Intraspecific Competition and High Food Availability Are Associated with Insular Gigantism in a Lizard
Total Page:16
File Type:pdf, Size:1020Kb
Naturwissenschaften (2009) 96:1107–1113 DOI 10.1007/s00114-009-0564-3 SHORT COMMUNICATION Intraspecific competition and high food availability are associated with insular gigantism in a lizard Panayiotis Pafilis & Shai Meiri & Johannes Foufopoulos & Efstratios Valakos Received: 27 February 2009 /Revised: 13 May 2009 /Accepted: 14 May 2009 /Published online: 2 June 2009 # Springer-Verlag 2009 Abstract Resource availability, competition, and predation stronger intraspecific competition; and (3) such aggression, commonly drive body size evolution. We assess the impact will be associated with larger average body size. We found of high food availability and the consequent increased a positive correlation between seabird and lizard densities. intraspecific competition, as expressed by tail injuries and Cannibalism and tail injuries were considerably higher in cannibalism, on body size in Skyros wall lizards (Podarcis dense populations. Increases in cannibalism and tail loss gaigeae). Lizard populations on islets surrounding Skyros were associated with large body sizes. Adult cannibalism (Aegean Sea) all have fewer predators and competitors than on juveniles may select for rapid growth, fuelled by high on Skyros but differ in the numbers of nesting seabirds. We food abundance, setting thus the stage for the evolution of predicted the following: (1) the presence of breeding gigantism. seabirds (providing nutrients) will increase lizard popula- tion densities; (2) dense lizard populations will experience Keywords Podarcis gaigeae . Insularity . Cannibalism . Seabird subsidy. Population density : P. Pafilis J. Foufopoulos Introduction School of Natural Resources & Environment, University of Michigan, Dana Building, 440 East University, Interpopulation variations in body size have been reported Ann Arbor, MI 48109-1041, USA for numerous taxa (Meiri et al. 2005;Wuetal.2006), J. Foufopoulos especially on islands (Sondaar 1977;Lomolino2005). e-mail: [email protected] Evolution of both dwarfism and gigantism has been attributed to the particularities of island environments P. Pafilis (*) Department of Classical Studies, Modern Greek Program, such as relaxed predation regime, low food availability, University of Michigan, and strong competition (Meiri et al. 2004; Dayan and 2160 Angell Hall, 435 S. State, Simberloff 2005;Meiri2008). Ann Arbor, MI 48109-1041, USA Density-dependent intraspecific competition is thought e-mail: [email protected] to strongly impact population dynamics, as well as multiple S. Meiri aspects of an organism’s life history (Lack 1954; Sinervo et NERC Centre for Population Biology, Imperial College London, al. 2000). On islands, low predation pressure often results Silwood Park Campus, in higher densities and thus increased intraspecific compe- Ascot, Berkshire SL5 7PY, UK e-mail: [email protected] tition. Juvenile mortality, through potential infanticide, is also related to higher levels of intraspecific competition E. Valakos (Jenssen et al. 1989; Elgar and Crespi 1992). Population Department of Biology, University of Athens, densities and body sizes of many island vertebrates are Panepistimiopolis, 157-84 Athens, Greece correlated to the presence of seabirds (Sanchez-Piñero and e-mail: [email protected] Polis 2000; Bonnet et al. 2002) that, through accumulation 1108 Naturwissenschaften (2009) 96:1107–1113 of fecal material, food scraps, and carcasses boost nutrient to lizard tails, 20 mm from the cloaca, for 15 s and recorded availability in islands (Polis and Hurd 1996). whether this resulted in autotomy. In this study, we assessed the impact of such “seabird We estimated autotomy rates by determining the fraction subsidies,” population density, and intraspecific competi- of animals with regenerated tails in the field and among tion on body size of an endemic Mediterranean lizard museum specimens. Whereas simulated predation trials showing extreme interpopulation body size variation represent the intrinsic ability for tail shedding, field (Valakos et al. 2008). The study populations all live under autotomy rates reflect the combined influence of intrinsic the same climatic conditions but differ considerably in tendencies and the external stimuli to do so. The relative population density, levels of intraspecific competition, and difference between these two variables, expressed as the the size of yellow-legged gull (Larus michahellis) colonies percent to which laboratory autotomy rates differ from field on the islands they inhabit. We tested three predictions: (1) rates, lizard population density is positively correlated to seabird presence; (2) higher population density is associated with RDFLAR ¼ ½ÂðÞLaboratory rates À Field rates =Field rates 100 elevated levels of intraspecific competition (measured as tail autotomy and cannibalism rates); and (3) higher levels reflects the intensity of environmental opportunities (pre- of intraspecific competition and greater gull abundance dation and social interactions) to cause autotomy (Pafilis et select for