves showing interpolated (solid line) and extrapolated (dotted line) diversities for amphibians (first and second columns of panel from left to right) and reptiles (third and fourth panels) in north‐eastern Madagascar. Unburned (first and third panels) and burned (second and fourth panels) land‐use types are separated. The species richness represented by q = 0 (first row of panel from above), Shannon diversity, q = 1 (second row of panel) and Simpson diversity q = 2 (third row of panel) with 95% confidence intervals (shaded areas) for the amphibian and the reptile data of seven land‐use types. The solid dots, triangles, and diamonds represent the reference samples, that is the observed number of encounters and species richness. See icon reference in supplementary information (Free Icons Library).1TableEstimated and observed amphibian and reptile species richness (q = 0) for all land‐use types and separated per land‐use type showing the observed and extrapolated total species richnessSpecies Group (measure)Land‐use typeAll 80 plotsOld‐growth forestForest fragmentForest‐derived vanilla agroforestFallow‐derived vanilla agroforestWoody fallowHerbaceous fallowRice paddyAmphibian (observed)5832261416864Amphibian (extrapolated)NA60.0 (20.4–99.6)46.1 (12.4–79.8)18.0 (9.1–26.9)25.0 (7.0–42. 9)9.0 (7.4–10.6)6.5 (3.9–9.1)5.0 (4.3–5.6)Reptile (observed)613430301819811Reptile (extrapolated)NA42.0 (20.4–63.5)40.1 (10.2–69.9)35.1 (18.2–52.4)26.9 (8.2–45.7)46.6 (11.5–81.7)8.0 (6.8–9.2)18.9 (4.3–33.4)Total (observed)11966564434271415Each land‐use type is represented by 10 plots; for fallow‐derived vanilla, the 10 plots are down‐sampled from 20 plots. Extrapolated species richness is based on 5000 encounters and includes the lower and upper 95% confidence interval in italics in the brackets. See SI for results of q = 1 and q = 2 (amphibians and reptiles: Table S3).Encounter‐based accumulation curves for reptiles revealed the highest species richness and Shannon diversity in old‐growth forest and the lowest in herbaceous fallow. Observed species richness curves flattened off in all land‐use types except woody fallow and fallow‐derived vanilla agroforest. We also found an overlap of the 95% confidence intervals of extrapolated species richness as well as Shannon and Simpson diversity within unburned and burned land‐use types except for Shannon and Simpson diversity in burned land‐use types (Fig. 3 and Table 1).The species diversity dropped more strongly across hill number orders (q = 0 to q = 2) in amphibians than in reptiles after old‐growth forest transformation (Fig. 3, Table S3 and Figure S5), highlighting that amphibian communities were more uneven than reptile communities (Figure S5). Furthermore, the comparison shows that old‐growth forest communities of both amphibians and reptiles were most uneven (Figure S5).Species composition and forest dependencyThe composition of amphibian communities differed significantly across land‐use types (PERMANOVA: R2 = 0.50, P < 0.001, Df = 6), partly driven by differences in group dispersion (PERMDISP: F = 4.400, P = 0.004, Df = 6, Tables S18 and S19). Pairwise comparisons showed that amphibian communities in old‐growth forest and rice paddy were significantly different from those of other land‐use types. Forest fragment differed significantly from fallow‐derived vanilla agroforest, woody fallow, and herbaceous fallow, but no significant differences were observed between forest fragment and forest‐derived vanilla agroforest. We found no significant differences between forest‐derived vanilla agroforest, fallow‐derived vanilla agroforest, woody fallow, and herbaceous fallow for amphibian communities (Fig. 4a, see Table S4). We compared also the amphibian communities between unburned and burned land‐use types and found significant differences (PERMANOVA: R2 = 0.193, P < 0.001, Df = 1, Figure S13 and Table S14), again partly drive by dispersion (PERMDISP: F = 34.718, P = 0.001, Df = 1, Table S16). Similarly, amphibian communities differed between villages and Marojejy National Park communities (PERMANOVA: R2 = 0.416, P < 0.001, Df = 10, Figure S21, Tables S22 and S24), in this case not driven by differences in group dispersion (PERMDISP: F = 1.254, P = 0.283, Df = 11, Table S25). Here, communities were very similar between villages (Figure S21).4FigureSpecies composition across seven land‐use types in north‐eastern Madagascar. Non‐metric dimensional scaling (NMDS) showing community dissimilarity of amphibian (a) and reptile (b) communities. Old‐growth forest shows unique species composition for both amphibians and reptiles. See icon reference in supplementary information (Free Icons Library).Reptile species composition showed significant differences among land‐use types (PERMANOVA: R2 = 0.40, P < 0.001, Df = 6) not driven by differences in group dispersion (PERMDISP: F = 1.870, P = 0.088, Df = 6, Table S20) indicating the homogeneity of group dispersion. Pairwise comparisons showed that the reptile communities in old‐growth forest and forest fragment were significantly different from all other land‐use types. The communities in other land‐use types showed no significant differences (Fig. 4b and Table S4). Old‐growth forest showed unique communities of reptiles and amphibians. The reptile species composition between unburned and burned land‐use types showed significant differences (PERMANOVA: R2 = 0.210, P = .001, Df = 1, Table S15) not driven by differences in group dispersion (PERMDISP: F = 3.918, P = 0.057, Df = 1, Table S17). The reptile communities between villages and Marojejy National Park showed significant differences (PERMANOVA: R2 = 0.389, P < 0.001, Df = 10, Table S23 and S24) not driven by differences in group dispersion (PERMDISP: F = 0.835, P = 0.626, Df = 11, Table S26).Finally, the analyses of forest dependency revealed that 22 amphibian species (38%) were found exclusively in old‐growth forest, 11 amphibian species (19%) exclusively in forest fragment, three amphibian species (5%) exclusively in forest‐derived vanilla agroforest, and four amphibian species (7%) exclusively in fallow‐derived vanilla agroforest (Fig. 5a).5FigureProportion of species encounters across seven land‐use types (with 10 replicates each) in north‐eastern Madagascar. All 58 amphibian species with 3694 encounters (a) and all 61 reptile species (b) with 2521 encounters are displayed by forest dependency rank: 38% of amphibian species and 26% of reptile species exclusively occurred in old‐growth forest, despite that old‐growth forest is only accounting for 12.5% of the total plots. See icon reference in supplementary information (Free Icons Library).We observed 16 reptile species (26%) exclusively in old‐growth forest, three reptile species (5%) exclusively in forest fragment, four reptile species (7%) exclusively in forest‐derived vanilla agroforest, and one reptile species (2%) exclusively in fallow‐derived vanilla agroforest (Fig. 5b). No species were exclusively found in woody fallow, herbaceous fallow, and rice paddy for both groups.DiscussionIn this study, we provide a comprehensive assessment of the response of amphibian and reptile species diversity to land‐use change in the biodiversity hotspot of north‐eastern Madagascar. Overall, we document a highly diverse herpetofauna with 58 amphibian and 61 reptile species and presumed species, and a high proportion of endemic species (98% for amphibians and 83% for reptiles).At a plot‐level, observed species richness for both amphibians and reptiles was very high, with up to 12 and 14 species, respectively. For both groups, old‐growth forest was significantly different from all other land‐use types in terms of total estimated species richness and community composition. Rice paddy and herbaceous fallow harbored the lowest observed species richness for amphibians and reptiles, respectively. Reptile species diversity varied significantly between forest‐derived and fallow‐derived vanilla agroforests whereas amphibian species richness showed no differences. We found no unique species in woody fallow, herbaceous fallow, or rice paddy plots for either taxon. Importantly, we found that amphibians and reptiles responded differently to land‐use history. After any kind of old‐growth forest conversion, whether through burning or not, amphibian species communities both declined in richness and shifted in structure away from old‐growth forest. Deforestation for shifting cultivation through slash‐and‐burn led to species losses in reptiles as well, but changes were less pronounced if old‐growth forest was transformed to forest fragment or forest‐derived vanilla agroforest since these transformations refrain from using fire.Outstanding diversity of amphibians and reptiles in north‐eastern MadagascarThe diversity of amphibians and reptiles documented here was exceptional, both within Madagascar and compared with other tropical biodiversity hotspots. The few available studies on the response of amphibian communities to land‐use change in Madagascar reported lower overall values, with 32 and 28 species, respectively (Vallan, 2000; Ndriantsoa et al. 2017). Andreone et al. (2000) found 42 amphibian and 23 reptile species in north‐eastern Malagasy rainforest. Despite differences in methodology and study sites, our study provides additional evidence that the north‐east of Madagascar is one of the most species‐rich regions for herpetofauna in the country, as previously suggested (Brown et al. 2016). The documented diversity also exceeds values found in other tropical