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

Learn More →

The effect of rainfall upon the behaviour and use of under-road culverts in four amphibian species

The effect of rainfall upon the behaviour and use of under-road culverts in four amphibian species Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzz001/5481437 by guest on 17 August 2022 BioscienceHorizons Volume 11 2018 10.1093/biohorizons/hzz001 ............................................................................................ ..................................................................... Research article The effect of rainfall upon the behaviour and use of under-road culverts in four amphibian species Timothy Gleeson , Silviu Petrovan and Anna Muir University of Chester, Parkgate Road, Chester CH1 4BJ, UK *Corresponding author: Timothy Gleeson, 22 Isleham Close, L19 4XS Liverpool, UK. Tel: +44 07505267958. Email: timothy.j.gleeson@gmail.com Supervisor: Dr Anna Muir, Department of Biological Sciences, University of Chester, Parkgate Road, Chester CH1 4BJ, UK. Email: a.muir@chester.ac.uk ............................................................................................ ..................................................................... Habitat fragmentation and road mortalities are major contributors towards declines in amphibian populations. This has seen the introduction of culverts, passages that run under roads and provide safe passage for amphibians. Research investigating the effects of rainfall upon amphibian culvert use is limited. This study, conducted at Frankfield Loch in Glasgow, assesses how time elapsed since rainfall influences migration behaviour and the use of culverts across four different species; common toads (Bufo bufo), common frogs (Rana temporaria and newts, a group composed of smooth newts (Lissotriton vulgaris) and palmate newts (Lissotriton helveticus). Analysis of images taken by a custom made, time lapse camera found that significantly fewer common toads (r = 0.148, n = 468, p = 0.001) and common frogs (r = −0.175, n = 106, p = 0.037) used the culvert as time since rainfall increased. This may have been caused by the culvert not maintaining wet enough conditions for amphi- bians. The study also found that more newts (r = 0.272, n = 92, p = 0.004) and common toads (r = 0.531, n = 19, p = 0.010) were using the culvert to move away from Frankfield Loch as time since rainfall increased. An increase in juvenile newts was also observed as time since rainfall increased (r = 0.214, n = 92, p = 0.020). This may have been caused by a decrease in baro- metric pressure, which follows a decrease in rainfall, acting as a cue for migration and juvenile dispersal. The study recom- mends careful consideration of the design of each culvert, incorporating species-specific preferences and the requirements of juveniles. The study also suggests that where possible the culvert should be designed to hold water for longer. Key words: Bufo bufo, Rana temporaria, culverts, amphibian, migration behaviour, rainfall Submitted on 23 May 2018 ............................................................................................ ..................................................................... individuals generally follow the same migratory routes Introduction (Gamble, McGarigal and Compton, 2007; Santos et al., 2007). As a result, intercepting these migration routes pre- Roads have been recognized as a major factor in the global sents amphibians with a barrier that prevents them from population decline of amphibians (Glista, DeVault and accessing breeding sites and reproducing (Santos et al. 2007). DeWoody, 2008; Petrovan and Schmidt, 2016). Amphibians are strongly impacted by road mortalities and roads can fur- Road fragmentation can isolate populations. This reduces ther lead to the fragmentation of amphibian habitat and can gene flow between populations and can lead to a reduction in intersect their migratory routes (Glista, DeVault and genetic variability. This in turn can increase the population’s DeWoody, 2008). Semlitsch (2002) describes the seasonal vulnerability to extinction (Keller and Largiadèr, 2003; cycles of amphibian migration; overwintering in terrestrial Cushman, 2006). Dispersal may also be reduced in fragmen- areas before moving to aquatic breeding sites in the spring. ted landscapes meaning that immigrants cannot bolster strug- Breeding site fidelity is often maintained across the years and gling populations or recolonize and re-establish populations ............................................................................................... .................................................................. ©The Author(s) 2018. Published by Oxford University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzz001/5481437 by guest on 17 August 2022 Research article Bioscience Horizons � Volume 11 2018 ............................................................................................... .................................................................. following a local extinction (Lehtinen, Galatowitsch and Tester, Network, 2013). Following a series of planning applications, 1999; Cushman, 2006). It is, therefore, vital that connectivity work was carried out to upgrade Loch Road (North between populations, breeding sites, overwintering sites and Lanarkshire Council, 2003). The plans were designed so that migration routes is maintained (Lehtinen, Galatowitsch and the road would act as less of a barrier to wildlife and incorpo- Tester, 1999; Semlitsch, 2002; Santos et al. 2007). rated the construction of the three culverts that pass under the road (North Lanarkshire Council, 2006). The culvert is an In response to this, ecological mitigation measures can be amphibian road tunnel, built by ACO Germany, which has a installed. The most promising measure for amphibians are gridded roof to allow rain and natural light into the culvert road tunnels or culverts (Beebee, 2013). These are passages, and is a widely used design type in Europe (Glista, DeVault varying from 0.3 to 2 m in diameter, that run under the road and DeWoody, 2009; White, Mayes and Petrovan, 2017). (Glista, DeVault and DeWoody, 2009). It has been shown that pipe culverts can reduce road mortality on a section of Method of observation motorway in France from almost 100% down to 23% in common toads, Bufo bufo (Lode, 2000). Similarly, Dodd, To observe the amphibians, a custom made, infra-red, time Barichivich and Smith (2004) found that in the Paynes Prairie lapse camera was developed and set up inside the culvert by State Preserve in Florida the presence of culverts reduced road Froglife, a UK-based charity dedicated to amphibian and rep- mortalities in the local wildlife, including southern toads, tile conservation (Froglife, 2017). Rather than use standard Bufo terrestris, and various other frog species, by 93.5%. motion detection, which has been shown to be largely inef- fective for slow moving and very small species such as amphi- Rainfall has been shown to alter migration behaviour in bians (Pagnucco, Paszkowski and Scrimgeour, 2011), these amphibians and lead to earlier breeding (Beebee et al. 2002; cameras record images every 10 s, 24 h/day during the moni- Todd and Winne, 2006; Todd et al. 2010). Equally, for some toring period. Each image captured by the camera included species of amphibians, migration can be triggered by a the time and date that it was taken. The images are later ana- decrease in barometric pressure caused by a decrease in rain- lysed using a custom made software script that applies set fall (Elewa, 2005; Osbourn, 2012). However, the effect of levels of pixel change detection between consecutive frames precipitation upon amphibian movements is frequently and is shown as a green outline highlighting the affected area, ignored (Todd et al. 2010) and very little information exists indicating movement (Figure 1). Images from the camera on how rainfall patterns affect amphibian usage of road miti- were collected between the 11 September and the 30 gation structures. Therefore, the aim of this study was to iden- September 2016. Froglife granted the ethical approval for the tify if and how rainfall affects the behaviour of amphibians study. using culverts. The study investigated how each of the follow- ing parameters correlated with time since rainfall: the total Previous research supports the use of camera traps as the number of amphibians that entered the culvert; the number of most effective method of observing amphibians inside a cul- amphibians that made a full crossing; the direction of vert, as unlike the alternative methods (sand beds, ink beds amphibian travel and the numbers of juveniles and adults that and pitfall traps), camera traps are not intrusive, can identify were observed in the culvert. amphibians to a species level and can provide data about time, date, behaviour and age (Veenbaas and Brandjes, 1999; Methods Hobbs and Brehme, 2017). Study site Located in Seven Lochs Wetland Park in Glasgow, UK, the culvert used for this investigation is one of three that runs underneath Loch Road. Loch Road runs north-east of Frankfield Loch and separates the loch from a large area of woodland, grassland and wetlands. Both the loch and the area of woodland, grassland and reed beds accommodate amphibian populations of common frogs (Rana temporaria), common toads (B. bufo), smooth newts (Lissotriton vulgaris) and palmate newts (Lissotriton helveticus) (S. Petrovan, per- sonal communication; glasgownaturalhistory.org.uk, 2016). As these species all exhibit seasonal migration between aquatic breeding sites (such as Frankfield Loch) and terrestrial habitat for overwintering and foraging (Semlitsch, 2002), then the recently constructed Loch Road is intersecting a key Figure 1. Image taken inside the culvert under Loch Road. section of the migratory route for these populations. Highlighted in green is a common frog. Recorded at the bottom of the Seven Lochs Wetland Park seek to protect the heritage of image is information about which camera took the image, the date and the time. the site and conserve local wildlife habitat (GCV Green ............................................................................................... .................................................................. 