- Email: [email protected]
Distribution of the Mallorcan midwife toad (Alytes muletensis) in relation to landscape topography and introduced predators
Biological Conservation 116 (2004) 327–332 www.elsevier.com/locate/biocon
Distribution of the Mallorcan midwife toad (Alytes muletensis) in relation to landscape topography and introduced predators Robin D. Moorea, Richard A. Griffithsa,*, Alvaro Roma´nb b
a The Durrell Institute of Conservation and Ecology, University of Kent, Canterbury, Kent CT2 7NS, UK Associacio´ per a la Recuperacio´ del Ferreret, Mateu E. Llado´ 34B. 1 C 07002, Palma de Mallorca, Balearic Islands, Spain
Received 18 January 2003; received in revised form 14 April 2003; accepted 24 April 2003
Abstract The endemic midwife toad of Mallorca (Alytes muletensis) is restricted to a small number of breeding populations in the mountainous northwest of the island. The decline of the species has been attributed to the impacts of introduced species such as the viperine snake (Natrix maura) and green frog (Rana perezi), and toads may be surviving only in areas that are suboptimal for these predators. The influence of landscape features (elevation, aspect and maximum slope) on the distribution of toads and associated predators was therefore investigated using GIS. The presence of toads was positively associated with steep slopes. At sites where they occurred with toads, the distribution of predators was negatively associated with elevation. Reproductive success within toad populations was strongly associated with the number of pools at each site, while reproductive success within individual pools was positively associated with elevation. These findings may be used to optimise the design and location of future reintroduction sites. # 2003 Elsevier Ltd. All rights reserved. Keywords: Endangered species; GIS; Introduced predator
1. Introduction According to Diamond’s (1989) ‘evil quartet’, the four principal causes of species’ extinctions are habitat fragmentation, overexploitation, introduced predators and chains of extinctions. Species on islands that have evolved in the absence of any major predators or competitors are likely to be particularly vulnerable to the introduction of predators. Periods of coevolution between predator and prey play an important role in reducing dynamic responses of one species to the other (Case and Bolger, 1991). However, novel predators can severely impact populations of prey species because there has been no opportunity for coevolution (e.g. Savidge, 1987; Rodda and Fritts, 1992; Keisecker and Blaustein, 1997; Fritts and Rodda, 1998; Short et al., 2002). There are many cases of introduced species negatively impacting native island fauna (e.g. Gaston, 1994; Wilson et al., 1998; Dowding and Murphy, 2001; Jackson, 2001; Towns et al., 2001; Burbridge and * Corresponding author. Tel.: +44-1227-823434; fax: +44-1227827839. E-mail address: r.a.griffi[email protected] (R.A. Griffiths). 0006-3207/03/$ - see front matter # 2003 Elsevier Ltd. All rights reserved. doi:10.1016/S0006-3207(03)00202-7
Manly, 2002; Hall et al., 2002), although most studies to date have focused on introduced mammals and there are few documented cases of reptile introductions. Exceptions include the introduced boa (Boa constrictor), believed to threaten endemic fauna on Cozumel Island in Mexico (Martinez-Morales and Cuaron, 1999) and the brown tree snake (Boiga irregularis), introduced accidentally to Guam in the 1950s and responsible for the extinction of many species of native fauna (Savidge, 1987; Rodda and Fritts, 1992; Fritts and Rodda, 1998; Rodda et al., 1999). The endemic midwife toad of Mallorca (Alytes muletensis) was once widespread over the island (Alcover et al., 1984), but today persists only as a small number of isolated breeding populations. These breeding populations are largely confined to the limestone gorges of the Serra de Tramuntana mountains in the northwest of the island, where populations may use one or more pools scoured by ephemeral torrents. The decline of the toad has been attributed by many workers to predation and competition from species such as the viperine snake (Natrix maura) and green frog (Rana perezi) (Tonge, 1986; Bloxam and Tonge, 1995), both of which were introduced to Mallorca around 2000 years ago. Both of
328
R.D. Moore et al. / Biological Conservation 116 (2004) 327–332
these species consume tadpole and adult stages of A. muletensis (personal observations); in addition, it is possible that interspecific competition occurs between the tadpole stages of A. muletensis and R. perezi (Spence, 2002). While some endemic fauna went extinct following the introduction of these and other predators during Roman times (Alcover et al., 1981; Tonge, 1986), A. muletensis was one of the few species to survive, although the reasons for this remain unclear. One hypothesis proposed by some authors is that toad populations have persisted in gorges with steep sides, because introduced predators find it difficult to colonise and establish populations in such areas (Bloxam and Tonge, 1995; Schley and Griffiths, 1998). A better understanding of the factors influencing the distribution of the toad and its associated predators is required before any firm conclusions can be reached. Recent advances in geographical information systems (GIS) and remote sensing technologies have facilitated the analysis of animal and plant distributions in relation to spatial environmental attributes (e.g. Farina, 1997; Corsi et al., 1999; Knutson et al., 1999; Howell et al., 2000; Lenton et al., 2000). Relatively simple environmental descriptors such as altitude, slope and aspect have been used to explain distribution and abundance patterns of taxa such as birds (e.g. Farina, 1997; Detmers and Bart, 1999) and plants (Janet, 1998), and it is likely that these variables strongly influence the microclimate of a site (Stocks and Heywood, 1994). Knowledge of how such environmental features influence the distribution of rare species and associated predators may be an important consideration in the management of the species. This study used GIS to analyse the distribution of Mallorcan midwife toads and their principle predators in relation to landscape topography. We therefore tested the hypothesis that toad populations have persisted in sites that are suboptimal for predators. In addition, we determined which site characteristics were the best predictors of reproductive success within toad populations.