biodiversity hotspots (Cordier et al. 2021; Murray & Nowakowski, 2021) – for example, Urbina‐Cardona et al. (2006) found 21 amphibian and 33 reptile species in the region of Veracruz, Mexico, and Kurz et al. (2014) found 25 amphibian and 20 reptile species in a region of north‐eastern Costa Rica (versus 58 amphibian and 61 reptile species in this study). The high species diversity may partly be driven by our extensive sampling effort, in terms of number of plots (total n = 80), diversity of different land‐use types (seven), and search effort (total of 270 min searching time by two observers per plot), and is likely also influenced by the use of genetic samples for species identification.Response of amphibian diversity to land‐use changeWe found a strong negative response of amphibian species richness to any anthropogenic land use. Old‐growth forest had a significantly higher observed amphibian species richness with 38% species exclusively occurring there. The amphibian species composition is unique compared with all other land‐use types, driven by a high share belonging to the Microhylidae family and the Spinomantis, Guibemantis, and Mantella genera (Fig. 5a). Only forest fragments harbored comparable accumulated amphibian species richness. Subsequently, we infer that Malagasy amphibians are very sensitive to habitat change. This result is in line with findings from Ndriantsoa et al. (2017) in disturbed habitats of eastern Madagascar which documented strongly impoverished frog communities in secondary vegetation and rice fields compared with forests. Forest fragments are thus valuable for Malagasy amphibians to maintain diversity within the agricultural landscape but cannot substitute large and continuous old‐growth forest (Vallan, 2000).The other land‐use types (fallows that form part of the slash‐and‐burn cycle, vanilla agroforests, and rice paddies) are of minor importance for amphibian conservation, given the low species diversity that consists mainly of common and widely ranging species. Nonetheless, amphibians could play an important functional role in these habitats: the abundance of amphibians is high throughout, reflected by the high number of encounters, particularly in rice paddies (see Fig. 3b and Figure S9). As such, they may be an important food source for other taxa or could provide pest control services in crops (Hocking et al. 2014).The strong negative response of amphibians to deforestation could be explained by the fact that numerous species rely on moist environments to avoid dehydration (Watling & Braga, 2015), especially species living in evergreen forest habitat (Hof et al. 2011). Given that selective logging, forest fragmentation, and deforestation severely change microclimate conditions (Ewers & Banks‐Leite, 2013), many species may struggle to cope. Furthermore, amphibians have different habitat requirements throughout their life cycle stages (Hof et al. 2011; Hocking et al. 2014). Such heterogeneity is often lost in concert with the simplification in forest structure (Wanger et al. 2009), making new habitats unsuitable for many species.Response of reptile diversity to land‐use changeWe found a strong effect of land‐use history on reptile diversity. Unburned land‐use types had a significantly higher average species richness, more uneven communities, more unique species, and a distinct species composition compared with burned land‐use types and rice paddy. The distinct species composition in unburned land‐use types was caused by an increased number of species belonging to the Chamaeleonidae family and the Uroplatus genus (Fig. 5b) compared with other land‐use types. This is, to our knowledge, the first study shedding light on the response of reptiles to land‐use change in Madagascar’s humid eastern escarpment. Studies from the drier southern region (Scott et al. 2006; Gardner et al. 2016; Nopper et al. 2018) show that Malagasy reptile species react less strongly to habitat change than other taxa, especially if diurnal (Nopper et al. 2018).Interestingly, the less pronounced effects of land‐use change on reptiles compared with amphibians in tropical landscapes has also been demonstrated by others (Wanger et al. 2009; Kurz et al. 2014) and was found in a review by Palacios et al. (2013). This may be due to the high thermotolerance in reptiles (Thompson & Donnelly, 2018), thanks to a fairly impermeable skin covered with scales. This structure prevents excessive evaporative water loss and adapts reptiles to various microclimates (Watling & Braga, 2015). Furthermore, egg development is not heavily impacted by temperature rise, which reduces the size of hatchlings but accelerates incubation (Packard et al. 1982). Reptile traits associated with more open, hotter, and drier