2 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzz001/5481437 by guest on 17 August 2022 Bioscience Horizons � Volume 11 2018 Research article ............................................................................................... .................................................................. Following automated analysis to select images with Common frogs were identified as having webbed feet, trav- detected movement, the images were analysed individually elling using a series of hops and having a pointed snout and manually using the programme FastStone (FastStone (Cooke, 2004; Halliday, 2012). Common toads were distin- soft, 2016). Upon identifying an amphibian, information guished by their toes being separated rather than webbed, about the image was recorded. This included the date and moving by walking and having a more rounded snout time of photo, the species of amphibian, the direction of tra- (Cooke, 2004; Halliday, 2012). Using the dorsal angle of vel, the age class of the amphibian, the behaviour being exhib- view of the camera, juvenile and female smooth newts were ited, the number of images the individual was recorded in and indistinguishable from palmate newts, and as no great crested the position of the amphibian in the culvert. The position that newts were recorded in this study, all observations of newts an individual was in the culvert was recorded as either left, were grouped into a single ‘newt’ category. centre left, centre, centre right or right and was determined by Rainfall data was retrieved from www.timeanddate.com, dividing the image up equally into vertical bands, from left to an organization that monitors and records weather data right. The position of an individual could change as it moved (Time And Date, 2017). These data were collected from the through the culvert and this would be recorded as the range weather station at Glasgow Airport, 11 miles away from of positions the individual moved across. The direction was Loch Road. These data were recorded twice every hour, at recorded as either moving ‘in’, towards the top of the image 30 min intervals. This meant that rainfall data for exact times and towards the loch, or as ‘out’, towards the bottom of the was not available. Due to this all the data from the images image and away from the loch. If an individual was observed were organized so that it was analysed per hour (Table 2). entering the image, turning around and leaving the way it The amount of time elapsed since last rainfall, measured in came it in, then this was recorded as either ‘in/out’, if it was hours, was then incorporated into the data. During rainfall, initially moving towards the top of the image, or if the oppos- the amount of time since last rainfall was recorded as 0. For ite was true, ‘out/in’. The age of each individual was subject- the hours where there was no rainfall, the amount of time ively determined by its size. Large individuals were recorded since the last known rainfall was recorded. The rainfall data as ‘adult’ and smaller individuals were recorded as ‘juvenile’. was recorded in 30 min intervals at either 20 min past the To classify the behaviour of an individual, an ethogram con- hour or 50 min past the hour. structed by Froglife was used (Table 1). An individual may exhibit multiple behaviours in the culvert, all of which would Statistical analyses were carried out using IBM SPSS ver- be recorded. sion 22.0 (IBM Corp., 2013). As the variable that was Table 1. Ethogram for behaviour of amphibians using the culvert Behaviour Description Moving Direct movement through the culvert Creeping A combination of small movements and hesitations Turns 180 degree change in direction heading either back into, or out of the tunnel Hesitant Stationary for 1 or 2 frames before moving on (10–20 s) Waiting Pausing for 3 or more frames (30 s or more), often seen to pivot on the spot Freezes Absolutely motionless—does not trigger motion detection until moving on again Exploring Strong lateral or stochastic movements across the tunnel, no clear heading Table 2. Small section of database arranged to an hourly basis, spanning hours 66–69 Hour Amphibian count Common frog count Newt count Common toad count Time elapsed since rainfall (h) 66 2 0 0 2 0.67 67 1 0 0 1 1.67 68 3 1 2 0 2.67 69 5 4 1 0 0 ............................................................................................... .................................................................. 3 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzz001/5481437 by guest on 17 August 2022 Research article Bioscience Horizons � Volume 11 2018 ............................................................................................... .................................................................. investigated sought to measure the strength of an association between two sets of non-parametric, scale data, a 1-tailed Spearman’s rank correlation test was used to analyse each hypothesis. The Spearman’s rank correlations conducted were 1-tailed tests as the hypotheses stated which direction the correlation would be. A Kolmogorov–Smirnov test for normality was conducted for each set of data and a histogram was produced. The following variables were tested for correl- ation with time since rainfall: the total number of amphibians that entered the culvert; the number of amphibians that moved through the culvert to make a full crossing; the direc- tion of amphibian travel and the numbers of juveniles and adults that were observed in the culvert. Each parameter was studied considering all amphibian species collectively as well as considering each species individually. Results After the automated selection from over 92 000 images, 7427 ‘positive’ images were analysed manually, and 431 amphi- Figure 2. Scatter plot, with line of best fit, showing a marginally bians were recorded by the camera across twenty days. The significant correlation between the number of ‘moving’ newts and the majority were newts which were recorded on 205 occasions. amount of time elapsed since rainfall (R = 0.014). There were many The next most common species were common frogs, totalling hours during rainfall, or shortly after, where no newts were observed 204 observations, whilst common toads were recorded on 26 ‘moving’, however as time since rainfall increased more newts were occasions. ‘Movement’ was the most commonly recorded ‘moving’ through the culvert each hour. Results of 1-tailed Spearman’s behaviour, being displayed in 404 instances and 338 amphi- rank correlation were r = 0.173, n = 92, p = 0.049. Circle thickness bians were moving out (away from the loch), more than any indicates frequency of data point. other direction. There were 295 juveniles and 135 adults recorded. On the 18th day of the investigation, the lens became obscured by condensation due to a high level of mist A highly significant correlation was found between the num- in the tunnel and until the camera was serviced no data could ber of amphibians moving in the ‘in’ direction, towards be collected from the affected images. As the study was con- Frankfield Loch and the amount of time since rainfall (r = ducted during a rainy period, large amounts of the data are −0.213, n = 157, p = 0.004) (Figure 3). Splitting the data by concentrated during rainfall or shortly after rainfall. species revealed that a highly significant correlation existed between common toads moving into the culvert and the amount of time since rainfall (r = −0.728, n = 19, p = Hypothesis 1—The number of amphibians that moved <0.001) (Figure 4). No significant correlations were found for through the culvert (in either direction) will increase as common frogs (r = −0.118, n = 106, p = 0.114) or for newts time since rainfall increases. (r = −0.125, n = 92, p = 0.118). There was no significant correlation between the number of Hypothesis 3—More amphibians will be moving away amphibians observed ‘moving’ through the culvert and the from the body of water (out) as time since rainfall time since the last rainfall (r = −0.53, n = 157, p = 0.225). increases. When splitting the data by species there was no significant correlation between time since rainfall and the number of ‘moving’ common frogs (r = −0.118, n = 106, p = 0.114) or There was no significant correlation between the number of common toads (r = −0.331, n = 19, p = 0.083). There was, amphibians moving in the ‘out’ direction, away from the however, a marginal significant correlation between the num- Loch, with the amount of time elapsed since the last rainfall (r ber of ‘moving’ newts and the time since rainfall (r = 0.173, n = 0.37, n = 157, p = 0.325). Splitting the data by species how- = 92, p = 0.049) (Figure 2). A large number of data points are ever found that highly significant correlations existed for concentrated during rainfall or shortly after rainfall but these common toads (r = 0.531, n = 19, p = 0.010) and newts (r = data predominantly show few or no newts were observed dur- 0.272, n = 92, p = 0.004) (Figure 5). No significant correl- ing these hours whilst higher numbers of newts were seen per ation was found for common frogs (r = −0.099, n = 106, p = hour as time since rainfall increased. 0.156). Hypothesis 2—Fewer amphibians will be moving towards Hypothesis 4—Fewer amphibians will be observed using the body of water (in) as time since rainfall increases. the culvert as time since rainfall increases. ............................................................................................... .................................................................. 4 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzz001/5481437 by guest on 17 August 2022 Bioscience Horizons � Volume 11 2018 Research article ............................................................................................... .................................................................. (Figure 6). Splitting the data according to species found that highly significant correlations existed for common frogs (r = −0.190, n = 468, p=<0.001) and common toads (r = 0.148, n = 468, p = 0.001) (Figure 7). Hypothesis 5—The number of juvenile amphibians using the culvert will decrease as time since rainfall increases. There was a significant correlation between number of juve- niles and the amount of time since the last rainfall (r = −0.106, n = 468, p = 0.011) (Figure 8). Dividing the data by species revealed that there was a significant correlation for newts (r = 0.214, n = 92, p = 0.020) (Figure 9) but not for common frogs (r = 0.016, n = 106, p = 0.436) and common toads (r = 0.281, n = 19, p = 0.122). Hypothesis 6—The number of adult amphibians using the culvert will decrease as time since rainfall increases. There was a highly significant correlation between the num- Figure 3. Scatter plot, with line of best fit, showing a significant ber of adults using the culvert each hour and the amount of correlation between the number of amphibians that moved into the time elapsed since rainfall (r = −0.157, n = 168, p =< 0.001) culvert, towards Frankfield Loch and the amount of time since rainfall (Figure 10). Splitting the data by species revealed that signifi- (R = 0.025). Results of 1-tailed Spearman’s rank correlation were as follows: r = −0.213, n = 157, p = 0.004. Circle thickness indicates cant correlations existed for both common frogs (r = −0.175, frequency of data point. n = 106, p = 0.037) and common toads (r = −0.445, n = 19, p = 0.028) (Figures 11). No significant correlation was found for newts (r = −0.087, n = 92, p = 0.204). Discussion Culvert usage A significant correlation was found when the number of amphibians using the culvert was compared with the amount of time elapsed since rainfall. As