2. Methods 2.1. Census methods Annual population surveys carried out at all known sites since 1991 provided information on the status of A. muletensis (Garcia and Buley, 1997; Roma´n and Mayol, 1997; Roma´n, 2001, 2002). As adults are cryptic and spend most of their lives in cracks and fissures underground, annual censuses focused on counts of tadpoles in the breeding pools. These counts were made in July each year, immediately following spring breeding activity. While tadpoles may take anything from several weeks to over a year to complete development, tadpole
numbers were thought to be at their highest during these census periods. Most breeding pools are clear and devoid of vegetation, so it is relatively easy to make direct counts of individual tadpoles (Schley et al., 1998) and counts made by different observers are quite consistent (Mun˜oz, 2000). Mean number of tadpoles counted in each pool at each site from 1991 to 2000 was taken as a measure of reproductive success of each population over this period. However, as the number of breeding pools used by a population of toads at each site varies between 1 and 19, reproductive success was expressed in two ways. First, reproductive success per population was calculated as the mean number of tadpoles based on the total tadpole count in all pools at a site. Second, reproductive success per pool per population was calculated by dividing the mean number of tadpoles counted per population by the number of pools used by toads in that population. During the annual censuses the presence or absence of Natrix maura and Rana perezi was noted. Populations were classified as having ‘predators present’ if predators had been observed within 100 m of the population on at least two separate visits. 2.2. Mapping procedure Scanned paper maps of the Serra de Tramuntana on 1:5000 scale were obtained from the Conselleria de Medi Ambient in June 2000. Torrents and 25 m contours were digitised from these scanned maps using Arcview 3.2 across the entire geographical range of A. muletensis. From these contours a digital terrain model (DTM) was constructed using Triangulated Irregular Networks (TIN). DTMs such as this serve as an important tool for deriving secondary spatial data sets such as altitude, slope and aspect (Stocks and Heywood, 1994). Slope and aspect surfaces with a grid size of 0.01 km were generated from the TIN for analysis. The locations of all toad populations were mapped using information already available on the geographic location of all known populations in the form of GPS readings and paper maps (Garcia and Buley, 1997). To provide a comparison with torrent habitats not occupied by toad populations, 50 points were selected along torrents using a random points generator extension for ArcView 3.2 and by combining all torrents so that the localities were randomly located throughout the range of the toad. 2.3. Statistical analysis Stepwise multiple logistic regression (Hosmer and Lemeshow, 1989) was used to test a number of topographical factors considered to be potentially important in influencing the distribution of the midwife toad and of the introduced predators N. maura and R. perezi. The distribution of the two predators amongst toad sites was
329
R.D. Moore et al. / Biological Conservation 116 (2004) 327–332
perfectly correlated (r=1.0) indicating that wherever N. maura are present, R. perezi also occurs nearby and vice versa. Logistic regression was considered the most appropriate analytical procedure for assessing the distribution of toads and predators because it permits the prediction of binary attributes such as presence/absence (McCullagh and Nelder, 1983) and it has proved to be a powerful technique for analysing species distributions (e.g. Jokima¨ki, 1999). The dependent or response variable in the regression—presence or absence of toads—was scored as 1 (all torrent-dwelling toad populations; n=17) or 0 (randomly generated points along torrents; n=50), respectively. Because all the random points were generated along torrents, any toad populations that inhabited artificial water bodies outside natural torrents were excluded from the analysis. The independent or explanatory variables were three topographical features associated with each locality derived from surfaces constructed using GIS; elevation (m), aspect (in degrees, where 0 =true north and 180 =true south) and maximum slope within a horizontal radius of 100 m (equal to an area of 31,416 m2). Maximum slope—rather than mean slope—was chosen as an independent variable as it was thought that just one steep slope in the vicinity of breeding site may be limiting for predators, in that it would prevent emigration of predators from the site. A radius of 100 m was chosen because this was considered to provide an appropriate representation of the immediate topography at each site. Scribner et al. (2001), for example, also characterised habitat within a 100 m radius around each pond when investigating environmental correlates of common toad (Bufo bufo) abundance and genetic diversity. The number of independent variables in the analysis was kept low due to the limited number of sites being compared; this is in accordance with the suggested minimum requirement of at least four to five times more cases than independent variables (Tabachnick and Fidell, 1983). Using the stepwise logistic regression procedure, selection of variables for inclusion in the final model was based on the Wald test (Hosmer and Lemershow, 1989; Jokima¨ki,
1999). All variables were normally distributed and therefore met the assumptions of the analysis. Factors influencing the distribution of predators were investigated by comparing landscape topography between all mapped toad populations (17 inhabiting natural torrent pools and four inhabiting artificial water bodies). Seven toad populations with predators were compared against 14 predator-free toad populations and predator presence or absence (scored as 1 and 0 respectively) was regressed against the three independent topographical variables described above. In order to assess the factors influencing reproductive success of A. muletensis over the period 1991–2000, stepwise linear regressions were made for all toad populations of (1) mean tadpole count per population, and (2) mean tadpole count per pool per population against three site characteristics (tadpole counts were log-transformed to ensure normality). In the absence of data on adult population sizes, adult fecundity could not be estimated and larval counts were considered to provide the best index of reproductive success between populations. The site characteristics tested were elevation, number of pools per site (taken from Garcia and Buley, 1997) and degree of isolation (calculated as mean horizontal distance to the five nearest toad populations).
3. Results 3.1. Distribution of toads and predators in relation to topography Out of the three variables entered into the forward stepwise logistic regression model, only maximum slope—which was significantly positively associated with toad presence—was included in the overall model (Table 1). The model was significant (w21=16.2; P < 0.001) and resulted in 76.1% correct predictions of toad occurrence. A further stepwise logistic regression on the same three topographical variables to predict the presence or absence of predators at toad sites found that only elevation—which was negatively associated with
Table 1 Topographical characteristics of torrent localities containing toad populations (toads present) and randomly allocated torrent localities (toads absent) and results of logistic regression to predict presence/absence of toads at each locality based on the same three topographical features Variable
( )a
Max slope Elevation (m)b Aspect ( )b
Toads present (n=17)
Toads absent (n=50)
Logistic regression based on toad presence
Mean
S.D.
Mean
S.D.
b
R
P
69.6 313.7 149.4
11.4 199.9 34.1
46.1 135.2 135.2
18.2 59.7 59.7
0.07 – –
0.35 0.14 0.00
<0.001 >0.05 >0.05
df=1 for each analysis. S.D., standard deviation; b standardised regression coefficient; R, unstandardised regression coefficient; P, level of significance. a Variable included in final model. b Variables excluded from final model.
330
R.D. Moore et al. / Biological Conservation 116 (2004) 327–332
for toads. Although A. muletensis is an excellent climber and can negotiate vertical rock faces, it has been suggested that steep sides may make sites inaccessible to introduced predators (Tonge, 1986; Schley and Griffiths, 1998). However, this theory was not supported by the lack of a significant negative association between the presence of predators and maximum slope. An alternative explanation is that the steepness of the immediate terrain influences the microclimate of a site; steep limestone rock faces surrounding a breeding pool may provide a more shaded and cooler environment that favours longer hydroperiods and provides optimal conditions for growth and development of tadpoles. The mean preferred temperature and optimum temperatures for development of A. muletensis tadpoles are 21.6 C (Martens, 1984) and 21–24 C (Kadel and Hemmer, 1984) respectively, both of which are low relative to other European amphibians including A. obstetricans (Martens, 1984). Pools in open areas are more likely to be exposed to the sun, making them prone to heating and desiccation during the summer. Because of the low optimum temperature for the growth and development of tadpoles—and since the distribution of the toad is likely to be constrained by water availability—environments favouring the persistence of cool plunge pools at the base of ephemeral waterfalls are most likely to favour the presence of the toad. Additionally, steep
predator presence—was included in the model (Table 2). The overall model was significant (w21=6.4; P < 0.05) and resulted in 71.4% of correct predictions of predator occurrence. 3.2. Factors influencing reproductive success Multiple linear regression showed that only the number of pools at each site was a significant predictor of overall reproductive success of a population: this variable showed a strong positive association with the mean number of tadpoles counted in all pools at each site over 1991–2000 (Table 3). When reproductive success per pool per population was regressed against the same three independent variables, elevation emerged as a significant predictor, and displayed a positive relationship with the dependent variable (Table 3).
4. Discussion Of the three topographical variables included in the analysis of toad distribution, only maximum slope was included in the final model. Toad presence was strongly positively associated with this variable, indicating that the immediate relief surrounding a torrent may be important in dictating its suitability as a breeding site
Table 2 Topographical characteristics of toad populations where predatory snakes and frogs are known to occur (predators present) and those where predators have never been seen (predators absent) and results of logistic regression to predict presence/absence of predators based on the same three topographical features Variable
a
Elevation (m) Max slope ( )b Aspect ( )b
Predators present (n=7)
Predators absent (n=14)
Logistic regression based on predator presence
Mean
S.D.