environments (Morin, 2005) may make them less vulnerable to land‐use change.Land‐use history of vanilla agroforests affects reptiles but not amphibiansIn our study, we have separated forest‐derived vanilla from fallow‐derived vanilla agroforests, thereby explicitly accounting for land‐use history (Martin et al. 2020). For amphibians, we found no differences between the two agroforestry systems across metrics. Reptile communities, on the other hand, were significantly more diverse in forest‐derived vanilla agroforests on plot‐level, more species‐rich overall, more uneven, and compositionally different compared with fallow‐derived vanilla agroforest. Reptile communities observed in forest‐derived vanilla agroforests are thus more similar to old‐growth forest and forest fragment. The reptile communities recorded in fallow‐derived vanilla agroforests were comparable with fallow land. Both kinds of agroforest thus resembled the land‐use types they were derived from. However, we observed an increase in accumulated amphibian species richness from fallow land to fallow‐derived vanilla agroforest, which suggests rehabilitation through agroforestry (Osen et al. 2021). These findings are in line with the prediction from a recent study by Martin et al. (2020) that found that the land‐use history of agroforestry systems matters for biodiversity.We further hypothesize that the strong importance of forest‐derived vanilla agroforests for reptiles may be in part driven by leaf‐litter depth, which is known to positively influence reptile diversity and abundance (Urbina‐Cardona et al. 2006), and which is typically deeper in forest‐derived vanilla agroforests. According to the same study, these microenvironmental variables also showed positive effects on amphibians and thus cannot explain why there are no differences between the two kinds of agroforests for amphibians. Elsewhere, the loss of canopy cover had negative impacts on amphibian and reptile species richness (Scott et al. 2006). Since canopy cover differed significantly between forest‐derived vanilla agroforest (higher) and fallow‐derived vanilla agroforest (lower) in the same study site (Osen et al. 2021), this calls for further investigation combining habitat characteristics with species traits (Oliveira et al. 2017) to elucidate the drivers of change in both species groups.Conservation implicationsThe strong negative response of amphibians to old‐growth forest modification and the high old‐growth forest dependency of amphibians and reptiles calls for the effective protection of the last remaining old‐growth forests. Additionally, conserving forest fragments within the agricultural landscape will be important to many reptile and amphibian species that are absent from other land‐use types. This is particularly important given that numerous species are micro‐endemics (Brown et al. 2014), meaning that they only occur very locally (Vieites et al. 2009), possibly in only a handful of forest fragments. Management strategies are thus needed to safeguard the long‐term existence of forest fragments and to improve the connectivity between fragments. The protection of large continuous forests throughout the region is also important under a changing climate (Hof et al. 2011) and in light of emerging threats, such as the chytrid fungi (Kolby & Skerratt, 2015) and the recent spread of the invasive Asian common toad (Duttaphrynus melanostictus, Soorae et al. 2020). These conservation needs are underscored by the exceptional diversity of reptiles and amphibians in north‐eastern Madagascar as well as by the high proportion of endemic species.Our findings from north‐eastern Madagascar also have broader relevance to herpetofauna conservation in other tropical biodiversity hotspots. The differential response of amphibians and reptiles to land‐use change shows that conservationists should not treat herpetofauna as a homogenous group when devising conservation programs. Furthermore, the distinct differences in reptile communities between plots of contrasting land‐use history shows that past land‐use must also be considered, and suggests that the maintenance of remnant forest fragments and forest‐derived agroforests is important. Lastly, we highlight the importance of unburned compared with burned land‐use types in the agricultural matrix for the conservation of reptiles.AcknowledgmentsThe authors are grateful to all chef de fokontany, landowners, and Madagascar National Parks for granting them access to sites and information. They thank Prof. Miguel Vences and his laboratory staff for invaluable support with the DNA barcoding, Saskia Dröge for preparing panel C of Fig. 1, James Herrera for feedback, and Prof. Miguel Acevedo, Prof. Arne Mooers, Philippe Fernandez‐Fournier, and an anonymous