the time since rainfall increased, fewer common toads and common frogs were observed using the culvert. This correlation was not found to exist for newts. Glista, DeVault and DeWoody (2009) and Jackson (1996) report that many amphibians require wet con- ditions in culverts before they use them. The culvert under Loch Road is grated to allow in rain. However, the rainfall is unlikely to remain in the culvert for long as it was designed to maintain hydrological links between the loch and the sur- rounding environment (North Lanarkshire Council, 2006; GCV Green Network, 2013). Therefore, it is possible that the culvert not being wet enough could be a possible explanation for the negative correlation between common frogs and com- mon toads using the culvert and the time elapsed since the last Figure 4. Scatter plot, with line of best fit, showing how fewer common rainfall. toads moved into the culvert and towards Frankfield Loch in correlation with the amount of time elapsed since rainfall (R = 0.249). Results of Furthermore, it has been reported that amphibians prefer 1-tailed Spearman’s rank correlation were as follows: r = −0.728, n = 19, wet conditions for migration in general, which may also be a p=< 0.001. Circle thickness indicates frequency of data point. factor influencing the negative correlation (Jackson, 1996). To better understand the factors responsible for fewer com- Therewas ahighlysignificant correlation between the number mon frogs and common toads using the culvert, further study, of amphibians using the culvert with the amount of time elapsed which contrasts both the numbers of amphibians moving out- since the last rainfall (r = −0.158, n = 468, p =< 0.001) side the culvert and across the road with the numbers of ............................................................................................... .................................................................. 5 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzz001/5481437 by guest on 17 August 2022 Research article Bioscience Horizons � Volume 11 2018 ............................................................................................... .................................................................. Figure 5. Scatter plots, with line of best fit, showing correlation between the amount of time since rainfall and (left) the number of common toads moving out of the culvert away and from Frankfield Loch (R = 0.093) (r = 0.531, n = 19, p = 0.010) and (right) the number of newts moving out of the culvert (R = 0.028) (r = 0.272, n = 92, p = 0.004). Circle thickness indicates frequency of data point. maintain hydrological links) would help clarify the effect that the internal conditions of the culvert had on the results. No significant correlation was found to exist between the number of newts using the culvert and the time elapsed since rainfall. One of the limitations of the study was that the cam- era was unable to distinguish between smooth newts and pal- mate newts. The newt category therefore was a combination of both smooth and palmate newts and this may well have influenced the results. To better understand how time elapsed since rainfall influences smooth and palmate newt usage of culverts, further investigation will be required. An alternative method to the infra-red camera, potentially pitfall traps, would be needed so as to distinguish the two newt species but this has substantial bias implications in terms of influencing the behaviour of the newts. Another limitation of the study is that the rainfall was not recorded at the study site but at the weather station at Glasgow Airport which may result in slight disparities between the recorded rainfall and the actual rainfall. These differences, however, would be minimal and unlikely to Figure 6. Scatter plot, with line of best fit, showing that fewer amphibians used the culvert in correlation with the amount of time impact the results. It would be preferable if in future studies elapsed since the rainfall (R = 0.010). Results of 1-tailed Spearman’s rainfall could be recorded at the study site. rank correlation were as follows: r = −0.158, n = 468, p =< 0.000. Circle thickness indicates frequency of data point. The study took place over a 20-day period in September which is a relatively short timeframe. Although previous research has shown that amphibian migration away from amphibians using the culvert, in correlation with the time breeding sites does occur during this timeframe (Sinsch, 1988; elapsed since last rainfall, would be required. In this study, Miaud, Guyétant and Elmberg,1999; Matos et al. 2017), fur- only one culvert was used but it would be beneficial if future ther research would benefit from extending the duration of studies could involve more culverts. Being able to compare the study. different culvert designs (such as ones that are not designed to ............................................................................................... .................................................................. 6 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzz001/5481437 by guest on 17 August 2022 Bioscience Horizons � Volume 11 2018 Research article ............................................................................................... .................................................................. Figure 7. Scatter plots, with line of best fit, showing correlation between the amount of time elapsed since rainfall and (left) the number of common frogs that used the culvert (R = 0.026) (r = −0.190, n = 468, p=<0.001) and (right) between the number of common toads that used the culvert (R = 0.017) (r = 0.148, n = 468, p = 0.001). Circle thickness indicates frequency of data point. Figure 8. Scatter plot, with line of best fit, showing correlation Figure 9. Scatter plot, with line of best fit, showing correlation between the number of juveniles that used the culvert and the between the number of juvenile newts that used the culvert and the amount of time since rainfall (R = 0.003). Results of 1-tailed amount of time elapsed since the last rainfall (R = 0.018). Results of Spearman’s rank correlation were as follows: r = −0.106, n = 468, 1-tailed Spearman’s rank correlation were as follows: r = 0.214, n = 92, p = 0.011. Circle thickness indicates frequency of data point. p = 0.020. Circle thickness indicates frequency of data point. in the ‘out’ direction, away from Frankfield Loch. No correl- Direction of travel ation, in either direction of travel, was found to exist when As time since rainfall increased, significantly fewer common common frogs were compared with time since rainfall. The toads were travelling in the ‘in’ direction, towards Frankfield results would suggest that an increasing absence of rainfall Loch. Alongside this, as time since rainfall increased, signifi- encourages common toads and newts to move away from the cantly more common toads and newts were observed moving loch, in the ‘out’ direction. ............................................................................................... .................................................................. 7 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzz001/5481437 by guest on 17 August 2022 Research article Bioscience Horizons � Volume 11 2018 ............................................................................................... .................................................................. As common toads and newts migrate away from breeding research that states that common toads and other amphibians sites and towards terrestrial areas, the correlations that exist typically prefer wet conditions for migration (Sinsch, 1988; between direction of travel and time since rainfall may suggest Jackson, 1996; Glista, DeVault and DeWoody, 2009). that an increasing absence of rainfall encourages migration Significantly more juvenile newts and common frogs were towards terrestrial sites. This, however, contradicts previous observed in the culvert as time since rainfall increased whilst significantly fewer adult common toads were found using the culvert as time since rainfall increased. This, combined with more amphibians moving away from Frankfield Loch, would indicate that an increasing absence of rainfall may result in an increase in juvenile dispersal (Rothermel, 2004; Cushman, 2006; Matos et al. 2017). This too, however, would contra- dict what previous research has found as juveniles also favour wet conditions for dispersal (Rothermel, 2004; Cushman, 2006). The habitat surrounding Frankfield Loch is a wetland and the culverts that run under Loch Road are designed to estab- lish a hydrological link between the loch and the surrounding wetlands (North Lanarkshire Council, 2006; GCV Green Network, 2013). It may, therefore, be possible that the sur- rounding area remains wet enough to still favour dispersal, even after extended periods of no rainfall (Lesbarrères, Lodé and Merilä, 2004). As discussed by Elewa (2005), a decrease in barometric pressure, which is brought about by a decrease in rainfall, may encourage amphibians to initiate migration (Tao et al. 2011). Juvenile dispersal may also be prompted by a decrease Figure 10. Scatter plot, with line of best fit, showing correlation in barometric pressure (Osbourn, 2012). It may therefore be between the number of adults that used the culvert and the amount of possible that as the culvert maintained hydrological links time since rainfall (R = 0.017). Results of 1-tailed Spearman’s rank between the loch and the surrounding environment, the con- correlation were as follows: r = −0.157, n = 168, p =< 0.001. Circle ditions remained wet and juvenile newts instead used a thickness indicates frequency of data point. decrease in barometric pressure as a cue to begin migrating Figure 11. Scatter plots, with line of best fit, showing correlation between the amount of time elapsed since the last rainfall and (left) the number of adult common frogs that used the culvert (R = 0.028) (r = −0.175, n = 106, p = 0.037) and (right) between the number of adult common toads that used the culvert (R = 0.409) (r = −0.445, n = 19, p = 0.028). Circle thickness indicates frequency of data point. ............................................................................................... .................................................................. 