Mean
S.D.
b
211.5 62.7 141.8
126.4 14.7 26.5
437.8 68.4 156.0
218.8 12.2 62.3
– –
R 0.01
P 0.27 0.00 0.00
<0.05 >0.05 >0.05
df=1 for each analysis. S.D., standard deviation; Beta, standardised regression coefficient; R, unstandardised regression coefficient; P, level of significance. a Variable included in final model. b Variables excluded from final model.
Table 3 Results of multiple linear regression of A. muletensis reproductive success (log means of annual counts 1999–2000) against elevation, number of pools and degree of isolation (mean horizontal distance to the five nearest populations) Variable
Reproductive success per population Standardized Coefficient (b)/Partial Correlation
Elevation No. pools Isolation
0.37b 0.66a 0.05b
Reproductive success per pool per population t
P
1.69 16.67 0.30
>0.05 <0.001 >0.05
Significance of overall models: reproductive success per population, R2=0.43; F1 R2=0.20; F1 19=4.7; P< 0.05. a Variable included in final model. b Variables excluded from final model.
Standardized Coefficient (b)/Partial Correlation 0.42a 0.23b 0.01b 19=14.4;
t
P
2.16 1.02 0.04
<0.05 >0.05 >0.05
P=0.001; reproductive success per pool per population,
R.D. Moore et al. / Biological Conservation 116 (2004) 327–332
sides surrounding a site may provide more suitable habitat for adults, which often take refuge in humid cracks and crevices within the walls of the torrent gorges (personal observation). Within the geographical range of A. muletensis, the distribution of two introduced predators N. maura and R. perezi was virtually identical. This suggests that snake populations may require the presence of R. perezi to sustain high population numbers and that the presence of A. muletensis alone is not sufficient. One study conducted on mainland Spain found that N. maura preyed heavily upon adults and tadpoles of R. perezi (Santos et al., 2000), and the presence of the frog may be a significant factor in the successful establishment and continued existence of the snake on Mallorca. Moreover, there was a strong negative association between predator presence and elevation at toad sites. Although both the frog and the snake were found together at elevations as high as 1000 m (personal observation), it is likely that the environmental and climatic conditions associated with lower elevations are preferred by both predators. N. maura is restricted to lower elevations in the northern part of its range; in Italy it is constrained to elevations below 700 m and in the Iberian Peninsula, where it reaches its highest elevation, it is found up to 1700 m (Gasc et al., 1997). Although R. perezi is known to occur up to elevations of 1500 m in the Pyrenees and over 2000 m in parts of Spain (Gasc et al., 1997), it may be constrained by a lack of suitable habitat at higher altitudes on Mallorca. A reduction in predation pressure and competition with increasing elevation may be an important factor in the persistence of toad populations in montane torrent gorges and an important consideration in the design of future reintroduction sites. The distribution of predators may additionally be influenced by human activity, and the creation of artificial reservoirs in the vicinity of natural torrent habitats may increase the chance of these areas being colonised by predators (Mayol, personal communication). Data are lacking on the ranges of N. maura and R. perezi on Mallorca outside localities occupied by A. muletensis, and this precludes a more comprehensive analysis of biotic and abiotic factors influencing their distributions. The viability of isolated populations of a species typically depends upon factors such as population size and degree of inter-population connectivity (Hanski, 1999), which in turn are a function of both physical distance and the characteristics of intervening habitat (Forman, 1995). Of the three site characteristics considered—elevation, isolation and the number of pools within a site—reproductive success of Mallorcan midwife toad populations was associated most strongly with the number of pools. This is an important finding because it indicates that the viability of a population may be greatly enhanced by increasing the number of
331
potential breeding pools within a gorge. Adult toads are capable of travelling considerable distances from breeding pools (a distance of 222 m has been recorded; Roma´n, 2000) and it is highly likely that torrent gorges provide ideal corridors for the movement of adults. Individuals seem more likely to move between pools within the same torrent gorge than to traverse open areas and mountain tops to reach new potential breeding sites, so the number of pools within a torrent gorge is likely to be an important factor determining the viability of a population. When the number of tadpoles per pool was regressed against the same three independent variables, the latter showed a significant positive association with elevation. There are two possible explanations for this finding. First, environmental and climatic conditions associated with high elevations may enhance the viability of toad populations, for the microclimatic reasons described above. Secondly, because the presence of introduced predators is negatively associated with altitude, populations of tadpoles and adult toads may suffer higher rates of predation at lower elevations. The results of this study have a number of implications for the conservation programme of the Mallorcan midwife toad. As the optimal habitat for the species is high altitude torrent gorges containing many pools and surrounded by steep sides, these factors should be considered in the design and location of reintroduction sites for the toad. Additionally, the viability of toad populations may be significantly enhanced by the creation of more pools within the vicinity of breeding populations, since overall reproductive success appears to be strongly associated with the number of pools available. More data are required on the distributions of N. maura and R. perezi on Mallorca to allow a better understanding of the factors influencing the distributions of these predators.
Acknowledgements This work was carried out with the co-operation of La Conselleria de Medi Ambient, who also supplied digitised maps of the Serra de Tramuntana. We thank D. Jay, V. Mun˜oz, C. O’Brien, C. Zayas and R. Barber for logistical support in the field. J. Mayol, T. Beebee and B. Davis provided helpful comments on earlier drafts of the manuscript. The work was supported by the Natural Environment Research Council.
References Alcover, J.A., Moya`-Sola`, S., Pons-Moya`, J., 1981. Les Quimeres del passat. Els vertebrats fossils del Plio-Quaternari de les Balears i pitiuses. Editorial Moll Palma de Mallorca.
