Grants & Prizes Ecological Opportunities Fund Honours, Awards and Prizes Building Capacity for Ecology Fund Honorary Member List Research grants General notes Grants to attend meetings & courses Grants to run meetings & events Grant reports List of Tansley Lectures

SEPG 1357 - Date Awarded 1996

The impact of road development on a toad population – a test of density dependence in the terrestrial habitat

R. S. Oldham
Department of Biological Sciences
De Montfort University
Leicester LE7 9SU
e-mail: rsoldham@dmu.ac.uk

Abstract

In 1992 the toad population living in the hinterland to the north of a breeding site was estimated to contain about 600 adults of breeding age.  To the south the sub-population was similar, estimated as about 500.  Indices of somatic condition for the animals in the two sub-populations were similar.  Mean egg production in a sample of females was also the same in both sub-populations.  Construction of a new road in 1993 adjacent to a toad breeding site, with inadequate “toad tunnels”, partially isolated the northern part of the population and changed the relative population densities north and south of the road, although the road acted as less of a barrier than had been anticipated.  There was a strong tendency for toads to return to their former home range to the north during the summer, despite the new road.  By 1998 both sub-populations had increased, but the density in the north was estimated to be less than half that in the south.  Somatic condition had changed, with animals of both sexes having significantly better condition (p<0.05) in the north.  However, there was no significant difference in the egg production.  There is, therefore, some evidence consistent with terrestrial density dependent damping of growth in the relatively dense southern sub-population and growth enhancement in the north, but with no evidence, at this stage of separation, of the crucial aspect of impact on female productivity.

Introduction

Density dependence is well known in the larval phase of the amphibian life cycle (e.g. Wilbur, 1977) and has been demonstrated in the adult aquatic phase of Triturus alpestris (Veith, 1997).  It has also been postulated in the terrestrial phase of the common toad, Bufo bufo, from indirect evidence (Reading, 1995), but there is no field demonstration.  Large-scale field experiments to investigate density dependent effects in vertebrate populations are inherently difficult in the normal agricultural habitats of the British amphibians, but the construction of a new road through the middle of a long-term study site provided an opportunity for such an investigation.  The study also contributed to our understanding of the impact of such development on an amphibian population.

A population of common toads at Coleorton in NW Leicestershire (SK 399171) has been the subject of study for nearly 20 years (Oldham & Swan, 1991).  For convenience the terrestrial hinterland in the study site can be divided into two more or less equal parts, north and south of the A512.  This road is tangential to the small lake which is the breeding site (Fig. 1).  All the toads use the same breeding site.  It is isolated from other toad breeding sites by at least 1,000 m., at which distance dispersal from adjacent sites is minimal (Reading et al., 1991,  Halley et al., 1996).

In the breeding season of 1982 the population was estimated to contain 7,000 adults (Oldham & Swan, 1991).  Subsequently the population size has fluctuated, probably the result of changes in land use within the terrestrial population range.  Between 1988 and 1992 it was at a relatively low level, estimated as between 1,900 and 2,250 adults.

In 1992 a new road was built on the northern side of the lake (Fig. 1).  Work on the new road was scheduled to last from spring 1992 to early 1993.  If toads were to breed in the pond in 1992, massive mortality of both adults and emergent juveniles was likely.  In an effort to reduce this mortality, adults were intercepted at a perimeter fence in the spring of 1992 and held in single sex enclosures until after the breeding season, when they were released.  Under these conditions most females do not ovulate, but resorb their eggs.  As expected, some adults circumvented the perimeter fence and spawned in the pond, but there was a much reduced breeding population and a very small emergence of metamorphs in 1992.

Barriers consisting of paving slabs (Fig.2) were installed along both sides of the new road for 500 m, adjacent to the breeding site, to prevent toads getting on to the new carriageway.  Three toad tunnels were constructed to permit access of adults to and from the lake during the spring, and exodus of the metamorphs during the early summer.  The recommended specifications of the tunnels (Oldham, 1988) were scaled down (to 600 mm diameter) by the road designers with the consequence that the road now acts as a filter.  It seems likely that there is inadequate air circulation in the tunnels and they warm up too slowly in the spring.  Adult toads use the tunnels to a negligible extent (as determined by fences and pitfall traps on each side of the road), but the metamorphs use them freely.  Consequently, each spring there is a build up of adult animals at the roadside barriers on the northern side of the new road, without access to the breeding site.  To ameliorate this problem annual population management involves translocation into the lake of all the adults arriving at the barriers on the northern side of the road. 

