habitat selection and thermal ecology of the sympatric lizards podarcis ...

1

-

HERPETOLOGICAL JOURNAL: V;1.5,

pp. 181-188 (1995)

HABITAT SELECTION AND THERMAL ECOLOGY OF THE SYMPATRIC LIZARDS PODARCIS MURALIS AND PODARCIS HISPANICA IN A MOUNTAIN REGION OF CENTRAL SPAIN

De~artamentode Estadistica v Matemtitica Aplicadas, Universidad de Salamanca. C/ E s ~ e i o2. E-3 7007 Sala manca, SFlain a Animal, Universia'ad de Subamanca. k rtamento Ide BiologiF alamanca, Spain Presei

:

Departarmento de Ciencias

?S y

Recurros Naturales. Universidad dc? Alicante,, Spain

We studied habitat selection, field body tem]peratures (:TB) and activity cycles in sympatric populations of the lacertid lizards Podarcis murnlis and P odarcis hispanica in a mountain area - . in Central Spain. Different patterns of habitat selection for the two species were found: P. murali;s occupiesI N and N'JV facing tallus sites, vvhile P. hispanica shlows a lessI habitat-slpecific distribr tio on at tlle study site. Inter- and intrasipecific di:fferences :in TBs wt:re found, while thermoregulator]r precisio~ 1 analyses showed dlifferencesi between adult ma1e and fern ale P. --. murali.s. Hypotheses baseci on activity cycles,, habitat selection a g behaviolur are discussied to explain such Iresults in 1two closel:y related sipecies.

,.

Thermoregulatory behavic)ur and habitat selection

re intimately relate:d and str,ongly influence life his-C 1:---. )ry traits and fitness- U I ~laards (Ouboter, 1981; Huey, 982; van Berkum, Huey & Adams, 1986; Adolph, 990). Many authors have approached the study of in:ractions between the conditions of the habitat ccupied by a given species and its thermal behaviour Jillman, 1969; Moermond, 1979; Roughgarden, Por:r & Heckel, 198 1; Hertz & Huey, 1981; Tracy & Ihristian, 1986; van Damme, Bauwens, Castilla & 'erheyen, 1989; Marquet, Ortiz, Bozinovic & J aksic, 989), as well as focusing on the connection between rracv., tnermoregulation and activity time (Porter & I..._, 1983; Grant & Dunham, 1988). We here report a study of body temperatures, activity cycles and features of habitat selection in the lizards Podarcis muralis and Podarcis hispanica, that are sympatric in some areas of the Iberian Peninsula, including the Sierra de Guadarrama (Central System, Spain), where this study was undertaken. The analysis of thennoregulatory precision, activity cycles and basking behaviour have been developed through field observations, and the conclusions extracted are therefore mainly directed towards suggesting working hypotheses for- liiture studies. Our maill objectives were to: ( I ) quantify habitat occupation hy the two species in the area, and identify any intra- arid interspecific differences between them; (2) determine some aspects of their thermal ecology, and relatc !Iicse to the habitat selection features: and

..

spect whlether dift , De rouna between the two sp

11 ecolog

AND ME Lizards of thce lacertid genus Po darcis fialm the Ibe rian 1'eninsula seem to be very similar in basi . . -. .. ecological characteristics such as diet composltlor foraging behaviour, activity cycles and thermoregula tory behaviour (Arnold & Burton, 1978 PCrez-Mellado, 1983a,b; Arnold, 1987; Gosh, 1987,. These similarities are especially strong between P. muraliis and P. hispanica,two species so closely related that th --,ere are some problems for their systematic distinction over a wide part of their distribution range (e.g. Martinez-Rica & Laplaza, 1989). These two small lizards are similar in size (adult snout-vent length 49.7-67.6 mm and 46.7-66.8 mm respectively, personal data), morphology (Perez-Mellado & Galindo, 1986) and habitat requirements (Arnold, 1987). Podarcis muralis occupies a wide range of latitudes in Europe -the southwestern extreme of which is the Sierra de Guadarrama-, while P. hispanica is lirnited to the Iberian Peninsula, southern France and northern Africa. STUDY ARE,\