8 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzz001/5481437 by guest on 17 August 2022 Bioscience Horizons � Volume 11 2018 Research article ............................................................................................... .................................................................. and dispersing. In order to better understand this, further barometric pressure that is associated with a decrease in rain- research is required that focuses on comparing barometric fall. The study recommends careful consideration of the pressure with dispersal away from breeding sites. design of each culvert, particularly with regards to juveniles, and that where possible the culvert should hold water for longer. This study provides insight into how rainfall affects Implications for culvert design and the behaviour and culvert use of common frogs, common amphibian conservation toads and newts. Our study found that a lack of rainfall saw more juveniles moving through the culvert. Schmidt and Zumbach (2008) Author’s biography suggest that imperfections in culvert design are more pro- nounced for juveniles and are likely to be less effective for Tim Gleeson studied Animal Behaviour at the University of them, compared to adults. This would imply that an increas- Chester and this research is part of his undergraduate disser- ing lack of rainfall may see more juveniles avoiding or failing tation. His fields of interest include conservation and animal to use mitigation measures. It is, therefore, recommended that welfare. culverts in locations which experience extensive periods of no rainfall during times of juvenile dispersal, should pay careful Silviu Petrovan was Conservation Coordinator at Froglife attention to the specific preferences in culvert design exhibited where he remains a trustee and is a Research Associate at by the target species and particularly juveniles. University of Cambridge. He set up the monitoring work, One of the factors that may have been responsible for con- contributed to designing the equipment and automated ana- tradictions between the results of this study and previous lysis and provided the camera trap images. studies is that the culvert was designed to provide a hydro- logical link between the loch and the surrounding habitat Anna Muir is a lecturer in Conservation Biology at the (North Lanarkshire Council, 2006; GCV Green Network, University of Chester and supervised the research. 2013). Although this may have maintained wet conditions in the surrounding habitat, it may have caused water not to remain in the culvert for long, leaving the culvert too dry for References amphibians. Amphibians require wet conditions to prevent desiccation of their skin (Lesbarrères, Lodé and Merilä, 2004) Beebee, T. J. (2013) Effects of road mortality and mitigation measures and may, therefore, have avoided using the culvert if it was on amphibian populations, Conservation Biology, 27 (4), 657–668. not wet enough. Therefore, the ideal design would hold water Beebee, T. J., Blaustein, A. R., Root, T. L. et al. (2002) Amphibian phen- in the culvert for longer, as is also reported by Glista, DeVault ology and climate change, Conservation Biology, 16 (6), 1454. and DeWoody (2009) but without being flooded (Schmidt and Zumbach, 2008). This may be difficult to achieve in prac- Cooke, F. (2004) The Encyclopedia of Animals: A Complete Visual Guide, tice as water levels are likely to vary substantially at most Univ of California Press, Oakland, California. sites. The design feature with the slotted top surface as in the Cushman, S. A. (2006) Effects of habitat loss and fragmentation on studied tunnel is likely to have substantial advantages by amphibians: a review and prospectus, Biological Conservation, 128 allowing free temperature and humidity exchange between (2), 231–240. the tunnel and the outside environment but there are poten- tially problematic inflows of heavily polluted water from the Dodd, C. K., Barichivich, W. J. and Smith, L. L. (2004) Effectiveness of a bar- road surface (White, Mayes and Petrovan, 2017). This design rier wall and culverts in reducing wildlife mortality on a heavily trav- feature, however, should be considered in the development of eled highway in Florida, Biological Conservation, 118 (5), 619–631. other culverts to maximize their effectiveness but might Elewa A. M. (2005) Migration of Organisms, Springer, New York City, NY. require additional measures, such as annual cleaning in early spring to remove road salt and other pollutants that might FastStone MaxView. (2016). FastStone. accumulate over winter. Froglife. What we do. Accessed at: froglife.org (February 2017). Gamble, L. R., McGarigal, K. and Compton, B. W. (2007) Fidelity and dis- persal in the pond-breeding amphibian, Ambystoma opacum: Conclusion implications for spatio-temporal population dynamics and conser- vation, Biological Conservation, 139 (3), 247–257. This study found that as time since rainfall increased, fewer common toads and common frogs would use the culvert Glasgow & Clyde Valley (GCV). (2013). Green network, Seven Lochs Wetland under Loch Road. The results of the investigation also suggest Park vision and masterplan. Accessed at: issuu.com (Feb 2017). that more newts and common toads would move away from Frankfield Loch and more juvenile newts would use the cul- glasgownaturalhistory.org.uk., (2016). Frankfeild Loch area species list vert, as time since rainfall increased. This contradicts previous [online]. Accessed at https://www.glasgownaturalhistory.org.uk/ research and may have been a result of a decrease in biodiversity/Frankfield_splist.pdf (6 November 2017). ............................................................................................... .................................................................. 9 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzz001/5481437 by guest on 17 August 2022 Research article Bioscience Horizons � Volume 11 2018 ............................................................................................... .................................................................. Glista, D. J., DeVault, T. L. and DeWoody, J. A. (2008) Vertebrate road Pagnucco, K. S., Paszkowski, C. A. and Scrimgeour, G. J. (2011) Using mortality predominantly impacts amphibians, Herpetological cameras to monitor tunnel use by long-toed salamanders Conservation and Biology, 3 (1), 77–87. (Ambystoma macrodactylum): an informative, cost-efficient tech- nique, Herpetological Conservation and Biology, 6 (2), 277–286. Glista, D. J., DeVault, T. L. and DeWoody, J. A. (2009) A review of mitiga- tion measures for reducing wildlife mortality on roadways, Petrovan, S. O. and Schmidt, B. R. (2016) Volunteer conservation action Landscape and Urban Planning, 91 (1), 1–7. data reveals large-scale and long-term negative population trends of a widespread amphibian, the common toad (Bufo bufo), PLoS Halliday, T., (2012). In C. Uhlenbroek, eds, Illustrated Encyclopaedia of One, 11 (10), e0161943. Animal Life, Dorling Kindersley, London, United Kingdom, pp. 202–249. Rothermel, B. B. (2004) Migratory success of juveniles: a potential con- straint on connectivity for pond‐breeding amphibians, Ecological Hobbs, M. T. and Brehme, C. S. (2017) An improved camera trap for Applications, 14 (5), 1535–1546. amphibians, reptiles, small mammals, and large invertebrates, PLoS One, 12 (10), e0185026. Santos, X., Llorente, G. A., Montori, A. et al. (2007) Evaluating factors affecting amphibian mortality on roads: the case of the Common IBM SPSS. Statistics for Windows, Version 22.0. (2013). IBM Corp, Toad Bufo bufo, near a breeding place, Animal Biodiversity and Armonk, NY. Conservation, 30 (1), 97–104. Jackson, S. D. (1996 April) Underpass systems for amphibians. In Trends Schmidt, B. R. and Zumbach, S. (2008) Amphibian road mortality and in addressing transportation related wildlife mortality from the how to prevent it: a review, Urban Herpetology. Herpetological transportation related mortality seminar, Tallahassee, Fl. Conservation, St. Louis, Missouri, 3, 157–167. Keller, I. and Largiadèr, C. R. (2003) Recent habitat fragmentation Semlitsch, R. D. (2002) Critical elements for biologically based recovery caused by major roads leads to reduction of gene flow and loss of plans of aquatic‐breeding amphibians, Conservation Biology, 16 (3), genetic variability in ground beetles, Proceedings of the Royal 619–629. Society of London B: Biological Sciences, 270 (1513), 417–423. Sinsch, U. (1988) Seasonal changes in the migratory behaviour of the Lehtinen, R. M., Galatowitsch, S. M. and Tester, J. R. (1999) toad Bufo bufo: direction and magnitude of movements, Oecologia, Consequences of habitat loss and fragmentation for wetland 76 (3), 390–398. amphibian assemblages, Wetlands, 19 (1), 1–12. Tao, H., Gemmer, M., Bai, Y. et al. (2011) Trends of streamflow in the Lesbarrères, D., Lodé, T. and Merilä, J. (2004) What type of amphibian Tarim River Basin during the past 50years: Human impact or climate tunnel could reduce road kills? Oryx, 38 (2), 220–223. change? Journal of Hydrology, 400 (1), 1–9. Lode, T. (2000) Effect of a motorway on mortality and isolation of wild- Time and Date. (2017) Past Weather in Glasgow, Scotland, United Kingdom life populations, AMBIO: A Journal of the Human Environment, 29 (3), - September 2016. Accessed at: timeanddate.com (February 2017). 163–166. Todd, B. D., Scott, D. E., Pechmann, J. H. et al., 2010. Climate change cor- Matos, C., Petrovan, S., Ward, A. I. et al. (2017) Facilitating permeability relates with rapid delays and advancements in reproductive timing of landscapes impacted by roads for protected amphibians: pat- in an amphibian community. Proceedings of the Royal Society of terns of movement for the great crested newt, PeerJ, 5, e2922. London B: Biological Sciences, p.rspb20101768. Miaud, C., Guyétant, R. and Elmberg, J. (1999) Variations in life-history Todd, B. D. and Winne, C. T. (2006) Ontogenetic and interspecific vari- traits in the common frog Rana temporaria (Amphibia: Anura): a lit- ation in timing of movement and responses to climatic factors dur- erature review and new data from the French Alps, Journal of ing migrations by pond-breeding amphibians, Canadian Journal of Zoology, 249 (1), 61–73. Zoology, 84 (5), 715–722. North Lanarkshire Council. (2003) Applications for planning and environ- Veenbaas, G. and Brandjes, J. (1999) Use of fauna passages along ment committee 10 December 2003. Department of Planning and waterways under highways, in Proceedings of the International Environment. Accessed et : northlanarkshire.gov.uk (February 2017). Conference on Wildlife Ecology and Transportation, Florida Department of Transportation, Tallahassee, pp. 253–258. North Lanarkshire Council. (2006) Applications for planning and envir- onment committee 13 December 2006. Department of Planning and White, K. J., Mayes, W. M. and Petrovan, S. O. (2017) Identifying path- Environment. Accessed at: northlanarkshire.gov.uk (February 2017). ways of exposure to highway pollutants in great crested newt (Triturus cristatus) road mitigation tunnels, Water and Environment Osbourn, M. S., 2012. Initial juvenile movement of pond-breeding amphi- Journal, 31 (3), 310–316. bians in altered forest habitat. University of Missouri-Columbia. ............................................................................................... .................................................................. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png BioScience Horizons Oxford University Press

The effect of rainfall upon the behaviour and use of under-road culverts in four amphibian species

Loading next page...