reviewer for their constructive reviews and feedback. They also thank “CodonCode grants” for offering them a free license under the CodonCode Aligner License Grant Program. They collected data under research permits N100/17/MEEF/SG/DGF/DSAP/SCB. Re, N°163/17/MEEF/SG/DGF/DSAP/SCB. Re, N°18/18/MEEF/SG/DGF/DSAP/SCB. Re and N°254/18/MEEF/SG/DGF/DSAP/SCB. Re granted by the Ministry of Environment and Sustainable Development, Antananarivo. They transported DNA samples domestically under the transport permit N° 34/17‐MEEF/SG/DREEF/SAVA/SRF granted by the Regional Office of the Ministry of Environment and Sustainable Development, Sambava. They exported the samples from Madagascar under CITES permits N°323C‐EA05/MG18 and N°180C‐EA03/MG19 granted by the Ministry of Environment and Sustainable Development, Antananarivo, and imported the samples to Germany under CITES permits DE‐E‐03377/18, DE‐E‐03378/18, DE‐E‐02422/19, DE‐E‐02423/19, DE‐E‐02424/19, and DE‐E‐02425/19 granted by the German Federal Agency for Nature Conservation, Bonn. This study was financially supported by the Niedersächsisches Vorab of Volkswagen Foundation as part of the research project “Diversity Turn in Land Use Science” (Grant number 11‐76251‐99‐35/13 (ZN3119)) and by the German Academic Exchange Service (DAAD) within the “Partnerships for Supporting Biodiversity in Developing Countries” initiative (Project Nr. 57449386). Open Access funding enabled and organized by Projekt DEAL. WOA Institution: GEORG‐AUGUST‐UNIVERSITAET GOTTINGEN. Blended DEAL: Projekt DEALConflict of interestNo conflict of interest to declare.Author’s contributionsTRF, DAM, KO, AW, HK, AA, and FMR designed the study. TRF, EB, RB, and RR collected amphibian and reptile data under the lead of TRF. TRF and DAM analyzed the data. TRF and DAM wrote the first manuscript draft. All authors revised the manuscript.Data Availability StatementData are available from Zenodo: http://doi.org/10.5281/zenodo.4548955ReferencesAltschul, S.F., Gish, W., Miller, W., Myers, E.W. & Lipman, D.J. (1990). Basic local alignement search tool. J. Mol. Biol. 215, 403–410.AmphibiaWeb. (2020). List of amphibians of Madagascar. Retrieve from: AmphibiaWeb., org/cgi/amphib_query?rel‐isocc=like&orderbyaw=Order&where‐isocc=Madagascar. Accessed December 14, 2020.Anderson, M.J. & Walsh, D.C.I. (2013). PERMANOVA, ANOSIM, and the Mantel test in the face of heterogeneous dispersions: what null hypothesis are you testing? Ecol. Monogr. 83, 557–574.Andreone, F., Randrianirina, J.E., Jenkins, P.D. & Aprea, G. (2000). Species diversity of Amphibia, Reptilia and Lipotyphla (Mammalia) at Ambolokopatrika, a rainforest between the Anjanaharibe‐Sud and Marojejy massifs, NE Madagascar. Biodivers. Conserv. 9, 1587–1622.Brown, J.L., Cameron, A., Yoder, A.D. & Vences, M. (2014). A necessarily complex model to explain the biogeography of the amphibians and reptiles of Madagascar. Nat. Commun. 5, 1–10.Brown, J.L., Sillero, N., Glaw, F., Bora, P., Vieites, D.R. & Vences, M. (2016). Spatial biodiversity patterns of Madagascar’s amphibians and reptiles. PLoS One 11, 1–26.Bruford, M.W., Hanotte, O., Brookfield, J.F.Y. & Burke, T. (1992). Single‐locus and multilocus DNA fingerprinting. In: Molecular genetic analysis of populations: A practical approach: 225–269. Hoelzel, A.R. (Ed.). Oxford: IRL Press.Cacciali, P., Buongermini, E. & Köhler, G. (2019). Barcoding analysis of Paraguayan squamata. Diversity 11, 152.Chao, A., Gotelli, N.J., Hsieh, T.C., Sander, E.L., Ma, K.H., Colwell, R.K. & Ellison, A.M. (2014). Rarefaction and extrapolation with Hill numbers: a framework for sampling and estimation in species diversity studies. Ecol. Monogr. 84, 45–67.Chao, A. & Jost, L. (2012). Coverage‐based rarefaction and extrapolation: standardizing samples by completeness rather than size. Ecology 93, 2533–2547.Chao, A. & Jost, L. (2015). Estimating diversity and entropy profiles via discovery rates of new species. Methods Ecol. Evol. 6, 873–882.Cordier, J.M., Aguilar, R., Lescano, J.N., Leynaud, G.C., Bonino, A., Miloch, D., Loyola, R. & Nori, J. (2021). A global assessment of amphibian and reptile responses to land‐use changes. Biol. Conserv. 253, 108863.De Palma, A., Kuhlmann, M., Roberts, S.P.M., Potts, S.G., Börger, L., Hudson, L.N., Lysenko, I., Newbold, T. & Purvis, A. (2015). Ecological traits affect the sensitivity of bees to land‐use pressures in European agricultural landscapes. J. Appl. Ecol. 52, 1567–1577.Dröge, S., Martin, D.A., Andriafanomezantsoa, R., Burivalova, Z., Fulgence, T.R., Osen, K., Rakotomalala, E., Schwab, D., Wurz, A., Richter, T. & Kreft, H. (2021). Listening to a changing landscape: acoustic indices reflect bird species richness and plot‐scale vegetation structure across different land‐use types in north‐eastern Madagascar. Ecol. Indic. 