332
R.D. Moore et al. / Biological Conservation 116 (2004) 327–332
Alcover, J.A., Sanders, E., Sanchiz, B., 1984. El registro fo´sil de los sapos parteros (Anura, Discoglossidae) de Balears. In: Hemmer, H., Alcover J.A. (Eds.), Historia biologica del ferreret. Editorial Moll Palma de Mallorca, pp. 109–122. Bloxam, Q.M.C., Tonge, S.J., 1995. Amphibians—suitable candidates for breeding-release programs. Biodiversity and Conservation 4, 636–644. Burbidge, A.A., Manly, B.F.J., 2002. Mammal extinctions on Australian islands: causes and conservation implications. Journal of Biogeography 29, 465–473. Case, T.J., Bolger, D.T., 1991. The role of introduced species in shaping the distribution and abundance of island reptiles. Evolutionary Ecology 5, 272–290. Corsi, F., Dupre`, E., Boitani, L., 1999. A large scale model of wolf distribution in Italy for conservation planning. Conservation Biology 13, 150–159. Detmers, R., Bart, J., 1999. A GIS modelling method applied to predicting forest songbird habitat. Ecological Applications 9, 152–163. Diamond, J.M., 1989. Overview of recent extinctions. In: Western, D., Pearl, M. (Eds.), Conservation for the Twenty-first Century. Oxford University Press, New York, pp. 27–41. Dowding, J.E., Murphy, E.C., 2001. The impact of predation by introduced mammals on endemic shorebirds in New Zealand: a conservation perspective. Biological Conservation 99, 47–64. Farina, A., 1997. Landscape structure and breeding bird distribution in a sub-Mediterranean agro-ecosystem. Landscape Ecology 12, 365–378. Forman, R.T.T., 1995. Some general-principles of landscape and regional ecology. Landscape Ecology 10, 133–142. Fritts, T.H., Rodda, G.H., 1998. The role of introduced species in the degradation of island ecosystems: a case history of Guam. Annual Review of Ecological Systematics 29, 113–140. Garcia, G., Buley, K., 1997. Report on the Mallorcan Fieldwork 1997. Jersey Wildlife Preservation Trust (JWPT) Internal Report. Gasc, J., Cabela, A., Crnobrnja-Isailovic, J., Dolmen, D., Grossenbacher, K., Haffner, P., Lescure, J., Martens, H., Martı´nez Rica, J.P., Maurin, H., Oliveira, M.E., Sofianidou, T.S., Veith, M., Zuiderwijk, A., 1997. Atlas of Amphibians and Reptiles of Europe. Societas Europea Herpetologica. Muse´um National D’Histoire Naturelle, Paris. Gaston, A.J., 1994. Status of the ancient murrelet, Synthliboramphusantiquus, in Canada and the effects of introduced predators. Canadian Field Naturalist 108, 211–222. Hall, J.R., Woods, R.W., Brooke, M.D., Hilton, G.M., 2002. Factors affecting the distribution of landbirds on the Falkland Islands. Bird Conservation International 12, 151–167. Hanski, I., 1999. Habitat connectivity, habitat continuity, and metapopulations in dynamic landscapes. Oikos 87, 209–219. Hosmer, D.W., Lemeshow, S., 1989. Applied Logistic Regression. John Wiley and Sons, New York. Howell, C.A., Latta, S.C., Donovan, T.M., Porneluzi, P.A., Parks, G.R., Faaborg, J., 2000. Landscape effects mediate breeding bird abundance in Midwestern forests. Landscape Ecology 15, 547–562. Jackson, D.B., 2001. Experimental removal of introduced hedgehogs improves wader nest success in the Western Isles, Scotland. Journal of Applied Ecology 38, 802–812. Janet, F., 1998. Predicting the distribution of shrub species in southern California from climate and terrain-derived variables. Journal of Vegetation Science 9, 733–748. Jokima¨ki, J., 1999. Occurrence of breeding bird species in urban parks: effects of park structure and broad-scale variables. Urban Ecosystems 3, 21–24. Kadel, K., Hemmer, H., 1984. Temperature dependence of larval development in the Mallorcan midwife toad, Baleaphryne muletensis. In: Hemmer, H., Alcover, J.A. (Eds.), Historia Biologica del Ferreret. Editorial Moll Palma de Mallorca, pp. 169–174. Kiesecker, J.M., Blaustein, A.R., 1997. Population differences in responses of red-legged frogs (Rana aurora) to introduced bullfrogs. Ecology 78, 1752–1760. Knutson, M.G., Sauer, J.R., Olsen, D.A., Mossman, M.J., Hemesath,
L.M., Lannoo, M.J., 1999. Effects of landscape composition and wetland fragmentation on frog and toad abundance and species richness in Iowa and Wisconsin, U. S.A. Conservation Biology 13, 1437–1446. Lenton, S.M., Fa, J.E., Perez del Val, J., 2000. A simple non-parametric GIS model for predicting species distribution: endemic birds in Bioko Island, West Africa. Biodiversity and Conservation 9, 869–885. Martens, H., 1984. Temperature selection in tadpoles of Baleaphryne muletensis. In: Hemmer, H., Alcover, J.A. (Eds.), Historia Biologica del Ferreret. Editorial Moll Palma de Mallorca, pp. 163–168. Martinez-Morales, M.A., Cuaron, A.D., 1999. Boa constrictor, an introduced predator threatening the endemic fauna on Cozumel Island, Mexico. Biodiveristy and Conservation 8, 957–993. McCullagh, P., Nelder, J.A., 1983. Generalized Linear Models. Chapman and Hall, London. Mun˜oz, V., 2000. Assessing Monitoring Methods for the Mallorcan Midwife Toad, Alytes muletensis. MSc Dissertation, University of Kent, UK. Rodda, G.H., Fritts, T.H., 1992. The impact of the introduction of the colubrid snake Boiga irregularis on Guam lizards. Journal of Herpetology 26, 166–174. Rodda, G.H., Sawai, Y., Chiszar, D., Tanaka, H. (Eds.), 1999. Problem Snake Management: the Habu and the Brown Treesnake. Cornell University Press, Ithaca and London. Roma´n, A., 2000. Estudi dels mecanismes dispersius del ferreret (Alytes muletensis). Memo`ria de beca de formacio´ per a la investigacio´ a la zona 5.B. Conselleria de Treball i Formacio´. Govern Balear. Roma´n, A., 2001. Report on the Fourth International Meeting on the Alytes muletensis Conservation. 22–23 February 2001. Roma´n, A., 2002. Recompte de les poblacions naturals i introduı¨ des de ferreret (Alytes muletensis) en totes les localitats (2001). Govern Balear. Roma´n, A., Mayol, J., 1997. Documents te`cnics de conservacio´. La recuperacio´n del ferreret, Alytes muletensis. Proyecto life ferreret, Palma de Mallorca.. Santos, X., Gonzalez-Solis, J., Llorente, G.A., 2000. Variation in the diet of the viperine snake Natrix maura in relation to prey availability. Ecography 23, 185–192. Savidge, J.A., 1987. Extinction of an island forest avifauna by an introduced snake. Ecology 68, 660–668. Schley, L., Griffiths, R.A., 1998. Midwife toads (Alytes muletensis) avoid chemical cues from snakes (Natrix maura). Journal of Herpetology 32, 572–574. Schley, L., Griffiths, R.A., Roma´n, A., 1998. Activity patterns and microhabitat selection of the Mallorcan midwife toad (Alytes muletensis) tadpoles in natural torrent pools. Amphibia-Reptilia 19, 143–151. Scribner, K.T., Arntzen, J.W., Cruddace, N., Oldham, R.S., Burke, T., 2001. Environmental correlates of toad abundance and population genetic diversity. Biological Conservation 98, 201–210. Short, J., Kinnear, J.E., Robley, A., 2002. Surplus killing by introduced predators in Australia—evidence for ineffective anti-predator adaptations in native prey species? Biological Conservation 103, 283–301. Spence, D., 2002. A Comparison of Fitness in Founder and Descendent Stock of Mallorcan Midwife Toad (Alytes muletensis) Tadpoles. MSc Dissertation, University of Kent, UK. Stocks, A.M., Heywood, D.I., 1994. Terrain modelling for mountains In: Price, M.F., Heywood, D.I. (Eds.), Mountain Environments and Geographical Information Systems. pp. 25–40. Tabachnick, B.G., Fidell, L.S., 1983. Using Multivariate Statistics. Harper and Row, NY. Tonge, S.J., 1986. Collecting the Mallorcan midwife toad. Oryx 20, 74–78. Towns, D.R., Daugherty, C.H., Cree, A., 2001. Raising the prospects for a forgotten fauna: a review of 10 years of conservation effort for New Zealand reptiles. Biological Conservation 99, 3–16. Wilson, P.R., Karl, B.J., Toft, R.J., Beggs, J.R., Taylor, R.H., 1998. The role of introduced predators and competitors in the decline of kaka (Nestor meridionalis) populations in New Zealand. Biological Conservation 83, 175–185.