 
/lp/oxford-university-press/the-effect-of-rainfall-upon-the-behaviour-and-use-of-under-road-071MajQ3JY

References (27)

Publisher
Oxford University Press
Copyright
Copyright © 2022 Oxford University Press
eISSN
1754-7431
DOI
10.1093/biohorizons/hzz001
Publisher site
See Article on Publisher Site

Abstract

Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzz001/5481437 by guest on 17 August 2022 BioscienceHorizons Volume 11 2018 10.1093/biohorizons/hzz001 ............................................................................................ ..................................................................... Research article The effect of rainfall upon the behaviour and use of under-road culverts in four amphibian species Timothy Gleeson , Silviu Petrovan and Anna Muir University of Chester, Parkgate Road, Chester CH1 4BJ, UK *Corresponding author: Timothy Gleeson, 22 Isleham Close, L19 4XS Liverpool, UK. Tel: +44 07505267958. Email: timothy.j.gleeson@gmail.com Supervisor: Dr Anna Muir, Department of Biological Sciences, University of Chester, Parkgate Road, Chester CH1 4BJ, UK. Email: a.muir@chester.ac.uk ............................................................................................ ..................................................................... Habitat fragmentation and road mortalities are major contributors towards declines in amphibian populations. This has seen the introduction of culverts, passages that run under roads and provide safe passage for amphibians. Research investigating the effects of rainfall upon amphibian culvert use is limited. This study, conducted at Frankfield Loch in Glasgow, assesses how time elapsed since rainfall influences migration behaviour and the use of culverts across four different species; common toads (Bufo bufo), common frogs (Rana temporaria and newts, a group composed of smooth newts (Lissotriton vulgaris) and palmate newts (Lissotriton helveticus). Analysis of images taken by a custom made, time lapse camera found that significantly fewer common toads (r = 0.148, n = 468, p = 0.001) and common frogs (r = −0.175, n = 106, p = 0.037) used the culvert as time since rainfall increased. This may have been caused by the culvert not maintaining wet enough conditions for amphi- bians. The study also found that more newts (r = 0.272, n = 92, p = 0.004) and common toads (r = 0.531, n = 19, p = 0.010) were using the culvert to move away from Frankfield Loch as time since rainfall increased. An increase in juvenile newts was also observed as time since rainfall increased (r = 0.214, n = 92, p = 0.020). This may have been caused by a decrease in baro- metric pressure, which follows a decrease in rainfall, acting as a cue for migration and juvenile dispersal. The study recom- mends careful consideration of the design of each culvert, incorporating species-specific preferences and the requirements of juveniles. The study also suggests that where possible the culvert should be designed to hold water for longer. Key words: Bufo bufo, Rana temporaria, culverts, amphibian, migration behaviour, rainfall Submitted on 23 May 2018 ............................................................................................ ..................................................................... individuals generally follow the same migratory routes Introduction (Gamble, McGarigal and Compton, 2007; Santos et al., 2007). As a result, intercepting these migration routes pre- Roads have been recognized as a major factor in the global sents amphibians with a barrier that prevents them from population decline of amphibians (Glista, DeVault and accessing breeding sites and reproducing (Santos et al. 2007). DeWoody, 2008; Petrovan and Schmidt, 2016). Amphibians are strongly impacted by road mortalities and roads can fur- Road fragmentation can isolate populations. This reduces ther lead to the fragmentation of amphibian habitat and can gene flow between populations and can lead to a reduction in intersect their migratory routes (Glista, DeVault and genetic variability. This in turn can increase the population’s DeWoody, 2008). Semlitsch (2002) describes the seasonal vulnerability to extinction (Keller and Largiadèr, 2003; cycles of amphibian migration; overwintering in terrestrial Cushman, 2006). Dispersal may also be reduced in fragmen- areas before moving to aquatic breeding sites in the spring. ted landscapes meaning that immigrants cannot bolster strug- Breeding site fidelity is often maintained across the years and gling populations or recolonize and re-establish populations ............................................................................................... .................................................................. ©The Author(s) 2018. Published by Oxford University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzz001/5481437 by guest on 17 August 2022 Research article Bioscience Horizons � Volume 11 2018 ............................................................................................... .................................................................. following a local extinction (Lehtinen, Galatowitsch and Tester, Network, 2013). Following a series of planning applications, 1999; Cushman, 2006). It is, therefore, vital that connectivity work was carried out to upgrade Loch Road (North between populations, breeding sites, overwintering sites and Lanarkshire Council, 2003). The plans were designed so that migration routes is maintained (Lehtinen, Galatowitsch and the road would act as less of a barrier to wildlife and incorpo- Tester, 1999; Semlitsch, 2002; Santos et al. 2007). rated the construction of the three culverts that pass under the road (North Lanarkshire Council, 2006). The culvert is an In response to this, ecological mitigation measures can be amphibian road tunnel, built by ACO Germany, which has a installed. The most promising measure for amphibians are gridded roof to allow rain and natural light into the culvert road tunnels or culverts (Beebee, 2013). These are passages, and is a widely used design type in Europe (Glista, DeVault varying from 0.3 to 2 m in diameter, that run under the road and DeWoody, 2009; White, Mayes and Petrovan, 2017). (Glista, DeVault and DeWoody, 2009). It has been shown that pipe culverts can reduce road mortality on a section of Method of observation motorway in France from almost 100% down to 23% in common toads, Bufo bufo (Lode, 2000). Similarly, Dodd, To observe the amphibians, a custom made, infra-red, time Barichivich and Smith (2004) found that in the Paynes Prairie lapse camera was developed and set up inside the culvert by State Preserve in Florida the presence of culverts reduced road Froglife, a UK-based charity dedicated to amphibian and rep- mortalities in the local wildlife, including southern toads, tile conservation (Froglife, 2017). Rather than use standard Bufo terrestris, and various other frog species, by 93.5%. motion detection, which has been shown to be largely inef- fective for slow moving and very small species such as amphi- Rainfall has been shown to alter migration behaviour in bians (Pagnucco, Paszkowski and Scrimgeour, 2011), these amphibians and lead to earlier breeding (Beebee et al. 2002; cameras record images every 10 s, 24 h/day during the moni- Todd and Winne, 2006; Todd et al. 2010). Equally, for some toring period. Each image captured by the camera included species of amphibians, migration can be triggered by a the time and date that it was taken. The images are later ana- decrease in barometric pressure caused by a decrease in rain- lysed using a custom made software script that applies set fall (Elewa, 2005; Osbourn, 2012). However, the effect of levels of pixel change detection between consecutive frames precipitation upon amphibian movements is frequently and is shown as a green outline highlighting the affected area, ignored (Todd et al. 2010) and very little information exists indicating movement (Figure 1). Images from the camera on how rainfall patterns affect amphibian usage of road miti- were collected between the 11 September and the 30 gation structures. Therefore, the aim of this study was to iden- September 2016. Froglife granted the ethical approval for the tify if and how rainfall affects the behaviour of amphibians study. using culverts. The study investigated how each of the follow- ing parameters correlated with time since rainfall: the total Previous research supports the use of camera traps as the number of amphibians that entered the culvert; the number of most effective method of observing amphibians inside a cul- amphibians that made a full crossing; the direction of vert, as unlike the alternative methods (sand beds, ink beds amphibian travel and the numbers of juveniles and adults that and pitfall traps), camera traps are not intrusive, can identify were observed in the culvert. amphibians to a species level and can provide data about time, date, behaviour and age (Veenbaas and Brandjes, 1999; Methods Hobbs and Brehme, 2017). Study site Located in Seven Lochs Wetland Park in Glasgow, UK, the culvert used for this investigation is one of three that runs underneath Loch Road. Loch Road runs north-east of Frankfield Loch and separates the loch from a large area of woodland, grassland and wetlands. Both the loch and the area of woodland, grassland and reed beds accommodate amphibian populations of common frogs (Rana temporaria), common toads (B. bufo), smooth newts (Lissotriton vulgaris) and palmate newts (Lissotriton helveticus) (S. Petrovan, per- sonal communication; glasgownaturalhistory.org.uk, 2016). As these species all exhibit seasonal migration between aquatic breeding sites (such as Frankfield Loch) and terrestrial habitat for overwintering and foraging (Semlitsch, 2002), then the recently constructed Loch Road is intersecting a key Figure 1. Image taken inside the culvert under Loch Road. section of the migratory route for these populations. Highlighted in green is a common frog. Recorded at the bottom of the Seven Lochs Wetland Park seek to protect the heritage of image is information about which camera took the image, the date and the time. the site and conserve local wildlife habitat (GCV Green ............................................................................................... .................................................................. 