120, 106929.Ewers, R.M. & Banks‐Leite, C. (2013). Fragmentation impairs the microclimate buffering effect of tropical forests. PLoS One 8, e58093.FAO. (2020). FAOSTAT. Food and Agriculture Organisation of the United Nations. http://www.FAO.org/faostat/en/#home.Gade, D.W. (1996). Deforestation and its effects in Highland Madagascar. Mt. Res. Dev. 16, 101–116.Gardner, C.J., Jasper, L.D., Eonintsoa, C., Duchene, J.J. & Davies, Z.G. (2016). The impact of natural resource use on bird and reptile communities within multiple‐use protected areas: evidence from sub‐arid Southern Madagascar. Biodivers. Conserv. 25, 1773–1793.Gardner, T.A., Barlow, J., Chazdon, R., Ewers, R.M., Harvey, C.A., Peres, C.A. & Sodhi, N.S. (2009). Prospects for tropical forest biodiversity in a human‐modified world. Ecol. Lett. 12, 561–582.Glaw, F. & Vences, M. (2007). In Ny Toro‐hay momba ny Amphibia sy ny Reptilia an’i Madagasikara: 3rd edn: 529. Vences, M. & Glaw, V. (Eds.). Cologne: Vences and Glaw Verlag.González, T.M., González‐Trujillo, J.D., Muñoz, A. & Armenteras, D. (2021). Differential effects of fire on the occupancy of small mammals in neotropical savanna‐gallery forests. Perspect. Ecol. Conserv. 19, 179–188.Goodman, S.M. & Benstead, J.P. (2005). Updated estimates of biotic diversity and endemism for Madagascar. Oryx 39, 73–77.Gray, C.L., Hill, S.L.L., Newbold, T., Hudson, L.N., Börger, L., Contu, S., Hoskins, A.J., Ferrier, S., Purvis, A. & Scharlemann, J.P.W. (2016). Local biodiversity is higher inside than outside terrestrial protected areas worldwide. Nat. Commun. 7, 1–7.Hänke, H., Barkmann, J., Blum, L., Franke, Y., Martin, D.A., Niens, J., et al. (2018). Socio‐ economic, land use and value chain perspectives on vanilla farming in the SAVA Region (north‐ eastern Madagascar): The Diversity Turn Baseline Study (DTBS). Discussion Paper, 1806. https://doi.org/10.13140/RG.2.2.22059.80163Hansen, M.C., Wang, L., Song, X.P., Tyukavina, A., Turubanova, S., Potapov, P.V. & Stehman, S.V. (2020). The fate of tropical forest fragments. Sci. Adv. 6, 1–10.Hocking, D.J., Babbitt, K.J. & Hocking, D.J. (2014). Amphibian contributions to ecosystem services. Herpetol. Conserv. Biol. 9, 1–17.Hof, C., Araújo, M.B., Jetz, W. & Rahbek, C. (2011). Additive threats from pathogens, climate and land‐use change for global amphibian diversity. Nature 480, 516–519.Hothorn, T., Bretz, F. & Westfall, P. (2008). Simultaneous inference in general parametric models. Biom. J. 50, 346–363.Hsieh, T.C., Ma, K.H. & Chao, A. (2016). iNEXT: an R package for rarefaction and extrapolation of species diversity (Hill numbers). Methods Ecol. Evol. 7, 1451–1456.Irwin, M.T., Wright, P.C., Birkinshaw, C., Fisher, B.L., Gardner, C.J., Glos, J., Goodman, S.M., Loiselle, P., Rabeson, P., Raharison, J.‐L., Raherilalao, M.J., Rakotondravony, D., Raselimanana, A., Ratsimbazafy, J., Sparks, J.S., Wilmé, L. & Ganzhorn, J.U. (2010). Patterns of species change in anthropogenically disturbed forests of Madagascar. Biol. Conserv. 143, 2351–2362.IUCN. (2019). The IUCN. red list of threatened species. Version 2019‐1. IUCN., Gland, Switzerland. Available from http://www.IUCN.redlist.org. Accessed March 2019.Kadlec, T., Tropek, R. & Konvicka, M. (2012). Timed surveys and transect walks as comparable methods for monitoring butterflies in small plots. J. Insect. Conserv. 16, 275–280.Kolby, J.E. & Skerratt, L.F. (2015). Amphibian chytrid fungus in Madagascar neither shows widespread presence nor signs of certain establishment. PLoS One 10, 1–6.Kremen, C. & Merenlender, A.M. (2018). Landscapes that work for biodiversity and people. Science 362. https://doi.org/10.1126/science.aau6020Kumar, S., Stecher, G., Li, M., Knyaz, C. & Tamura, K. (2018). MEGA X: molecular evolutionary genetics analysis across computing platforms. Mol. Biol. Evol. 35, 1547–1549.Kurz, D.J., Nowakowski, A.J., Tingley, M.W., Donnelly, M.A. & Wilcove, D.S. (2014). Forest‐land use complementarity modifies community structure of a tropical herpetofauna. Biol. Conserv. 170, 246–255.Li, D., Olden, J.D., Lockwood, J.L., Record, S., McKinney, M.L. & Baiser, B. (2020). Changes in taxonomic and phylogenetic diversity in the Anthropocene: changes in biodiversity. Proc. R. Soc. B Biol. Sci. 287, 20200777.Llopis, J.C., Harimalala, P.C., Bär, R., Heinimann, A., Rabemananjara, Z.H. & Zaehringer, J.G. (2019). Effects of protected area establishment and cash crop price dynamics on land use transitions 1990–2017 in north‐eastern Madagascar. J. Land Use Sci. 14, 52–80.Martin, D.A., Andriafanomezantsoa, R., Dröge, S., Osen, K., Rakotomalala, E., Wurz, A., Andrianarimisa, A. & Kreft, H. (2021). Bird diversity and endemism along a land‐use gradient in Madagascar: the conservation value of vanilla