Distribution of the Mallorcan midwife toad (Alytes muletensis) in relation to landscape topography and introduced predators Robin D. Moorea, Richard A. Griffithsa,*, Alvaro Roma´nb b
a The Durrell Institute of Conservation and Ecology, University of Kent, Canterbury, Kent CT2 7NS, UK Associacio´ per a la Recuperacio´ del Ferreret, Mateu E. Llado´ 34B. 1 C 07002, Palma de Mallorca, Balearic Islands, Spain
Received 18 January 2003; received in revised form 14 April 2003; accepted 24 April 2003
Abstract The endemic midwife toad of Mallorca (Alytes muletensis) is restricted to a small number of breeding populations in the mountainous northwest of the island. The decline of the species has been attributed to the impacts of introduced species such as the viperine snake (Natrix maura) and green frog (Rana perezi), and toads may be surviving only in areas that are suboptimal for these predators. The influence of landscape features (elevation, aspect and maximum slope) on the distribution of toads and associated predators was therefore investigated using GIS. The presence of toads was positively associated with steep slopes. At sites where they occurred with toads, the distribution of predators was negatively associated with elevation. Reproductive success within toad populations was strongly associated with the number of pools at each site, while reproductive success within individual pools was positively associated with elevation. These findings may be used to optimise the design and location of future reintroduction sites. # 2003 Elsevier Ltd. All rights reserved. Keywords: Endangered species; GIS; Introduced predator
1. Introduction According to Diamond’s (1989) ‘evil quartet’, the four principal causes of species’ extinctions are habitat fragmentation, overexploitation, introduced predators and chains of extinctions. Species on islands that have evolved in the absence of any major predators or competitors are likely to be particularly vulnerable to the introduction of predators. Periods of coevolution between predator and prey play an important role in reducing dynamic responses of one species to the other (Case and Bolger, 1991). However, novel predators can severely impact populations of prey species because there has been no opportunity for coevolution (e.g. Savidge, 1987; Rodda and Fritts, 1992; Keisecker and Blaustein, 1997; Fritts and Rodda, 1998; Short et al., 2002). There are many cases of introduced species negatively impacting native island fauna (e.g. Gaston, 1994; Wilson et al., 1998; Dowding and Murphy, 2001; Jackson, 2001; Towns et al., 2001; Burbridge and * Corresponding author. Tel.: +44-1227-823434; fax: +44-1227827839. E-mail address: r.a.griffi[email protected] (R.A. Griffiths). 0006-3207/03/$ - see front matter # 2003 Elsevier Ltd. All rights reserved. doi:10.1016/S0006-3207(03)00202-7
Manly, 2002; Hall et al., 2002), although most studies to date have focused on introduced mammals and there are few documented cases of reptile introductions. Exceptions include the introduced boa (Boa constrictor), believed to threaten endemic fauna on Cozumel Island in Mexico (Martinez-Morales and Cuaron, 1999) and the brown tree snake (Boiga irregularis), introduced accidentally to Guam in the 1950s and responsible for the extinction of many species of native fauna (Savidge, 1987; Rodda and Fritts, 1992; Fritts and Rodda, 1998; Rodda et al., 1999). The endemic midwife toad of Mallorca (Alytes muletensis) was once widespread over the island (Alcover et al., 1984), but today persists only as a small number of isolated breeding populations. These breeding populations are largely confined to the limestone gorges of the Serra de Tramuntana mountains in the northwest of the island, where populations may use one or more pools scoured by ephemeral torrents. The decline of the toad has been attributed by many workers to predation and competition from species such as the viperine snake (Natrix maura) and green frog (Rana perezi) (Tonge, 1986; Bloxam and Tonge, 1995), both of which were introduced to Mallorca around 2000 years ago. Both of
328
R.D. Moore et al. / Biological Conservation 116 (2004) 327–332
these species consume tadpole and adult stages of A. muletensis (personal observations); in addition, it is possible that interspecific competition occurs between the tadpole stages of A. muletensis and R. perezi (Spence, 2002). While some endemic fauna went extinct following the introduction of these and other predators during Roman times (Alcover et al., 1981; Tonge, 1986), A. muletensis was one of the few species to survive, although the reasons for this remain unclear. One hypothesis proposed by some authors is that toad populations have persisted in gorges with steep sides, because introduced predators find it difficult to colonise and establish populations in such areas (Bloxam and Tonge, 1995; Schley and Griffiths, 1998). A better understanding of the factors influencing the distribution of the toad and its associated predators is required before any firm conclusions can be reached. Recent advances in geographical information systems (GIS) and remote sensing technologies have facilitated the analysis of animal and plant distributions in relation to spatial environmental attributes (e.g. Farina, 1997; Corsi et al., 1999; Knutson et al., 1999; Howell et al., 2000; Lenton et al., 2000). Relatively simple environmental descriptors such as altitude, slope and aspect have been used to explain distribution and abundance patterns of taxa such as birds (e.g. Farina, 1997; Detmers and Bart, 1999) and plants (Janet, 1998), and it is likely that these variables strongly influence the microclimate of a site (Stocks and Heywood, 1994). Knowledge of how such environmental features influence the distribution of rare species and associated predators may be an important consideration in the management of the species. This study used GIS to analyse the distribution of Mallorcan midwife toads and their principle predators in relation to landscape topography. We therefore tested the hypothesis that toad populations have persisted in sites that are suboptimal for predators. In addition, we determined which site characteristics were the best predictors of reproductive success within toad populations.