After breeding, toads normally return to the habitat they occupied in the previous summer (Haapanen, 1974, Latham, 1997), but they disperse from the lake gradually, again using the tunnels to a negligible extent.  The northern toads do not aggregate at the proximal barriers and there is no translocation back across the road.  In the absence of any evidence to the contrary an assumption was made that the toads did not circumvent the barriers and that the northern toads remained with the southern sub-population.  If so the population composition north and south of the road would be different, with a deficiency of adult cohorts in the north, but with the southern adult cohorts augmented by the translocations.  This provided an opportunity for a field test of the hypothesis that there is terrestrial density dependent damping of growth and productivity.

The work was approached by monitoring the influx of toads in the springs of 1997 and 1998, and comparing the sizes and egg production of northern and southern sub-populations with those obtained before the road was built.

Methods

For convenience the northern sub-population is defined as comprising those toads living to the north of the new road, and assumed to disperse up to 1,000 m. during the summer; this incorporates about 160 ha.  The same definition is used for the period before 1993, based upon the route of the new road, although a small proportion of catches before 1993 cannot reliably be assigned to the appropriate sub-population.

Each year after 1993, toads were captured in 31 pitfall traps (10 l plastic buckets) placed flush with the roadside toad barrier to the north of the pond.  Drift fences with pitfall traps were placed across the southern entrances to the three tunnels to monitor toad movements.  Additionally a fence around the entire perimeter of the pond, set about one metre from the water’s edge, was in use from 1992 to 1994.  Prior to 1993 animals were trapped each year in 115 pitfall traps at 30 drift fences, constructed of poultry netting (Arntzen et al. 1995), total length 160 m, set at a mean distance of 67 m all the way round the pond.  Those traps not destroyed by the new road (82) continued in use after 1993.  The pitfall traps were inspected each morning during the four-week period of study in spring.  Additionally some toads were collected by hand at night during their migration towards the pond and samples were taken from the pond itself as part of a mark and recapture exercise.  Animals from the roadside barriers were released into the pond, except for a sample returned to the distal side of the barriers.  Those caught at the drift fences and the perimeter fence were released on the pond side of the respective fences.

An index of population size was obtained by extrapolation from the catch obtained at the fences, after correcting for fence inefficiency on the basis of data in Arntzen et al. (1995).  Population size was also estimated using mark and recapture techniques, although because of the displacement of the northern animals this was only applicable to the southern sub-population.  Toads were marked in groups according to site of capture, by a combination of toe clipping (under licence) and Panjet (Wright Health Group, Dundee) dye marks.

Individuals were measured from the anterior tip of the snout to the posterior tip of the urostyle, using an angled ruler, to the nearest 0.5 mm, and weighed to the nearest 0.1 g, after being held in water for at least one hour.   The ratio of mass to length provides an index of the somatic condition of the animals, which would be expected to reflect resource availability.  Condition is expressed using the index devised by Hemmer and Kadel (1972), and modified for Bufo bufo by Kuhn (1994) ([mass in g] x 10⁶ / [length in mm] ³).  For females, condition indices refer to unovulated animals.  Samples of phalanges were taken for subsequent age determination, the results of which will be reported elsewhere.

Female toads caught during immigration to the breeding site have rarely ovulated, especially early in the season.  If caught and held in captivity they may fail to ovulate and eventually resorb their eggs (Swan, 1986).  However, females at the breeding site and those caught later in the breeding season, or those held under suitable conditions in captivity, do spontaneously ovulate and oviposit, whether or not a male is present.  Samples of 38 such unspawned female toads (13 from 1992 and 25 from 1998) were held individually in 10 l containers in 0.1 m water and the deposited eggs counted.

Results

If there is density dependent damping of growth and productivity after 1992 it would be expected to apply with steadily increasing emphasis year by year.  Data below are predominantly those of 1997 with the exception of egg production for which a better sample was available in 1998.  Again, most of the data refer to males, which comprise the major part of the population; they also spend longer at the breeding site and are more likely than females to be recaptured within the breeding season.

Population size

Despite the roadworks and population manipulation in 1992 (see Methods), the estimated population size increased (albeit from a particularly low base) in both sub-populations in 1993 and 1994 (Fig. 3), probably the result of recruitment from the terrestrial juvenile stock already present before the road was built.  These values indicate an overall density, within a 1,000 m radius, of between 4 and 22 adults per hectare.  Most animals probably live within 500 m of the pond, in which case the adult density would range up to 65 adults per hectare.  In 1995 both sub-populations declined in numbers, probably the result of low recruitment from breeding in 1992.  Subsequently both have increased, probably related to recovery from the disruption of population manipulation and road building.  As noted above, inward dispersal from the adjacent breeding sites is thought to be minimal, so the data probably reflect the increasing success of juvenile recruitment.  Between 1995 and 1998 the rate of increase was considerably higher in the south than in the north (Fig.3).