agroforests. Biotropica 53, 179–190.Martin, D.A., Osen, K., Grass, I., Hölscher, D., Tscharntke, T., Wurz, A. & Kreft, H. (2020). Land‐use history determines ecosystem services and conservation value in tropical agroforestry. Conserv. Lett. 13, 1–12.Martinez, A.P. pairwise.Adonis: Pairwise Multilevel Comparison Using Adonis. R Package Version 0.0.1. Available online: https://github.com/pmartinezarbizu/pairwiseAdonis. Accessed on 6 December 2020.Matuoka, M.A., Benchimol, M., de Almeida‐Rocha, J.M. & Morante‐Filho, J.C. (2020). Effects of anthropogenic disturbances on bird functional diversity: a global meta‐analysis. Ecol. Indic 116, 106471.Mendenhall, C.D., Shields‐Estrada, A., Krishnaswami, A.J. & Daily, G.C. (2016). Quantifying and sustaining biodiversity in tropical agricultural landscapes. Proc. Natl Acad. Sci. USA 113, 14544–14551.Morin, K.C. (2005). Herpetofaunal responses to prescibed fire in upland pine communities of Northeast Florida: 77 pp. PhD Thesis. University of Florida.Murray, A.H., Nowakowski, A.J. & Frishkoff, L.O. (2021). Climate and land‐use change severity alter trait‐based responses to habitat conversion. Global Ecol. Biogeogr. 30, 598–610.Myers, N., Mittermeler, R.A., Mittermeler, C.G., Da Fonseca, G.A.B. & Kent, J. (2000). Biodiversity hotspots for conservation priorities. Nature 403, 853–858.Ndriantsoa, S.H., Riemann, J.C., Raminosoa, N., Rödel, M.O. & Glos, J.S. (2017). Amphibian diversity in the matrix of a fragmented landscape around Ranomafana in Madagascar depends on matrix quality. Trop. Conserv. Sci. 10, 194008291668606.Newbold, T., Hudson, L.N., Hill, S.L.L., Contu, S., Gray, C.L., Scharlemann, J.P.W., Börger, L., Phillips, H.R.P., Sheil, D., Lysenko, I. & Purvis, A. (2016). Global patterns of terrestrial assemblage turnover within and among land uses. Ecography 39, 1151–1163.Newbold, T., Hudson, L.N., Phillips, H.R.P., Hill, S.L.L., Contu, S., Lysenko, I., Blandon, A., Butchart, S.H.M., Booth, H.L., Day, J., De Palma, A., Harrison, M.L.K., Kirkpatrick, L., Pynegar, E., Robinson, A., Simpson, J., Mace, G.M., Scharlemann, J.P.W. & Purvis, A. (2014). A global model of the response of tropical and sub‐tropical forest biodiversity to anthropogenic pressures. Proc. R. Soc. B Biol. Sci. 281, 20141371.Newbold, T., Oppenheimer, P., Etard, A. & Williams, J.J. (2020). Tropical and Mediterranean biodiversity is disproportionately sensitive to land‐use and climate change. Nat. Ecol. Evol. 4, 1630–1638.Nopper, J., Riemann, J.C., Brinkmann, K., Rödel, M.O. & Ganzhorn, J.U. (2018). Differences in land cover – biodiversity relationships complicate the assignment of conservation values in human‐used landscapes. Ecol. Indic. 90, 112–119.Nori, J., Villalobos, F. & Loyola, R. (2018). Global priority areas for amphibian research. J. Biogeogr. 45, 2588–2594.Oksanen, J., Blanchet, F.G., Friendly, M., Kindt, R., Legendre, P., Mcglinn, D., Minchin, P.R., Hara, R.B.O., Simpson, G.L., Solymos, P., Stevens, M.H.H. & Szoecs, E. (2020). Community Ecology Package. 1–299.Oliveira, B.F., São‐Pedro, V.A., Santos‐Barrera, G., Penone, C. & Costa, G.C. (2017). AmphiBIO, a global database for amphibian ecological traits. Sci. Data 4, 1–7.Osen, K., Soazafy, M.R., Martin, D.A., Wurz, A., März, A., Ranarijaona, H.L.T. & Hölscher, D. (2021). Land‐use history determines stand structure and tree diversity in vanilla agroforests of northeastern Madagascar. Appl. Veg. Sci. 24, 1–14.Packard, M.J., Packard, G.C. & Boardman, T.J. (1982). Structure of eggshells and water relations of reptilian eggs. Herpetologica 38, 136–155.Palacios, C.P., Agüero, B. & Simonetti, J.A. (2013). Agroforestry systems as habitat for herpetofauna: Is there supporting evidence? Agrofor. Syst. 87, 517–523.Palumbi, S., Romano, S., Mcmillan, W.O. & Grabowski, G. (1991). the Simple Fool ’ S Guide To PCR. October 96822:1–45.Perfecto, I. & Vandermeer, J. (2010). The agroecological matrix as alternative to the land‐sparing/agriculture intensification model. Proc. Natl Acad. Sci. USA 107, 5786–5791.Powers, J.S., Corre, M.D., Twine, T.E. & Veldkamp, E. (2011). Geographic bias of field observations of soil carbon stocks with tropical land‐Use changes precludes spatial extrapolation. Proc. Natl Acad. Sci. USA 108, 6318–6322.Powers, R.P. & Jetz, W. (2019). Global habitat loss and extinction risk of terrestrial vertebrates under future land‐use‐change scenarios. Nat. Clim. Chang. 