2. Methods 2.1. Census methods Annual population surveys carried out at all known sites since 1991 provided information on the status of A. muletensis (Garcia and Buley, 1997; Roma´n and Mayol, 1997; Roma´n, 2001, 2002). As adults are cryptic and spend most of their lives in cracks and fissures underground, annual censuses focused on counts of tadpoles in the breeding pools. These counts were made in July each year, immediately following spring breeding activity. While tadpoles may take anything from several weeks to over a year to complete development, tadpole
numbers were thought to be at their highest during these census periods. Most breeding pools are clear and devoid of vegetation, so it is relatively easy to make direct counts of individual tadpoles (Schley et al., 1998) and counts made by different observers are quite consistent (Mun˜oz, 2000). Mean number of tadpoles counted in each pool at each site from 1991 to 2000 was taken as a measure of reproductive success of each population over this period. However, as the number of breeding pools used by a population of toads at each site varies between 1 and 19, reproductive success was expressed in two ways. First, reproductive success per population was calculated as the mean number of tadpoles based on the total tadpole count in all pools at a site. Second, reproductive success per pool per population was calculated by dividing the mean number of tadpoles counted per population by the number of pools used by toads in that population. During the annual censuses the presence or absence of Natrix maura and Rana perezi was noted. Populations were classified as having ‘predators present’ if predators had been observed within 100 m of the population on at least two separate visits. 2.2. Mapping procedure Scanned paper maps of the Serra de Tramuntana on 1:5000 scale were obtained from the Conselleria de Medi Ambient in June 2000. Torrents and 25 m contours were digitised from these scanned maps using Arcview 3.2 across the entire geographical range of A. muletensis. From these contours a digital terrain model (DTM) was constructed using Triangulated Irregular Networks (TIN). DTMs such as this serve as an important tool for deriving secondary spatial data sets such as altitude, slope and aspect (Stocks and Heywood, 1994). Slope and aspect surfaces with a grid size of 0.01 km were generated from the TIN for analysis. The locations of all toad populations were mapped using information already available on the geographic location of all known populations in the form of GPS readings and paper maps (Garcia and Buley, 1997). To provide a comparison with torrent habitats not occupied by toad populations, 50 points were selected along torrents using a random points generator extension for ArcView 3.2 and by combining all torrents so that the localities were randomly located throughout the range of the toad. 2.3. Statistical analysis Stepwise multiple logistic regression (Hosmer and Lemeshow, 1989) was used to test a number of topographical factors considered to be potentially important in influencing the distribution of the midwife toad and of the introduced predators N. maura and R. perezi. The distribution of the two predators amongst toad sites was
329
R.D. Moore et al. / Biological Conservation 116 (2004) 327–332
perfectly correlated (r=1.0) indicating that wherever N. maura are present, R. perezi also occurs nearby and vice versa. Logistic regression was considered the most appropriate analytical procedure for assessing the distribution of toads and predators because it permits the prediction of binary attributes such as presence/absence (McCullagh and Nelder, 1983) and it has proved to be a powerful technique for analysing species distributions (e.g. Jokima¨ki, 1999). The dependent or response variable in the regression—presence or absence of toads—was scored as 1 (all torrent-dwelling toad populations; n=17) or 0 (randomly generated points along torrents; n=50), respectively. Because all the random points were generated along torrents, any toad populations that inhabited artificial water bodies outside natural torrents were excluded from the analysis. The independent or explanatory variables were three topographical features associated with each locality derived from surfaces constructed using GIS; elevation (m), aspect (in degrees, where 0 =true north and 180 =true south) and maximum slope within a horizontal radius of 100 m (equal to an area of 31,416 m2). Maximum slope—rather than mean slope—was chosen as an independent variable as it was thought that just one steep slope in the vicinity of breeding site may be limiting for predators, in that it would prevent emigration of predators from the site. A radius of 100 m was chosen because this was considered to provide an appropriate representation of the immediate topography at each site. Scribner et al. (2001), for example, also characterised habitat within a 100 m radius around each pond when investigating environmental correlates of common toad (Bufo bufo) abundance and genetic diversity. The number of independent variables in the analysis was kept low due to the limited number of sites being compared; this is in accordance with the suggested minimum requirement of at least four to five times more cases than independent variables (Tabachnick and Fidell, 1983). Using the stepwise logistic regression procedure, selection of variables for inclusion in the final model was based on the Wald test (Hosmer and Lemershow, 1989; Jokima¨ki,
1999). All variables were normally distributed and therefore met the assumptions of the analysis. Factors influencing the distribution of predators were investigated by comparing landscape topography between all mapped toad populations (17 inhabiting natural torrent pools and four inhabiting artificial water bodies). Seven toad populations with predators were compared against 14 predator-free toad populations and predator presence or absence (scored as 1 and 0 respectively) was regressed against the three independent topographical variables described above. In order to assess the factors influencing reproductive success of A. muletensis over the period 1991–2000, stepwise linear regressions were made for all toad populations of (1) mean tadpole count per population, and (2) mean tadpole count per pool per population against three site characteristics (tadpole counts were log-transformed to ensure normality). In the absence of data on adult population sizes, adult fecundity could not be estimated and larval counts were considered to provide the best index of reproductive success between populations. The site characteristics tested were elevation, number of pools per site (taken from Garcia and Buley, 1997) and degree of isolation (calculated as mean horizontal distance to the five nearest toad populations).
3. Results 3.1. Distribution of toads and predators in relation to topography Out of the three variables entered into the forward stepwise logistic regression model, only maximum slope—which was significantly positively associated with toad presence—was included in the overall model (Table 1). The model was significant (w21=16.2; P < 0.001) and resulted in 76.1% correct predictions of toad occurrence. A further stepwise logistic regression on the same three topographical variables to predict the presence or absence of predators at toad sites found that only elevation—which was negatively associated with
Table 1 Topographical characteristics of torrent localities containing toad populations (toads present) and randomly allocated torrent localities (toads absent) and results of logistic regression to predict presence/absence of toads at each locality based on the same three topographical features Variable
( )a
Max slope Elevation (m)b Aspect ( )b
Toads present (n=17)
Toads absent (n=50)
Logistic regression based on toad presence
Mean
S.D.
Mean
S.D.
b
R
P
69.6 313.7 149.4
11.4 199.9 34.1
46.1 135.2 135.2
18.2 59.7 59.7
0.07 – –
0.35 0.14 0.00
<0.001 >0.05 >0.05
df=1 for each analysis. S.D., standard deviation; b standardised regression coefficient; R, unstandardised regression coefficient; P, level of significance. a Variable included in final model. b Variables excluded from final model.
330
R.D. Moore et al. / Biological Conservation 116 (2004) 327–332
for toads. Although A. muletensis is an excellent climber and can negotiate vertical rock faces, it has been suggested that steep sides may make sites inaccessible to introduced predators (Tonge, 1986; Schley and Griffiths, 1998). However, this theory was not supported by the lack of a significant negative association between the presence of predators and maximum slope. An alternative explanation is that the steepness of the immediate terrain influences the microclimate of a site; steep limestone rock faces surrounding a breeding pool may provide a more shaded and cooler environment that favours longer hydroperiods and provides optimal conditions for growth and development of tadpoles. The mean preferred temperature and optimum temperatures for development of A. muletensis tadpoles are 21.6 C (Martens, 1984) and 21–24 C (Kadel and Hemmer, 1984) respectively, both of which are low relative to other European amphibians including A. obstetricans (Martens, 1984). Pools in open areas are more likely to be exposed to the sun, making them prone to heating and desiccation during the summer. Because of the low optimum temperature for the growth and development of tadpoles—and since the distribution of the toad is likely to be constrained by water availability—environments favouring the persistence of cool plunge pools at the base of ephemeral waterfalls are most likely to favour the presence of the toad. Additionally, steep
predator presence—was included in the model (Table 2). The overall model was significant (w21=6.4; P < 0.05) and resulted in 71.4% of correct predictions of predator occurrence. 3.2. Factors influencing reproductive success Multiple linear regression showed that only the number of pools at each site was a significant predictor of overall reproductive success of a population: this variable showed a strong positive association with the mean number of tadpoles counted in all pools at each site over 1991–2000 (Table 3). When reproductive success per pool per population was regressed against the same three independent variables, elevation emerged as a significant predictor, and displayed a positive relationship with the dependent variable (Table 3).
4. Discussion Of the three topographical variables included in the analysis of toad distribution, only maximum slope was included in the final model. Toad presence was strongly positively associated with this variable, indicating that the immediate relief surrounding a torrent may be important in dictating its suitability as a breeding site
Table 2 Topographical characteristics of toad populations where predatory snakes and frogs are known to occur (predators present) and those where predators have never been seen (predators absent) and results of logistic regression to predict presence/absence of predators based on the same three topographical features Variable
a
Elevation (m) Max slope ( )b Aspect ( )b
Predators present (n=7)
Predators absent (n=14)
Logistic regression based on predator presence
Mean
S.D.