The study area is located in the westernmost part of the Sierra de Guadarrama, Spain, a mountain range reaching a maximum altitude of 1902 m. The altitudinal range surveyed extended from 1450 to 1902

182

J. M A R T ~ N - V A L L E J OETA L.

m. The dominant vegetation consists of Pinus sylvestris forest, widely spread throughout the whole mountain range. In the high mountain areas, the forest was gradually replaced by small shrubs, with Jzrniperus commztnis subsp. r~anabeing the dominant species. Below 1550 m, the forest becomes very open, and Quercus ilex subsp. ballota is the predominant tree species. Important habitat differences occurred resulting from the topography: the southern slopes of the mountains present a greater number of open habitats with Mediterranean shrubs such as Cistus laurifolius, Halymium viscosum, Erica arborea and Thymus bracteatus. The northern slopes exhibit closed Pinus forests with a few accompanying scrubs (mainly Arctostaphyllos uva-ursi) and have much higher humidity. METHODS

Observations were made at monthly intervals from May 1988 to September 1989. We captured lizards by hand or noose, and recorded the following data upon each capture: species, date, time, age, sex, reproductive condition of females, behaviour (basking or active), orientation of slope (N, NW, W, SW, S, SE, E, NE), altitude, body temperature (cloaca1 = TB), air temperature (shaded bulb, 10 cm above substrate = TA), and substrate temperature (shaded bulb = TS). Temperatures were measured with a Schultheis thermometer to the nearest 0.1 "C. Three age classes were considered: juveniles, the individuals born in the year of capture; subadults, oneyear-old individuals, which have not yet reached sexual maturity; and !adults, sexually mature individuI als. j Activity rhythms were determined from the frequencies of lizard o~servationsmade in the different months and time 'intervals (data from 1988 and 1989 were pooled since no differences were detected). Correction indices (Telleria, 1986) were used to avoid possible biases caused by an irregular survey intensity. The microhabitat selection field study was developed between April and August 1989, at monthly intervals. A 5 x 5 m square was used as the survey unit, the centre of which was a lizard previously observed (these observations were made randomly, using two hour intervals and considering the fust lizard seen after each of these periods of time). Within each square we estimated 28 variables: (1-5) percentage cover at ground level of gravel, grass, fallen leaves, rock and soil; (6-10) percentage cover of vegetation layers with heights 2 m; (1 1- 15) percentage cover of walls, and rocks with heights lo0 cm; (16) percentage cover of talus structures; (17-1 8) percentage cover of dead tree trunks and branches; (19-20) percentage cover of 0-20" and 20-40" slopes; (21-28) SW, S, SE, E, NE, N, NW or W dominant orientations. Percentages were calculated as means of estimates made by eye by two independent observers. Similarly -

-

we estimated the availability of microhabitats in the environment by measuring the same variables in squares for which the centres were the points occupied by us after stopping every two minutes during random directed walks.

We used one-way analysis of variance (ANOVA) and Bonferroni's multiple comparison test (see Keppel, 1991) on arcsin x' transformed data to examine differences between samples (P. muralis, P. hispanica and availability) for each microhabitat variable. When variances were heterogeneous across samples (Bartlett's test) we used the non-parametric Kruskal-Wallis test, followed by Dunn's multiple comparison test (Dunn, 1964; see Hollander & Wolfe, 1973). Thermoregulatory precision was estimated by the slope of the least square regressions of TB on TA or TS. Differences between lizard species and agelsex groups were assessed by analyses of covariance (ANCOVA). As another indicator of thennoregulatory precision we employed the interquartile range. To analyse differences between two means we used the two-tailed Student t-test, with the Welch approximation (tw) in case of heterogeneity of variances, or a 2-test for n>50. Frequency distributions of distinct behavioural categories were compared with x2 tests. The significance level used was a= 0.05. RESULTS HABITAT SELECTION

Podarcis muralis occupied a narrower altitudinal range in the study area (1490-1725 m; Fig. l), while P. hispanica was present along the whole surveyed range (1450-1900 m).