9, 323–329.R Core Team (2019). R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing.Rakotoarison, A., Scherz, M.D., Glaw, F., Köhler, J., Andreone, F., Franzen, M., Glos, J., Hawlitschek, O., Jono, T., Mori, A., Ndriantsoa, S.H., Raminosoa, N.R., Riemann, J.C., Rödel, M.O., Rosa, G.M., Vieites, D.R., Crottini, A. & Vences, M. (2017). Describing the smaller majority: Integrative taxonomy reveals twenty‐six new species of tiny microhylid frogs (genus Stumpffia) from Madagascar. Vertebr. Zool. 67, 271–398.Ratsoavina, F.M., Raselimanana, A.P., Scherz, M.D., Rakotoarison, A., Razafindraibe, J.H., Glaw, F. & Vences, M. (2019). Finaritra! A splendid new leaf‐tailed gecko (Uroplatus) species from Marojejy National Park in north‐eastern Madagascar. Zootaxa 4545, 563–577.Rembold, K., Mangopo, H., Tjitrosoedirdjo, S.S. & Kreft, H. (2017). Plant diversity, forest dependency, and alien plant invasions in tropical agricultural landscapes. Biol. Conserv. 213, 234–242.Ricciardi, V., Mehrabi, Z., Wittman, H., James, D. & Ramankutty, N. (2021). Higher yields and more biodiversity on smaller farms. Nat. Sustain. 4, 651–657.Roswell, M., Dushoff, J. & Winfree, R. (2021). A conceptual guide to measuring species diversity. Oikos 130, 321–338.Scales, B.R. & Marsden, S.J. (2008). Biodiversity in small‐scale tropical agroforests: a review of species richness and abundance shifts and the factors influencing them. Environ. Conserv. 35, 160–172.Scott, D.M., Brown, D., Mahood, S., Denton, B., Silburn, A. & Rakotondraparany, F. (2006). The impacts of forest clearance on lizard, small mammal and bird communities in the arid spiny forest, southern Madagascar. Biol. Conserv. 127, 72–87.Soorae, P.S., Frankham, G.J. & Mohamed, A.A. (2020). The first record of the asian common toad Duttaphrynus melanostictus schneider, 1799 in Abu Dhabi. United Arab Emirates. Bioinvasions Rec. 9, 434–443.Styger, E., Rakotondramasy, H.M., Pfeffer, M.J., Fernandes, E.C.M. & Bates, D.M. (2007). Influence of slash‐and‐burn farming practices on fallow succession and land degradation in the rainforest region of Madagascar. Agric. Ecosyst. Environ. 119, 257–269.Thompson, M.E. & Donnelly, M.A. (2018). Effects of secondary forest succession on amphibians and reptiles: a review and meta‐analysis. Copeia 106, 10–19.Tilman, D., Balzer, C., Hill, J. & Befort, B.L. (2011). Global food demand and the sustainable intensification of agriculture. Proc. Natl Acad. Sci. USA 108, 20260–20264.Tscharntke, T., Clough, Y., Wanger, T.C., Jackson, L., Motzke, I., Perfecto, I., Vandermeer, J. & Whitbread, A. (2012). Global food security, biodiversity conservation and the future of agricultural intensification. Biol. Conserv. 151, 53–59.Uetz, P., Freed, P., Aguilar, R. & Hošek, J. (eds.) (2021) The Reptile Database, http://www.reptile‐database.org. Accessed August 25, 2021.Urbina‐Cardona, J.N., Olivares‐Pérez, M. & Reynoso, V.H. (2006). Herpetofauna diversity and microenvironment correlates across a pasture‐edge‐interior ecotone in tropical rainforest fragments in the Los Tuxtlas Biosphere Reserve of Veracruz, Mexico. Biol. Conserv. 132, 61–75.Vallan, D. (2000). Influence of forest fragmentation on amphibian diversity in the nature reserve of Ambohitantely, highland Madagascar. Biol. Conserv. 96, 31–43.Vences, M., Thomas, M., Bonett, R.M. & Vieites, D.R. (2005). Deciphering amphibian diversity through DNA barcoding: chances and challenges. Philos. Trans. R. Soc. B Biol. Sci. 360, 1859–1868.Vieilledent, G., Grinand, C., Rakotomalala, F.A., Ranaivosoa, R., Rakotoarijaona, J.R., Allnutt, T.F. & Achard, F. (2018). Combining global tree cover loss data with historical national forest cover maps to look at six decades of deforestation and forest fragmentation in Madagascar. Biol. Conserv. 222, 189–197.Vieites, D.R., Wollenberg, K.C., Andreone, F., Köhler, J., Glaw, F. & Vences, M. (2009). Vast underestimation of Madagascar’s biodiversity evidenced by an integrative amphibian inventory. Proc. Natl Acad. Sci. USA 106, 8267–8272.Wanger, T.C., Saro, A., Iskandar, D.T., Brook, B.W., Sodhi, N.S., Clough, Y. & Tscharntke, T. (2009). Conservation value of cacao agroforestry for amphibians and reptiles in South‐East Asia: combining correlative models with follow‐up field experiments. J. Appl. Ecol. 46, 823–832.Watling, J.I. & Braga, L. (2015). Desiccation resistance explains amphibian distributions in a fragmented tropical forest landscape. Landsc. Ecol. 30, 1449–1459.Whitehurst, A.S., Sexton, J.O. & Dollar, L. (2009). Land cover change in western Madagascar’s dry deciduous forests: a comparison of forest changes in and around Kirindy Mite National Park. Oryx 43, 275.Wickham, H. (2016). Programming with ggplot2. 241–253. https://doi.org/10.1007/978‐3‐319‐24277‐4_12Williams, J.J. & Newbold, T. (2020). Local climatic changes affect biodiversity responses to land use: a review. Divers. Distrib. 26, 76–92.

Journal

Animal ConservationWiley

Published: Aug 1, 2022

Keywords: agroecology; agroforestry; community ecology; herpetofauna; human‐dominated landscape; land‐use history; slash‐and‐burn shifting cultivation; forest dependency

There are no references for this article.