In the northern sub-population annual survivorship, based upon mark and recapture data, was about half that in the south, probably reflecting combined impacts of road building and the continuing road barriers.  The fact that adult numbers increased at all indicates that outward migration of the post-metamorphic animals was successful.  Metamorph emigration was sampled in 1994, using pitfall traps at the distal ends of the three tunnels and large numbers are known to have used the tunnels.  It is difficult to determine whether metamorphs were able to cross on the surface of the carriageway itself.  The toad barriers were not as effective for metamorphs as for adults and it is possible that some animals were able to pass the barriers and cross the road on wet nights, when traffic density was low, and contribute to the population build up to the north of the road.

Patterns of toad movement

Evidence from observed road mortality suggested that the road-side barriers (Figs.1 & 2) were highly effective.  In the south, despite much lighter traffic on the old road, about 200 corpses were recorded in 1997, compared with less than 10 on the new road. 

Less than 10% of the northern toads used the tunnels during immigration towards the pond in 1997, the remainder were found on the distal side of the barrier.  Of those males captured at the barrier and returned to the north (n = 84) 93% were recaptured again at the barriers or had moved into the tunnels.  Of the males captured at the road barrier and released into the pond (n = 768) 8% were recaptured in the south, in the same breeding season.  This compares with 15% for males captured in, and returned to, the south (n = 877).  A handful of animals (seven: two of northern origin, five of southern origin) was captured in the north following release in the south earlier in the same breeding season, which was surprising.  This might have been the result of toads being swept over a weir and along an underground stream, which flows beneath the road towards the north.  The same explanation probably applies to a rather larger number of similar instances in the following year, which coincided with a period of heavy rainfall.  This route would not be readily available to toads during their terrestrial phase following the breeding season.

The annual, rather than seasonal, movements of toads may be examined by taking account of toads originally marked in 1996, when the same pattern of capture and release was used.  Of 569 males captured north of the barriers and released into the pond in the spring, 26% were recaptured in 1997, 25% of them back again in the north, only 1% in the south.  The corresponding values for 309 southern animals were 27% recaptured, 23% in the south, 4% in the north.  In view of the original assumptions, these were very surprising results and suggest that in the context of annual population movements, as distinct from spring movements, the road and associated barriers were ineffective.

Individual size and somatic condition

If high population density has an impact on individual growth rates, this should be reflected in smaller sized animals in the northern sub-population.  However, mean masses and lengths of adults within each sub-population will be strongly influenced by mortality and recruitment rates, which differ north and south (Fig.3).  The index of somatic condition overcomes this problem; condition values for 366 males and 158 females, measured in 1997, were virtually unaffected by body length (the slopes of the regression lines were 0.0003 and 0.0001 respectively).

In 1992, for both sexes, the condition indices did not differ in northern and southern sub-populations (p>0.1, Table 1), but five years after completion of the new road they were significantly higher in the north (p<0.05), again for both sexes.  For females in the south there was a reduction of the index value between 1992 and 1997, which may relate to increased competition in the now larger southern sub-population.  These data are therefore consistent with the hypothesis of density dependent damping of growth.

Table 1. Median somatic condition indices of toads from northern and southern sub-populations, before (1992) and after (1997) the new road was built, with the probability that the values do not differ.

Egg production

Egg numbers and egg mass are correlated in the common toad (Reading, 1986) and both are likely to be affected by female food intake.  If so, egg production provides a measure of response to density dependent population effects.  However, egg numbers are also influenced by female size (Reading, 1986) and there are several annual cohorts of females in both sub-populations.  The comparison of egg productivity in the two populations in 1998 was approached, therefore, by comparing the regression of egg counts on female length (Fig. 4).  Although the largest females and egg clutches occurred in the north, there was no significant difference in the slope of the regression lines plotted for females from northern and southern sub-populations, either in 1998 (Fig.4) or in 1992.

Discussion

The most striking result was the evidence that recaptured members of the northern sub-population had not remained in the south; the road and associated barriers were not as effective in separating the sub-populations as was first assumed.  Toads transferred manually from north to south in the spring evidently “leaked” back to the north of the road during the course of the subsequent summer.  There was no evidence that they used the tunnels.  It must be assumed that they travelled beyond the ends of the barriers or climbed over them and crossed the carriageway at night when there was low traffic flow.  Scaling the barriers would be easier during the summer than the spring, because of the growth of adjacent vegetation.  The observed leakage of northern toads back to the north during the summer emphasises the strength of their fidelity to the terrestrial range.