FIG. I. Absolute frequencies (on horizontal axis) of Podarcis muralis (left hand) and Podarcis hispanica (right hand) lizards seen at the different altitude intervals (on

vertical axis). .

.

- !

-

HABITAT SELECTION IN PODARCIS TABLE 1. Average percentages and standard deviations for the three groups considered: Availability, P. muralis and P. hispanica.

TABLE 2. Results obtained in the habitat selection study. (* P The regression slopes or gravla ana non-gravid females were analysed, and differences have not been detected between the two groups for either of the two species. Y V L

I

RUClAJ.

< 25cn 25-50~ 50-100cm > lOOc Talus Slope:

10.3 4.9 L ~ L

17.3. 15.0 11.3 2.4 2.7 7.8

12.6 14.6 5,7 7. 21

19.2 2.7 0.5 0 0.0 .8 30.2

'.2 l.2

46.5 31.3 53.4 31.3

ance and Bonferroni tests, In t he~ analyse L. p G r c G 1itages , obtained in the made Lwirn rrnt: microhabitat selection study (Table l), P. muralis exhibited significantly higher percentages for the variables soil, talus and 20-40" slope, and lower ones for grass and the 0-20" slope. Both P. muralis and P. hispanica showed significantly higher percentages than available for rocks under 25 cm (Table 2). In the group of variables analysed through nonparametric tests, significant results were found for the following variables: trees (>2 m high vegetation cover) and U IS);

1.50 TS); I,ooled dat;

Europe , where I? muralhi occurs s~ympatricallywith other 12~certidspecies, e.g ., in Gree:ce, with Podarcis s ., .. ,-. -* , " erllarari (StrljDoscn, Helmer & scohlte, 1989; and pers. obs.), and in the Balkans area with Lacerta horvathi (Amold, 1987). In both cases, P. muralis restricts its microhabitat to soil dwellings - in shaded zones, while the other slpecies occupies sulnny and rocky microhabitats. Hence, our results woulclJ suppurr me: LWV ~~ypotheses mentioned above, the second of which needs to be verified with an experimental design consisting of displacements of one species from the area (see, for instance, Nevo, Gorman, Soule, Yung Yang, Clover & Jovanovic, 1972). However, the restricted mounlainous distribution of P. murulis in the Iberian Peninsula nientioned above could favour thc acceptance of the first hypothesis of physiological constraints, indicating a residual situation of this rock lizard population at the very edge of the species' geographical range (PerezMellado & Galindo, 1986).

... .

-

2 .

J . M A R T ~ N - V A L L E J OET A L .

J

0

(A)

MARCH

APRIL

JULY

JUNE

AUGUST SEPTEMBER

_ I

JULY

INE

FIG. 3. AnnualI and daily activity rhythms of P. muralis (A and B:) and P. hi.spanica (CI and D). 'VVhite squares represt:nt adult males and black squares iadult females, triang!les stand fc3r subadullts.

. ...

RMAL ECC

'Ihe average TBs reported here for P. hispanica are

somewhat lower than those recorded at lower altitucdes (Pdr -..*ez-Mellado, 19836). These differences are Inot likely to be due to physiological changes, but rathe1r to a slightly less effective thermoregulation; or to phjrsical limitations derived ~. from the higher altitudes, that -- - : 773s thc:y exhibil woudd not lelt them ac:hieve thc lowcx sites (vim Damm~ e, Bauwel1s & Verhleyen, 199 ~

...,TORS

.-

-- .-

.-.

INFLUENCING I~HERMOREGULATORYYRECI

J IN MALE

5 Mum LI.

between two species could . .. ,. due to dltterenCeS in their tnermal prererences or thermal physiology. In this particular situation, intraspecific differences in selected body or optimal temperatures have not been detected (Bauwens, pers. corn.). Therefore, differences in TJ3s should not be due to physiological differences. Hertz & Huey (1981), van Damme et al. (1989, 1990) point out three possible behavioural causes to explain a greater independence from the environmental conditions in some reptiles: habitat shifts; increases in basking intensity; and changes in activity times. However, the hypotheses concerning habitat shifts would be valid only for Podarcis muralis adult males, since females achieve a thermoregulation pattern similar to Podarcis hispanica, i.e., greater dependence on

.