2 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzz001/5481437 by guest on 17 August 2022 Bioscience Horizons � Volume 11 2018 Research article ............................................................................................... .................................................................. Following automated analysis to select images with Common frogs were identified as having webbed feet, trav- detected movement, the images were analysed individually elling using a series of hops and having a pointed snout and manually using the programme FastStone (FastStone (Cooke, 2004; Halliday, 2012). Common toads were distin- soft, 2016). Upon identifying an amphibian, information guished by their toes being separated rather than webbed, about the image was recorded. This included the date and moving by walking and having a more rounded snout time of photo, the species of amphibian, the direction of tra- (Cooke, 2004; Halliday, 2012). Using the dorsal angle of vel, the age class of the amphibian, the behaviour being exhib- view of the camera, juvenile and female smooth newts were ited, the number of images the individual was recorded in and indistinguishable from palmate newts, and as no great crested the position of the amphibian in the culvert. The position that newts were recorded in this study, all observations of newts an individual was in the culvert was recorded as either left, were grouped into a single ‘newt’ category. centre left, centre, centre right or right and was determined by Rainfall data was retrieved from www.timeanddate.com, dividing the image up equally into vertical bands, from left to an organization that monitors and records weather data right. The position of an individual could change as it moved (Time And Date, 2017). These data were collected from the through the culvert and this would be recorded as the range weather station at Glasgow Airport, 11 miles away from of positions the individual moved across. The direction was Loch Road. These data were recorded twice every hour, at recorded as either moving ‘in’, towards the top of the image 30 min intervals. This meant that rainfall data for exact times and towards the loch, or as ‘out’, towards the bottom of the was not available. Due to this all the data from the images image and away from the loch. If an individual was observed were organized so that it was analysed per hour (Table 2). entering the image, turning around and leaving the way it The amount of time elapsed since last rainfall, measured in came it in, then this was recorded as either ‘in/out’, if it was hours, was then incorporated into the data. During rainfall, initially moving towards the top of the image, or if the oppos- the amount of time since last rainfall was recorded as 0. For ite was true, ‘out/in’. The age of each individual was subject- the hours where there was no rainfall, the amount of time ively determined by its size. Large individuals were recorded since the last known rainfall was recorded. The rainfall data as ‘adult’ and smaller individuals were recorded as ‘juvenile’. was recorded in 30 min intervals at either 20 min past the To classify the behaviour of an individual, an ethogram con- hour or 50 min past the hour. structed by Froglife was used (Table 1). An individual may exhibit multiple behaviours in the culvert, all of which would Statistical analyses were carried out using IBM SPSS ver- be recorded. sion 22.0 (IBM Corp., 2013). As the variable that was Table 1. Ethogram for behaviour of amphibians using the culvert Behaviour Description Moving Direct movement through the culvert Creeping A combination of small movements and hesitations Turns 180 degree change in direction heading either back into, or out of the tunnel Hesitant Stationary for 1 or 2 frames before moving on (10–20 s) Waiting Pausing for 3 or more frames (30 s or more), often seen to pivot on the spot Freezes Absolutely motionless—does not trigger motion detection until moving on again Exploring Strong lateral or stochastic movements across the tunnel, no clear heading Table 2. Small section of database arranged to an hourly basis, spanning hours 66–69 Hour Amphibian count Common frog count Newt count Common toad count Time elapsed since rainfall (h) 66 2 0 0 2 0.67 67 1 0 0 1 1.67 68 3 1 2 0 2.67 69 5 4 1 0 0 ............................................................................................... .................................................................. 3 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzz001/5481437 by guest on 17 August 2022 Research article Bioscience Horizons � Volume 11 2018 ............................................................................................... .................................................................. investigated sought to measure the strength of an association between two sets of non-parametric, scale data, a 1-tailed Spearman’s rank correlation test was used to analyse each hypothesis. The Spearman’s rank correlations conducted were 1-tailed tests as the hypotheses stated which direction the correlation would be. A Kolmogorov–Smirnov test for normality was conducted for each set of data and a histogram was produced. The following variables were tested for correl- ation with time since rainfall: the total number of amphibians that entered the culvert; the number of amphibians that moved through the culvert to make a full crossing; the direc- tion of amphibian travel and the numbers of juveniles and adults that were observed in the culvert. Each parameter was studied considering all amphibian species collectively as well as considering each species individually. Results After the automated selection from over 92 000 images, 7427 ‘positive’ images were analysed manually, and 431 amphi- Figure 2. Scatter plot, with line of best fit, showing a marginally bians were recorded by the camera across twenty days. The significant correlation between the number of ‘moving’ newts and the majority were newts which were recorded on 205 occasions. amount of time elapsed since rainfall (R = 0.014). There were many The next most common species were common frogs, totalling hours during rainfall, or shortly after, where no newts were observed 204 observations, whilst common toads were recorded on 26 ‘moving’, however as time since rainfall increased more newts were occasions. ‘Movement’ was the most commonly recorded ‘moving’ through the culvert each hour. Results of 1-tailed Spearman’s behaviour, being displayed in 404 instances and 338 amphi- rank correlation were r = 0.173, n = 92, p = 0.049. Circle thickness bians were moving out (away from the loch), more than any indicates frequency of data point. other direction. There were 295 juveniles and 135 adults recorded. On the 18th day of the investigation, the lens became obscured by condensation due to a high level of mist A highly significant correlation was found between the num- in the tunnel and until the camera was serviced no data could ber of amphibians moving in the ‘in’ direction, towards be collected from the affected images. As the study was con- Frankfield Loch and the amount of time since rainfall (r = ducted during a rainy period, large amounts of the data are −0.213, n = 157, p = 0.004) (Figure 3). Splitting the data by concentrated during rainfall or shortly after rainfall. species revealed that a highly significant correlation existed between common toads moving into the culvert and the amount of time since rainfall (r = −0.728, n = 19, p = Hypothesis 1—The number of amphibians that moved <0.001) (Figure 4). No significant correlations were found for through the culvert (in either direction) will increase as common frogs (r = −0.118, n = 106, p = 0.114) or for newts time since rainfall increases. (r = −0.125, n = 92, p = 0.118). There was no significant correlation between the number of Hypothesis 3—More amphibians will be moving away amphibians observed ‘moving’ through the culvert and the from the body of water (out) as time since rainfall time since the last rainfall (r = −0.53, n = 157, p = 0.225). increases. When splitting the data by species there was no significant correlation between time since rainfall and the number of ‘moving’ common frogs (r = −0.118, n = 106, p = 0.114) or There was no significant correlation between the number of common toads (r = −0.331, n = 19, p = 0.083). There was, amphibians moving in the ‘out’ direction, away from the however, a marginal significant correlation between the num- Loch, with the amount of time elapsed since the last rainfall (r ber of ‘moving’ newts and the time since rainfall (r = 0.173, n = 0.37, n = 157, p = 0.325). Splitting the data by species how- = 92, p = 0.049) (Figure 2). A large number of data points are ever found that highly significant correlations existed for concentrated during rainfall or shortly after rainfall but these common toads (r = 0.531, n = 19, p = 0.010) and newts (r = data predominantly show few or no newts were observed dur- 0.272, n = 92, p = 0.004) (Figure 5). No significant correl- ing these hours whilst higher numbers of newts were seen per ation was found for common frogs (r = −0.099, n = 106, p = hour as time since rainfall increased. 0.156). Hypothesis 2—Fewer amphibians will be moving towards Hypothesis 4—Fewer amphibians will be observed using the body of water (in) as time since rainfall increases. the culvert as time since rainfall increases. ............................................................................................... .................................................................. 4 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzz001/5481437 by guest on 17 August 2022 Bioscience Horizons � Volume 11 2018 Research article ............................................................................................... .................................................................. (Figure 6). Splitting the data according to species found that highly significant correlations existed for common frogs (r = −0.190, n = 468, p=<0.001) and common toads (r = 0.148, n = 468, p = 0.001) (Figure 7). Hypothesis 5—The number of juvenile amphibians using the culvert will decrease as time since rainfall increases. There was a significant correlation between number of juve- niles and the amount of time since the last rainfall (r = −0.106, n = 468, p = 0.011) (Figure 8). Dividing the data by species revealed that there was a significant correlation for newts (r = 0.214, n = 92, p = 0.020) (Figure 9) but not for common frogs (r = 0.016, n = 106, p = 0.436) and common toads (r = 0.281, n = 19, p = 0.122). Hypothesis 6—The number of adult amphibians using the culvert will decrease as time since rainfall increases. There was a highly significant correlation between the num- Figure 3. Scatter plot, with line of best fit, showing a significant ber of adults using the culvert each hour and the amount of correlation between the number of amphibians that moved into the time elapsed since rainfall (r = −0.157, n = 168, p =< 0.001) culvert, towards Frankfield Loch and the amount of time since rainfall (Figure 10). Splitting the data by species revealed that signifi- (R = 0.025). Results of 1-tailed Spearman’s rank correlation were as follows: r = −0.213, n = 157, p = 0.004. Circle thickness indicates cant correlations existed for both common frogs (r = −0.175, frequency of data point. n = 106, p = 0.037) and common toads (r = −0.445, n = 19, p = 0.028) (Figures 11). No significant correlation was found for newts (r = −0.087, n = 92, p = 0.204). Discussion Culvert usage A significant correlation was found when the number of amphibians using the culvert was