larger body size. al. 2009). Measurements Materials and methods We examined 218 preserved specimens at the Alexander Study organism and site Koenig Zoological Museum (Bonn) and 471 live field- caught animals. SVLs were measured using digital cal- The Skyros wall lizard (Podarcis gaigeae, Sauria: Lacertidae) lipers. No statistical difference was found between SVLs is a small [snout-vent length (SVL) ∼60 mm; mass, 6.6 g] of specimens measured in museums and in the field insectivorous lacertid endemic to the Skyros Archipelago (three-way analysis of variance with sex, museum/field, (38º51′ N, 24º33′ E; central Aegean Sea, Greece). We and island as factors: sex, F1,616=468.1, P<0.05; compare lizard populations living on Skyros Island and four museum/field, F1,616=0.27, P>0.05, island: F4,616= of its satellite islets. Lizards were collected in mid-July 378.0, P<0.05). That was also the case for autotomy rates of three consecutive years (2005, 2006, and 2007). To (χ2 test with Yate’s correction, df=1, all χ2≤2.5, P>0.05). avoid pseudoreplication and monitor population dynam- We therefore pooled museum and field measurements ics, we toe-clipped every lizard we caught and measured (Table 1). lizards only in the first year they were caught. Because We used mean population SVL as a measure of size the islands vary widely in levels of predation, competi- rather than maximum SVL, which can mask intraspecific tion, and presence of seabird colonies while experiencing variation and may underestimate true potential size (Meiri the same climate, they provide a unique opportunity to 2007). That said, population mean and maximum SVLs test the effects of the former three factors on lizard body were highly correlated with each other (juveniles, r=0.983; size (Table 1). females, r=0.974; males, r=0.969; n=5 in all cases), and we thus believe that different indices of size distributions Intraspecific competition will yield similar results. Population densities (lizards/hectare) were calculated by Intensity of intraspecific competition was assessed by the same observer (PP) using a single (some islands are too measuring rates of caudal autotomy and cannibalism. small for multiple transects), randomly placed, line transect Cannibalism was quantified by dissecting museum speci- of 4-m width and 100–200-m length in each island (see mens and examining the digestive track for lizard body Jaeger 1994). parts. Lizards shed their tail primarily to avoid predation and also to escape attacks by conspecifics (Arnold 1988). Statistical analysis We quantified ease of autotomy by simulating predation in the laboratory (Pérez-Mellado et al. 1997). Lizards were We used χ2 tests to compare the frequencies of tail placed in a terrarium with cork substrate to provide traction. autotomy and simple linear regression to estimate the To reduce pressure variation and standardize the experi- relationships between the ecological conditions on differ- mental conditions, we used digital callipers to apply ent islands and lizard density, cannibalism, autotomy comparable pressure (by comparing callipers’ indications) frequency, and body size. Naturwissenschaften (2009) 96:1107–1113 1109 Table 1 Geographic origin, snout vent length (in mm), autotomy, cannibalism percentages, and population density (individuals ha−1) of the animals used in this study Location Skyros Valaxa Mesa Diavates Lakonissi Exo Diavates Traits SVL males (museum and field pooled) 61.36 (185) 64.55 (33) 66.12 (46) 71.95 (64) 85.28 (69) SVL females (museum and field pooled) 55.65 (101) 57.07 (19) 57.98 (25) 62.27 (25) 70.34 (43) SVL juveniles (Field) 28.14 (37) 28.77 (7) 28.97 (7) 30.82 (10) 32.52 (18) Autotomy (museum and field pooled) 32.16% (92/286) 32.7% (17/52) 29.6% (21/71) 70.78% (63/89) 88.4% (99/112) Autotomy (Lab) 35.13% (39/111) 35% (7/20) 31.25% (10/32) 33.33% (6/18) 40% (4/10) RDFLAR 9.19 7.03 5.57 −52.95 −54.75 Cannibalism 1.20% (1/83) 0% (0/12) 0% (0/11) 4.54% (1/22) 21.42% (3/14) Densities 185 95 110 350 850 Species richness 4 (1, 2, 3, 4) 4 (1, 2, 3, 4) 2 (2, 4) 2 (2, 4) 2 (2, 4) Snake predators 3 (1, 2, 3) 1 (3) 0 0 0 Bird predators 6 (1, 2, 3, 4, 5, 6) 1 (3) 1 (3) 0 1 (3) Gull presence 0 5 8 10 50 Area (km2) 207 4.33 0.0384 0.016 0.019 Distance from Skyros (m) 0 140 306 666 1.400 Supporting references: Handrinos and Akriotis 1997; Valakos et al. 2008. Numbers in parentheses are sample sizes unless the identity of species is given. Snakes: 1, Zamenis situlus;2,Elaphe quatuorlineata,3Telescopus fallax. Birds: 1, Buteo buteo;2,Falco tinnunculus;3,Falco eleonorae;4, Athene noctua;5,Lanius senator;6,Corvus sp. Lizards: 1, Lacerta trilineata;2,Cyrtopodion kotschyi;3,Hemidactylus turcicus;4, Podarcis gaigeae While a multiple regression would have enabled us to difference disappeared when Lakonissi and Exo Diavates distinguish between different hypotheses regarding the were removed (χ2=0.199, df=2, P=0.91). Similarly, there ecological drivers of these four factors, we feel that, with were no differences between field and laboratory autotomy only five islands, a multiple regression model will be rates for most populations (all χ2≤0.01, P>0.05) with the grossly over-parameterized.