.

-

.

.

lenta~remlperatures. If we stilI1 would c a amerential micronaoitat. use as a hypotnesl!.s to explain the interspecific variations we ought tc verify whether males and females use space differently. We n have not detected intraspecific differen,,,0 - n ., microhablitat selection in eitl~ e of r the two species (see results). ...C*L , , " " : L I , ,C" +, , ...-*:---.a The ~ a.-.~. ~ U UL I IUG U p U 3 J I U I V l a b L U 1 3 IUGIILIUIIGU by van Damme et al. (1 990) is an iincrease in basking intensity. Significant intra- or interspecific significant -- -r differences in the fiequencies or various types of behaviour have not been detected, therefore this hypothesis would not be v alid eithc:r to expllain the thermoregulatory difference!S. . .. The third hypothesis explains tne airrerences WIUI changes in activity periods. P. muralis exhibits shorter activity periods than P. hispanica, which appears to be because the talus is less exposed to the sun (most talus areas are in shade (luring the first morning hornrs, from 8- 11 hr). This could also explain the 1thennoregylatory ., , .: U,E I Ir. D I I L I ~ I ~ L I ~ I UI hispanica, i.e., it --.. m a y "".. sadice a higher thermal control for longer activity times (Bowker, 1986). Another hypothesis is linked to food availability. Huey & Slatkin (1976) predict that increased habitat productivity should cause increased thermoregulatory benefits; in this sense, Lee (1980) affirms that in an environment with abundant trophic resources, lizards are expected to thermoregulate more precisely, since a

..

.

..--!

.

.-----em.

HABITAT SELECTION IN PODARCIS the time inversion in feeding behaviours decreases. The mean number of available prey items per day (Garcia-Fernandez, Martin-Vallejo & PBrez- Mellado, in prep.), an estimate of the relative abundance of prey in the environment, is lowest in habitats occupied by P. muralis, which contradicts this hypothesis. Finally, intraspecific differences in TBs found in Podarcis muralis could be due to differences in reproductive condition. More cryptic behaviour in females, related to the survivorship of themselves and their offspring could explain these differences, since they ould invest less time in thermoregulation. The h''gn' productive effort made by females of this speciies, ith very high clutch sizes (mode=8; unpub. dz~ta) -Id support this hypothesis. However, we found nn diff erences in TB-TA regression slopes between gravid and non-gravid femalces of P. nnuralis. We theref 1:rr----.... conclude that sexual ~IUG-IG-IIGG-s in thermoreeulatc precision in P. muralis are unlik:ely to be the resull: of possible behavioural modificatic)ns induu:d by the females reproductive condition. '1l e third :hypothesis mentiorled above (changes; i n i likewise: explain intraspecjific ) activity peric~ d s coulc diff erences in Podari:is muralis: adult males hiwe . . .per iter frequencies of activity at midday 3), wl - (Fig. . n with a :greater pr fem ales exhilbit more 1 encce at less th~ermallyfi hours. Hchwever, if --- *L-.---------IIIS~G-GL LIIG regresslull sca~iclpluis,wr; 11c)tice that points concerning these particular hours (with lo! TA and TS), show higher TBs again in males. T1 together with some occasional observations in fie11d, suggested a new hypothesis: adult ma Poalarcis muralis, because of their more aggressive i territorial behaviour (Boag, 1973; Mou, 19 Edsman, 1990), would exploit small sunny patche!;at early times, when the talus is almost completely in shade, so that in the moment of capture, they c auld hav e already achieved high TBs. Magnuson, Crower & Mecdvick (1979) and Edsman (1990) report the influencle of territoriality on the thermal ecology of lizard . species, in the sense that dominant animals occupy the best basking sites to achieve high temperatures as soon as possible. In our study area, subadults begin their annual cycle in May, with the highest frequency occurring in June, possibly to avoid interactions with adult males that would exhibit intense territorial behaviour in the months dedicated to reproductive behaviour. In summary, both species showed at the study site a similar thermal biology with only apparently slight interspecific differences linked to their different microhabitat selection and changes in activity periods. Microhabitat selection arises as a major factor in spccies segregation, but the observed pattern could be explained by several alternative hypotheses only testable within an experimental framework. 11"