compared with the amount of time elapsed since rainfall. As the time since rainfall increased, fewer common toads and common frogs were observed using the culvert. This correlation was not found to exist for newts. Glista, DeVault and DeWoody (2009) and Jackson (1996) report that many amphibians require wet con- ditions in culverts before they use them. The culvert under Loch Road is grated to allow in rain. However, the rainfall is unlikely to remain in the culvert for long as it was designed to maintain hydrological links between the loch and the sur- rounding environment (North Lanarkshire Council, 2006; GCV Green Network, 2013). Therefore, it is possible that the culvert not being wet enough could be a possible explanation for the negative correlation between common frogs and com- mon toads using the culvert and the time elapsed since the last Figure 4. Scatter plot, with line of best fit, showing how fewer common rainfall. toads moved into the culvert and towards Frankfield Loch in correlation with the amount of time elapsed since rainfall (R = 0.249). Results of Furthermore, it has been reported that amphibians prefer 1-tailed Spearman’s rank correlation were as follows: r = −0.728, n = 19, wet conditions for migration in general, which may also be a p=< 0.001. Circle thickness indicates frequency of data point. factor influencing the negative correlation (Jackson, 1996). To better understand the factors responsible for fewer com- Therewas ahighlysignificant correlation between the number mon frogs and common toads using the culvert, further study, of amphibians using the culvert with the amount of time elapsed which contrasts both the numbers of amphibians moving out- since the last rainfall (r = −0.158, n = 468, p =< 0.001) side the culvert and across the road with the numbers of ............................................................................................... .................................................................. 5 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzz001/5481437 by guest on 17 August 2022 Research article Bioscience Horizons � Volume 11 2018 ............................................................................................... .................................................................. Figure 5. Scatter plots, with line of best fit, showing correlation between the amount of time since rainfall and (left) the number of common toads moving out of the culvert away and from Frankfield Loch (R = 0.093) (r = 0.531, n = 19, p = 0.010) and (right) the number of newts moving out of the culvert (R = 0.028) (r = 0.272, n = 92, p = 0.004). Circle thickness indicates frequency of data point. maintain hydrological links) would help clarify the effect that the internal conditions of the culvert had on the results. No significant correlation was found to exist between the number of newts using the culvert and the time elapsed since rainfall. One of the limitations of the study was that the cam- era was unable to distinguish between smooth newts and pal- mate newts. The newt category therefore was a combination of both smooth and palmate newts and this may well have influenced the results. To better understand how time elapsed since rainfall influences smooth and palmate newt usage of culverts, further investigation will be required. An alternative method to the infra-red camera, potentially pitfall traps, would be needed so as to distinguish the two newt species but this has substantial bias implications in terms of influencing the behaviour of the newts. Another limitation of the study is that the rainfall was not recorded at the study site but at the weather station at Glasgow Airport which may result in slight disparities between the recorded rainfall and the actual rainfall. These differences, however, would be minimal and unlikely to Figure 6. Scatter plot, with line of best fit, showing that fewer amphibians used the culvert in correlation with the amount of time impact the results. It would be preferable if in future studies elapsed since the rainfall (R = 0.010). Results of 1-tailed Spearman’s rainfall could be recorded at the study site. rank correlation were as follows: r = −0.158, n = 468, p =< 0.000. Circle thickness indicates frequency of data point. The study took place over a 20-day period in September which is a relatively short timeframe. Although previous research has shown that amphibian migration away from amphibians using the culvert, in correlation with the time breeding sites does occur during this timeframe (Sinsch, 1988; elapsed since last rainfall, would be required. In this study, Miaud, Guyétant and Elmberg,1999; Matos et al. 2017), fur- only one culvert was used but it would be beneficial if future ther research would benefit from extending the duration of studies could involve more culverts. Being able to compare the study. different culvert designs (such as ones that are not designed to ............................................................................................... .................................................................. 6 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzz001/5481437 by guest on 17 August 2022 Bioscience Horizons � Volume 11 2018 Research article ............................................................................................... .................................................................. Figure 7. Scatter plots, with line of best fit, showing correlation between the amount of time elapsed since rainfall and (left) the number of common frogs that used the culvert (R = 0.026) (r = −0.190, n = 468, p=<0.001) and (right) between the number of common toads that used the culvert (R = 0.017) (r = 0.148, n = 468, p = 0.001). Circle thickness indicates frequency of data point. Figure 8. Scatter plot, with line of best fit, showing correlation Figure 9. Scatter plot, with line of best fit, showing correlation between the number of juveniles that used the culvert and the between the number of juvenile newts that used the culvert and the amount of time since rainfall (R = 0.003). Results of 1-tailed amount of time elapsed since the last rainfall (R = 0.018). Results of Spearman’s rank correlation were as follows: r = −0.106, n = 468, 1-tailed Spearman’s rank correlation were as follows: r = 0.214, n = 92, p = 0.011. Circle thickness indicates frequency of data point. p = 0.020. Circle thickness indicates frequency of data point. in the ‘out’ direction, away from Frankfield Loch. No correl- Direction of travel ation, in either direction of travel, was found to exist when As time since rainfall increased, significantly fewer common common frogs were compared with time since rainfall. The toads were travelling in the ‘in’ direction, towards Frankfield results would suggest that an increasing absence of rainfall Loch. Alongside this, as time since rainfall increased, signifi- encourages common toads and newts to move away from the cantly more common toads and newts were observed moving loch, in the ‘out’ direction. ............................................................................................... .................................................................. 7 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzz001/5481437 by guest on 17 August 2022 Research article Bioscience Horizons � Volume 11 2018 ............................................................................................... .................................................................. As common toads and newts migrate away from breeding research that states that common toads and other amphibians sites and towards terrestrial areas, the correlations that exist typically prefer wet conditions for migration (Sinsch, 1988; between direction of travel and time since rainfall may suggest Jackson, 1996; Glista, DeVault and DeWoody, 2009). that an increasing absence of rainfall encourages migration Significantly more juvenile newts and common frogs were towards terrestrial sites. This, however, contradicts previous observed in the culvert as time since rainfall increased whilst significantly fewer adult common toads were found using the culvert as time since rainfall increased. This, combined with more amphibians moving away from Frankfield Loch, would indicate that an increasing absence of rainfall may result in an increase in juvenile dispersal (Rothermel, 2004; Cushman, 2006; Matos et al. 2017). This too, however, would contra- dict what previous research has found as juveniles also favour wet conditions for dispersal (Rothermel, 2004; Cushman, 2006). The habitat surrounding Frankfield Loch is a wetland and the culverts that run under Loch Road are designed to estab- lish a hydrological link between the loch and the surrounding wetlands (North Lanarkshire Council, 2006; GCV Green Network, 2013). It may, therefore, be possible that the sur- rounding area remains wet enough to still favour dispersal, even after extended periods of no rainfall (Lesbarrères, Lodé and Merilä, 2004). As discussed by Elewa (2005), a decrease in barometric pressure, which is brought about by a decrease in rainfall, may encourage amphibians to initiate migration (Tao et al. 2011). Juvenile dispersal may also be prompted by a decrease Figure 10. Scatter plot, with line of best fit, showing correlation in barometric pressure (Osbourn, 2012). It may therefore be between the number of adults that used the culvert and the amount of possible that as the culvert maintained hydrological links time since rainfall (R = 0.017). Results of 1-tailed Spearman’s rank between the loch and the surrounding environment, the con- correlation were as follows: r = −0.157, n = 168, p =< 0.001. Circle ditions remained wet and juvenile newts instead used a thickness indicates frequency of data point. decrease in barometric pressure as a cue to begin migrating Figure 11. Scatter plots, with line of best fit, showing correlation between the amount of time elapsed since the last rainfall and (left) the number of adult common frogs that used the culvert (R = 0.028) (r = −0.175, n = 106, p = 0.037) and (right) between the number of adult common toads that used the culvert (R = 0.409) (r = −0.445, n = 19, p = 0.028). Circle thickness indicates frequency of data point. ............................................................................................... .................................................................. 