The proportions of animals returning back to the north during the summer, following spring translocation to the south (25%), compared with the proportion recaptured in the south after being tagged and left in the south (23%), suggests that the road is a totally ineffective barrier.  This would appear to undermine the basis of the field experiment, suggesting that the populations north and south had a similar composition.  However, the similarity of the above two values is misleading, since there was a marked difference in sampling efficiency north and south of the new road.  The paving slab barriers are very effective (probably >90% efficient during the spring) whereas the poultry netting fences are not only less efficient (perhaps 39%; Arntzen et al., 1995), but also sample a much smaller proportion of the southern sub-population.  Although northern toads apparently tried to return “home” they suffered high mortality in doing so, and this, probably coupled with lower recruitment, is reflected in the slower growth, and lower density, of the northern sub-populations (Fig.3).

Five years after partial separation of northern and southern sub-populations there are some indications of differences in the population parameters.  Reduction in intraspecific competition resulting from lower population density might be responsible for the higher somatic condition of adults in the north (Table 1), but this is not, as yet, reflected in differences in egg production (Fig. 4).  It is known that during the non-breeding season, males occupy niches closer to the breeding site than females (Oldham, unpublished, Latham, 1997).  It might be expected, therefore, that density dependent effects would be more intense amongst males.  At this stage of population separation there is no evidence to support this.  Monitoring in subsequent years might establish whether sex-specific differences in density dependent effects develop, and whether egg productivity is ultimately affected.

Acknowledgements

I am very grateful for help during fieldwork in 1992 by Andrew Alexander, Jane Doyle and Julie Johnson, for more recent site help by John Wilkinson, Jonathan Daws, Terry Green and Amanda Pepper, for site access by the late Mr. R. Collings and Mr. R. Woods, for comments on the manuscript by David Bullock, and for financial help during the 1997 field season by the BES.

References

Arntzen, J.W., Oldham, R.S. & Latham, D.M. (1995). Cost effective drift fences for toads and newts. Amphibia-Reptilia 16:137-145.

Haapanen, A. (1974). Site tenacity of the common toad, Bufo bufo (L.). Ann. Zool. Fennici. 11:251-252.

Halley, J.M., Oldham, R.S. & Arntzen, J.W. (1996). Predicting the persistence of amphibian populations with the help of a spatial model.  J. of Applied Ecology.  33: 455-470.

Hemmer, H. & Kadel, K. (1972). Gewichtszustand und Wachstumsverlauf bei der Kreuzkrote. Forma et Functio 5:113-120.

Kuhn, J. (1994).  Lebensgeschichte und demographie von erdkrotenweibchen Bufo bufo bufo (L.). Zeitschrift fur feldherpetologie. 1: 3-87.

Latham, D.M (1997).  The terrestrial habitat selection and utilisation by the common toad

(Bufo bufo L) in agricultural landscapes.  Unpublished PhD thesis. De Montfort University, Leicester.

Oldham R.S. (1988). Impact of road development on newt habitat and on a toad population.  Unpublished report to Leicestershire County Council.

Oldham R.S. & Swan M.J.S. (1991). In: Seitz A. & Loeschcke V. (Eds.) Species conservation: A population-biological approach.  Birkhauser Verlag, Basel: 141-158.  

Reading, C.J., (1986)  Egg production in the common toad, Bufo bufo.  J. Zool. Lond.(A) 208: 99-107.

Reading C J & Clarke R T (1995)  The effects of density, rainfall and environmental temperature on body condition and fecundity in the common toad, Bufo bufo.  Oecologia   102:453-459.

Reading, C.J., Loman, L. & Madsen, T. (1991). Breeding pond fidelity in the common toad, Bufo bufo.  J. Zool. Lond. 225: 201-211.

Swan, M.J.S. (1986). The conservation ecology of  Rana temporaria and Bufo bufo in Leicestershire. Unpublished PhD thesis. Leicester Polytechnic.

Veith, M. (1987). Weight condition in Triturus alpestris – methodological and ecological aspects. In: J.J.van Gelder, H.Strijbosch & P.J.M.Bergers (Eds.), Proc.4^th Meeting Soc. Europaea Herpetol.:429-432.

Wilbur H.M. (1977). Density-dependent aspects of growth and metamorphosis in Bufo americanus. Ecology 58: 196-200.

R. S. Oldham
Department of Biological Sciences
De Montfort University
Leicester LE7 9SU
e-mail: rsoldham@dmu.ac.uk

Figure Titles

Figure 1. The main features in the immediate vicinity of the toad breeding site at Coleorton, Leicestershire.

Figure 2. The paving slab barrier and a toad tunnel entrance on the northern side of the new road.

Figure 3.  Adult toad populations north and south of the new road, estimated on the basis of pitfall catches during spring each year.

Figure 4. Egg counts for 23 toads of different lengths, from northern (triangles) and southern sub-populations, six years after the new road was built.  The lines of best fit are provided using an exponential function.

Copyright 2007 British Ecological Society. All Rights Reserved