Mean
S.D.
b
211.5 62.7 141.8
126.4 14.7 26.5
437.8 68.4 156.0
218.8 12.2 62.3
– –
R 0.01
P 0.27 0.00 0.00
<0.05 >0.05 >0.05
df=1 for each analysis. S.D., standard deviation; Beta, standardised regression coefficient; R, unstandardised regression coefficient; P, level of significance. a Variable included in final model. b Variables excluded from final model.
Table 3 Results of multiple linear regression of A. muletensis reproductive success (log means of annual counts 1999–2000) against elevation, number of pools and degree of isolation (mean horizontal distance to the five nearest populations) Variable
Reproductive success per population Standardized Coefficient (b)/Partial Correlation
Elevation No. pools Isolation
0.37b 0.66a 0.05b
Reproductive success per pool per population t
P
1.69 16.67 0.30
>0.05 <0.001 >0.05
Significance of overall models: reproductive success per population, R2=0.43; F1 R2=0.20; F1 19=4.7; P< 0.05. a Variable included in final model. b Variables excluded from final model.
Standardized Coefficient (b)/Partial Correlation 0.42a 0.23b 0.01b 19=14.4;
t
P
2.16 1.02 0.04
<0.05 >0.05 >0.05
P=0.001; reproductive success per pool per population,
R.D. Moore et al. / Biological Conservation 116 (2004) 327–332
sides surrounding a site may provide more suitable habitat for adults, which often take refuge in humid cracks and crevices within the walls of the torrent gorges (personal observation). Within the geographical range of A. muletensis, the distribution of two introduced predators N. maura and R. perezi was virtually identical. This suggests that snake populations may require the presence of R. perezi to sustain high population numbers and that the presence of A. muletensis alone is not sufficient. One study conducted on mainland Spain found that N. maura preyed heavily upon adults and tadpoles of R. perezi (Santos et al., 2000), and the presence of the frog may be a significant factor in the successful establishment and continued existence of the snake on Mallorca. Moreover, there was a strong negative association between predator presence and elevation at toad sites. Although both the frog and the snake were found together at elevations as high as 1000 m (personal observation), it is likely that the environmental and climatic conditions associated with lower elevations are preferred by both predators. N. maura is restricted to lower elevations in the northern part of its range; in Italy it is constrained to elevations below 700 m and in the Iberian Peninsula, where it reaches its highest elevation, it is found up to 1700 m (Gasc et al., 1997). Although R. perezi is known to occur up to elevations of 1500 m in the Pyrenees and over 2000 m in parts of Spain (Gasc et al., 1997), it may be constrained by a lack of suitable habitat at higher altitudes on Mallorca. A reduction in predation pressure and competition with increasing elevation may be an important factor in the persistence of toad populations in montane torrent gorges and an important consideration in the design of future reintroduction sites. The distribution of predators may additionally be influenced by human activity, and the creation of artificial reservoirs in the vicinity of natural torrent habitats may increase the chance of these areas being colonised by predators (Mayol, personal communication). Data are lacking on the ranges of N. maura and R. perezi on Mallorca outside localities occupied by A. muletensis, and this precludes a more comprehensive analysis of biotic and abiotic factors influencing their distributions. The viability of isolated populations of a species typically depends upon factors such as population size and degree of inter-population connectivity (Hanski, 1999), which in turn are a function of both physical distance and the characteristics of intervening habitat (Forman, 1995). Of the three site characteristics considered—elevation, isolation and the number of pools within a site—reproductive success of Mallorcan midwife toad populations was associated most strongly with the number of pools. This is an important finding because it indicates that the viability of a population may be greatly enhanced by increasing the number of
331
potential breeding pools within a gorge. Adult toads are capable of travelling considerable distances from breeding pools (a distance of 222 m has been recorded; Roma´n, 2000) and it is highly likely that torrent gorges provide ideal corridors for the movement of adults. Individuals seem more likely to move between pools within the same torrent gorge than to traverse open areas and mountain tops to reach new potential breeding sites, so the number of pools within a torrent gorge is likely to be an important factor determining the viability of a population. When the number of tadpoles per pool was regressed against the same three independent variables, the latter showed a significant positive association with elevation. There are two possible explanations for this finding. First, environmental and climatic conditions associated with high elevations may enhance the viability of toad populations, for the microclimatic reasons described above. Secondly, because the presence of introduced predators is negatively associated with altitude, populations of tadpoles and adult toads may suffer higher rates of predation at lower elevations. The results of this study have a number of implications for the conservation programme of the Mallorcan midwife toad. As the optimal habitat for the species is high altitude torrent gorges containing many pools and surrounded by steep sides, these factors should be considered in the design and location of reintroduction sites for the toad. Additionally, the viability of toad populations may be significantly enhanced by the creation of more pools within the vicinity of breeding populations, since overall reproductive success appears to be strongly associated with the number of pools available. More data are required on the distributions of N. maura and R. perezi on Mallorca to allow a better understanding of the factors influencing the distributions of these predators.
Acknowledgements This work was carried out with the co-operation of La Conselleria de Medi Ambient, who also supplied digitised maps of the Serra de Tramuntana. We thank D. Jay, V. Mun˜oz, C. O’Brien, C. Zayas and R. Barber for logistical support in the field. J. Mayol, T. Beebee and B. Davis provided helpful comments on earlier drafts of the manuscript. The work was supported by the Natural Environment Research Council.
References Alcover, J.A., Moya`-Sola`, S., Pons-Moya`, J., 1981. Les Quimeres del passat. Els vertebrats fossils del Plio-Quaternari de les Balears i pitiuses. Editorial Moll Palma de Mallorca.