8 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzz001/5481437 by guest on 17 August 2022 Bioscience Horizons � Volume 11 2018 Research article ............................................................................................... .................................................................. and dispersing. In order to better understand this, further barometric pressure that is associated with a decrease in rain- research is required that focuses on comparing barometric fall. The study recommends careful consideration of the pressure with dispersal away from breeding sites. design of each culvert, particularly with regards to juveniles, and that where possible the culvert should hold water for longer. This study provides insight into how rainfall affects Implications for culvert design and the behaviour and culvert use of common frogs, common amphibian conservation toads and newts. Our study found that a lack of rainfall saw more juveniles moving through the culvert. Schmidt and Zumbach (2008) Author’s biography suggest that imperfections in culvert design are more pro- nounced for juveniles and are likely to be less effective for Tim Gleeson studied Animal Behaviour at the University of them, compared to adults. This would imply that an increas- Chester and this research is part of his undergraduate disser- ing lack of rainfall may see more juveniles avoiding or failing tation. His fields of interest include conservation and animal to use mitigation measures. It is, therefore, recommended that welfare. culverts in locations which experience extensive periods of no rainfall during times of juvenile dispersal, should pay careful Silviu Petrovan was Conservation Coordinator at Froglife attention to the specific preferences in culvert design exhibited where he remains a trustee and is a Research Associate at by the target species and particularly juveniles. University of Cambridge. He set up the monitoring work, One of the factors that may have been responsible for con- contributed to designing the equipment and automated ana- tradictions between the results of this study and previous lysis and provided the camera trap images. studies is that the culvert was designed to provide a hydro- logical link between the loch and the surrounding habitat Anna Muir is a lecturer in Conservation Biology at the (North Lanarkshire Council, 2006; GCV Green Network, University of Chester and supervised the research. 2013). Although this may have maintained wet conditions in the surrounding habitat, it may have caused water not to remain in the culvert for long, leaving the culvert too dry for References amphibians. Amphibians require wet conditions to prevent desiccation of their skin (Lesbarrères, Lodé and Merilä, 2004) Beebee, T. J. (2013) Effects of road mortality and mitigation measures and may, therefore, have avoided using the culvert if it was on amphibian populations, Conservation Biology, 27 (4), 657–668. not wet enough. Therefore, the ideal design would hold water Beebee, T. J., Blaustein, A. R., Root, T. L. et al. (2002) Amphibian phen- in the culvert for longer, as is also reported by Glista, DeVault ology and climate change, Conservation Biology, 16 (6), 1454. and DeWoody (2009) but without being flooded (Schmidt and Zumbach, 2008). This may be difficult to achieve in prac- Cooke, F. (2004) The Encyclopedia of Animals: A Complete Visual Guide, tice as water levels are likely to vary substantially at most Univ of California Press, Oakland, California. sites. The design feature with the slotted top surface as in the Cushman, S. A. (2006) Effects of habitat loss and fragmentation on studied tunnel is likely to have substantial advantages by amphibians: a review and prospectus, Biological Conservation, 128 allowing free temperature and humidity exchange between (2), 231–240. the tunnel and the outside environment but there are poten- tially problematic inflows of heavily polluted water from the Dodd, C. K., Barichivich, W. J. and Smith, L. L. (2004) Effectiveness of a bar- road surface (White, Mayes and Petrovan, 2017). This design rier wall and culverts in reducing wildlife mortality on a heavily trav- feature, however, should be considered in the development of eled highway in Florida, Biological Conservation, 118 (5), 619–631. other culverts to maximize their effectiveness but might Elewa A. M. (2005) Migration of Organisms, Springer, New York City, NY. require additional measures, such as annual cleaning in early spring to remove road salt and other pollutants that might FastStone MaxView. (2016). FastStone. accumulate over winter. Froglife. What we do. Accessed at: froglife.org (February 2017). Gamble, L. R., McGarigal, K. and Compton, B. W. (2007) Fidelity and dis- persal in the pond-breeding amphibian, Ambystoma opacum: Conclusion implications for spatio-temporal population dynamics and conser- vation, Biological Conservation, 139 (3), 247–257. This study found that as time since rainfall increased, fewer common toads and common frogs would use the culvert Glasgow & Clyde Valley (GCV). (2013). Green network, Seven Lochs Wetland under Loch Road. The results of the investigation also suggest Park vision and masterplan. Accessed at: issuu.com (Feb 2017). that more newts and common toads would move away from Frankfield Loch and more juvenile newts would use the cul- glasgownaturalhistory.org.uk., (2016). Frankfeild Loch area species list vert, as time since rainfall increased. This contradicts previous [online]. Accessed at https://www.glasgownaturalhistory.org.uk/ research and may have been a result of a decrease in biodiversity/Frankfield_splist.pdf (6 November 2017). ............................................................................................... .................................................................. 9 Downloaded from https://academic.oup.com/biohorizons/article/doi/10.1093/biohorizons/hzz001/5481437 by guest on 17 August 2022 Research article Bioscience Horizons � Volume 11 2018 ............................................................................................... .................................................................. Glista, D. J., DeVault, T. L. and DeWoody, J. A. (2008) Vertebrate road Pagnucco, K. S., Paszkowski, C. A. and Scrimgeour, G. J. (2011) Using mortality predominantly impacts amphibians, Herpetological cameras to monitor tunnel use by long-toed salamanders Conservation and Biology, 3 (1), 77–87. (Ambystoma macrodactylum): an informative, cost-efficient tech- nique, Herpetological Conservation and Biology, 6 (2), 277–286. Glista, D. J., DeVault, T. L. and DeWoody, J. A. (2009) A review of mitiga- tion measures for reducing wildlife mortality on roadways, Petrovan, S. O. and Schmidt, B. R. (2016) Volunteer conservation action Landscape and Urban Planning, 91 (1), 1–7. data reveals large-scale and long-term negative population trends of a widespread amphibian, the common toad (Bufo bufo), PLoS Halliday, T., (2012). In C. Uhlenbroek, eds, Illustrated Encyclopaedia of One, 11 (10), e0161943. Animal Life, Dorling Kindersley, London, United Kingdom, pp. 202–249. Rothermel, B. B. (2004) Migratory success of juveniles: a potential con- straint on connectivity for pond‐breeding amphibians, Ecological Hobbs, M. T. and Brehme, C. S. (2017) An improved camera trap for Applications, 14 (5), 1535–1546. amphibians, reptiles, small mammals, and large invertebrates, PLoS One, 12 (10), e0185026. Santos, X., Llorente, G. A., Montori, A. et al. (2007) Evaluating factors affecting amphibian mortality on roads: the case of the Common IBM SPSS. Statistics for Windows, Version 22.0. (2013). IBM Corp, Toad Bufo bufo, near a breeding place, Animal Biodiversity and Armonk, NY. Conservation, 30 (1), 97–104. Jackson, S. D. (1996 April) Underpass systems for amphibians. In Trends Schmidt, B. R. and Zumbach, S. (2008) Amphibian road mortality and in addressing transportation related wildlife mortality from the how to prevent it: a review, Urban Herpetology. Herpetological transportation related mortality seminar, Tallahassee, Fl. Conservation, St. Louis, Missouri, 3, 157–167. Keller, I. and Largiadèr, C. R. (2003) Recent habitat fragmentation Semlitsch, R. D. (2002) Critical elements for biologically based recovery caused by major roads leads to reduction of gene flow and loss of plans of aquatic‐breeding amphibians, Conservation Biology, 16 (3), genetic variability in ground beetles, Proceedings of the Royal 619–629. Society of London B: Biological Sciences, 270 (1513), 417–423. Sinsch, U. (1988) Seasonal changes in the migratory behaviour of the Lehtinen, R. M., Galatowitsch, S. M. and Tester, J. R. (1999) toad Bufo bufo: direction and magnitude of movements, Oecologia, Consequences of habitat loss and fragmentation for wetland 76 (3), 390–398. amphibian assemblages, Wetlands, 19 (1), 1–12. Tao, H., Gemmer, M., Bai, Y. et al. (2011) Trends of streamflow in the Lesbarrères, D., Lodé, T. and Merilä, J. (2004) What type of amphibian Tarim River Basin during the past 50years: Human impact or climate tunnel could reduce road kills? Oryx, 38 (2), 220–223. change? Journal of Hydrology, 400 (1), 1–9. Lode, T. (2000) Effect of a motorway on mortality and isolation of wild- Time and Date. (2017) Past Weather in Glasgow, Scotland, United Kingdom life populations, AMBIO: A Journal of the Human Environment, 29 (3), - September 2016. Accessed at: timeanddate.com (February 2017). 163–166. Todd, B. D., Scott, D. E., Pechmann, J. H. et al., 2010. Climate change cor- Matos, C., Petrovan, S., Ward, A. I. et al. (2017) Facilitating permeability relates with rapid delays and advancements in reproductive timing of landscapes impacted by roads for protected amphibians: pat- in an amphibian community. Proceedings of the Royal Society of terns of movement for the great crested newt, PeerJ, 5, e2922. London B: Biological Sciences, p.rspb20101768. Miaud, C., Guyétant, R. and Elmberg, J. (1999) Variations in life-history Todd, B. D. and Winne, C. T. (2006) Ontogenetic and interspecific vari- traits in the common frog Rana temporaria (Amphibia: Anura): a lit- ation in timing of movement and responses to climatic factors dur- erature review and new data from the French Alps, Journal of ing migrations by pond-breeding amphibians, Canadian Journal of Zoology, 249 (1), 61–73. Zoology, 84 (5), 715–722. North Lanarkshire Council. (2003) Applications for planning and environ- Veenbaas, G. and Brandjes, J. (1999) Use of fauna passages along ment committee 10 December 2003. Department of Planning and waterways under highways, in Proceedings of the International Environment. Accessed et : northlanarkshire.gov.uk (February 2017). Conference on Wildlife Ecology and Transportation, Florida Department of Transportation, Tallahassee, pp. 253–258. North Lanarkshire Council. (2006) Applications for planning and envir- onment committee 13 December 2006. Department of Planning and White, K. J., Mayes, W. M. and Petrovan, S. O. (2017) Identifying path- Environment. Accessed at: northlanarkshire.gov.uk (February 2017). ways of exposure to highway pollutants in great crested newt (Triturus cristatus) road mitigation tunnels, Water and Environment Osbourn, M. S., 2012. Initial juvenile movement of pond-breeding amphi- Journal, 31 (3), 310–316. bians in altered forest habitat. University of Missouri-Columbia. ............................................................................................... ..................................................................

Journal

BioScience HorizonsOxford University Press

Published: Jan 1, 2018

There are no references for this article.