332
R.D. Moore et al. / Biological Conservation 116 (2004) 327–332
Alcover, J.A., Sanders, E., Sanchiz, B., 1984. El registro fo´sil de los sapos parteros (Anura, Discoglossidae) de Balears. In: Hemmer, H., Alcover J.A. (Eds.), Historia biologica del ferreret. Editorial Moll Palma de Mallorca, pp. 109–122. Bloxam, Q.M.C., Tonge, S.J., 1995. Amphibians—suitable candidates for breeding-release programs. Biodiversity and Conservation 4, 636–644. Burbidge, A.A., Manly, B.F.J., 2002. Mammal extinctions on Australian islands: causes and conservation implications. Journal of Biogeography 29, 465–473. Case, T.J., Bolger, D.T., 1991. The role of introduced species in shaping the distribution and abundance of island reptiles. Evolutionary Ecology 5, 272–290. Corsi, F., Dupre`, E., Boitani, L., 1999. A large scale model of wolf distribution in Italy for conservation planning. Conservation Biology 13, 150–159. Detmers, R., Bart, J., 1999. A GIS modelling method applied to predicting forest songbird habitat. Ecological Applications 9, 152–163. Diamond, J.M., 1989. Overview of recent extinctions. In: Western, D., Pearl, M. (Eds.), Conservation for the Twenty-first Century. Oxford University Press, New York, pp. 27–41. Dowding, J.E., Murphy, E.C., 2001. The impact of predation by introduced mammals on endemic shorebirds in New Zealand: a conservation perspective. Biological Conservation 99, 47–64. Farina, A., 1997. Landscape structure and breeding bird distribution in a sub-Mediterranean agro-ecosystem. Landscape Ecology 12, 365–378. Forman, R.T.T., 1995. Some general-principles of landscape and regional ecology. Landscape Ecology 10, 133–142. Fritts, T.H., Rodda, G.H., 1998. The role of introduced species in the degradation of island ecosystems: a case history of Guam. Annual Review of Ecological Systematics 29, 113–140. Garcia, G., Buley, K., 1997. Report on the Mallorcan Fieldwork 1997. Jersey Wildlife Preservation Trust (JWPT) Internal Report. Gasc, J., Cabela, A., Crnobrnja-Isailovic, J., Dolmen, D., Grossenbacher, K., Haffner, P., Lescure, J., Martens, H., Martı´nez Rica, J.P., Maurin, H., Oliveira, M.E., Sofianidou, T.S., Veith, M., Zuiderwijk, A., 1997. Atlas of Amphibians and Reptiles of Europe. Societas Europea Herpetologica. Muse´um National D’Histoire Naturelle, Paris. Gaston, A.J., 1994. Status of the ancient murrelet, Synthliboramphusantiquus, in Canada and the effects of introduced predators. Canadian Field Naturalist 108, 211–222. Hall, J.R., Woods, R.W., Brooke, M.D., Hilton, G.M., 2002. Factors affecting the distribution of landbirds on the Falkland Islands. Bird Conservation International 12, 151–167. Hanski, I., 1999. Habitat connectivity, habitat continuity, and metapopulations in dynamic landscapes. Oikos 87, 209–219. Hosmer, D.W., Lemeshow, S., 1989. Applied Logistic Regression. John Wiley and Sons, New York. Howell, C.A., Latta, S.C., Donovan, T.M., Porneluzi, P.A., Parks, G.R., Faaborg, J., 2000. Landscape effects mediate breeding bird abundance in Midwestern forests. Landscape Ecology 15, 547–562. Jackson, D.B., 2001. Experimental removal of introduced hedgehogs improves wader nest success in the Western Isles, Scotland. Journal of Applied Ecology 38, 802–812. Janet, F., 1998. Predicting the distribution of shrub species in southern California from climate and terrain-derived variables. Journal of Vegetation Science 9, 733–748. Jokima¨ki, J., 1999. Occurrence of breeding bird species in urban parks: effects of park structure and broad-scale variables. Urban Ecosystems 3, 21–24. Kadel, K., Hemmer, H., 1984. Temperature dependence of larval development in the Mallorcan midwife toad, Baleaphryne muletensis. In: Hemmer, H., Alcover, J.A. (Eds.), Historia Biologica del Ferreret. Editorial Moll Palma de Mallorca, pp. 169–174. Kiesecker, J.M., Blaustein, A.R., 1997. Population differences in responses of red-legged frogs (Rana aurora) to introduced bullfrogs. Ecology 78, 1752–1760. Knutson, M.G., Sauer, J.R., Olsen, D.A., Mossman, M.J., Hemesath,
L.M., Lannoo, M.J., 1999. Effects of landscape composition and wetland fragmentation on frog and toad abundance and species richness in Iowa and Wisconsin, U. S.A. Conservation Biology 13, 1437–1446. Lenton, S.M., Fa, J.E., Perez del Val, J., 2000. A simple non-parametric GIS model for predicting species distribution: endemic birds in Bioko Island, West Africa. Biodiversity and Conservation 9, 869–885. Martens, H., 1984. Temperature selection in tadpoles of Baleaphryne muletensis. In: Hemmer, H., Alcover, J.A. (Eds.), Historia Biologica del Ferreret. Editorial Moll Palma de Mallorca, pp. 163–168. Martinez-Morales, M.A., Cuaron, A.D., 1999. Boa constrictor, an introduced predator threatening the endemic fauna on Cozumel Island, Mexico. Biodiveristy and Conservation 8, 957–993. McCullagh, P., Nelder, J.A., 1983. Generalized Linear Models. Chapman and Hall, London. Mun˜oz, V., 2000. Assessing Monitoring Methods for the Mallorcan Midwife Toad, Alytes muletensis. MSc Dissertation, University of Kent, UK. Rodda, G.H., Fritts, T.H., 1992. The impact of the introduction of the colubrid snake Boiga irregularis on Guam lizards. Journal of Herpetology 26, 166–174. Rodda, G.H., Sawai, Y., Chiszar, D., Tanaka, H. (Eds.), 1999. Problem Snake Management: the Habu and the Brown Treesnake. Cornell University Press, Ithaca and London. Roma´n, A., 2000. Estudi dels mecanismes dispersius del ferreret (Alytes muletensis). Memo`ria de beca de formacio´ per a la investigacio´ a la zona 5.B. Conselleria de Treball i Formacio´. Govern Balear. Roma´n, A., 2001. Report on the Fourth International Meeting on the Alytes muletensis Conservation. 22–23 February 2001. Roma´n, A., 2002. Recompte de les poblacions naturals i introduı¨ des de ferreret (Alytes muletensis) en totes les localitats (2001). Govern Balear. Roma´n, A., Mayol, J., 1997. Documents te`cnics de conservacio´. La recuperacio´n del ferreret, Alytes muletensis. Proyecto life ferreret, Palma de Mallorca.. Santos, X., Gonzalez-Solis, J., Llorente, G.A., 2000. Variation in the diet of the viperine snake Natrix maura in relation to prey availability. Ecography 23, 185–192. Savidge, J.A., 1987. Extinction of an island forest avifauna by an introduced snake. Ecology 68, 660–668. Schley, L., Griffiths, R.A., 1998. Midwife toads (Alytes muletensis) avoid chemical cues from snakes (Natrix maura). Journal of Herpetology 32, 572–574. Schley, L., Griffiths, R.A., Roma´n, A., 1998. Activity patterns and microhabitat selection of the Mallorcan midwife toad (Alytes muletensis) tadpoles in natural torrent pools. Amphibia-Reptilia 19, 143–151. Scribner, K.T., Arntzen, J.W., Cruddace, N., Oldham, R.S., Burke, T., 2001. Environmental correlates of toad abundance and population genetic diversity. Biological Conservation 98, 201–210. Short, J., Kinnear, J.E., Robley, A., 2002. Surplus killing by introduced predators in Australia—evidence for ineffective anti-predator adaptations in native prey species? Biological Conservation 103, 283–301. Spence, D., 2002. A Comparison of Fitness in Founder and Descendent Stock of Mallorcan Midwife Toad (Alytes muletensis) Tadpoles. MSc Dissertation, University of Kent, UK. Stocks, A.M., Heywood, D.I., 1994. Terrain modelling for mountains In: Price, M.F., Heywood, D.I. (Eds.), Mountain Environments and Geographical Information Systems. pp. 25–40. Tabachnick, B.G., Fidell, L.S., 1983. Using Multivariate Statistics. Harper and Row, NY. Tonge, S.J., 1986. Collecting the Mallorcan midwife toad. Oryx 20, 74–78. Towns, D.R., Daugherty, C.H., Cree, A., 2001. Raising the prospects for a forgotten fauna: a review of 10 years of conservation effort for New Zealand reptiles. Biological Conservation 99, 3–16. Wilson, P.R., Karl, B.J., Toft, R.J., Beggs, J.R., Taylor, R.H., 1998. The role of introduced predators and competitors in the decline of kaka (Nestor meridionalis) populations in New Zealand. Biological Conservation 83, 175–185.