UVU

-

.^^A

-~

.C-

--.A

187

ACKNOWLEDGEMENTS We are very grateful to Dirk Bauwens, Richard Brown and Aurora Castilla for their constructive comments on previous drafts of the manuscript. Purificaci6n Galindo provided useful advices on the statistical analyses, and several members of the Department of Animal Biology of the University of Salamanca h e l ~ e dus in many aspects of our research and prep,aration of 'the manuiscript. This study was supported b:Y two grants of theI Spanish CTCYT (projects PB 086-139 and PE190-0526-C02-0 1) ZNCES

.-.-

Adolph, Influence havioral .z---, rncrm.oregularion on micronaolrar use ~y two Scelol Dorus lizards. Ecology 71, 3 15-327. Arnold, E1. N. & Burton, J.A. (1978). Ajieldguide to the r L

I

Rrnti, ..,..les

and Amphibians of Britian and

f i g" - -, - "yc. u

Collins, Londor Arnold, E. N. (198 rce partitic>n among lacertid /Dl 1 170 1. I J 7 lizards in southcAslu u a u p r . 1 h7w- -w1s . Ul aV r-sAU . ("1 782. Barbadill0, L.J. (1 1987). L, a G u h de Incafo anjibiios y repitiles de It a Peninstcrra loerica, lsras Baleares y Canarias. Madrid, Incafco SA. Blasco, M. (1980). Contribuci6n a1 COInocimienbo de 10s lackrtidos de Andalucia. Monograj ias y trab,ajos del Dept. 2001. Univ. Mrilaga 1- 8 1. Boag, D. A. (1973). Spatial relationshil? among nnembers of a population of wall lizard. OecoIogia 12, 1-13 Bowker, R. G . (1986). Patterns of th~rnuvrr~ularrvr~ Podar.cis hispanica (Lac1ertilia: Lacertidae). Studies in Hesrpetology, 62 1-626. Rocek, Z. (Ed.). Prague: Char11 es Universitv. Brafia, F. (1984). Eliogeograj'ia, biolog ia y estructura de nicho,s de la tax ocenosis 6ie saurios de Asturias. Tesis Doct 1Univ Oviedo. Busack, S. D. (1977). Zoogeography of Amphibians and Reptiles in Chdiz Province, Spain. Ann. Carnegie Mus. Nut. Hist. 46 (17), 285-3 16. Dunn, 0 . J. (1964). Multipl e compariisons using range sums. Teclrnometrics 6,2!4 1-252. Edsman, L. (1 990). Terriforitdity and c,ompefition in wall I.'--..rrzurd~. PhD Thesis, n----*-ucparrrncnt of Zoology, University of Stockholm. Gosh, A. (1987). Las lagartijas del gCnero Podarcis en la costa del Pais Vasco (Vizcaya, Guiptlzcoaz y Lapurdi). Cuad. Secc. Cienc. Nut. (Aranzadi) 3, 3 3 3 -

.

Y.

._

r L I ~ . -

346.

Grant, B. W. & Dunham, A. E. (1988). Thermally imposed time constrains on the activity of the desert lizard Sceloporus mew-iami. Ecology 69, 167- 176. Ilcrtz, P. l i , I% kluey, 11. R. (1981). Compensation for altitudinal changes in the thermal environment by some Anolis lizards on Hispaniola. Ecology 62. 5 IS521.

ALLEJO ET AL.

-.

I-lillrnan. P. E. (1969). Ilahitat specificity in t h r c ~ syrnpatric specics of A~nciva (Reptilia: 'l'ciidae! Ecolo,yy S0(3), 470-48 1. I-ioll~rnde~-. bl. & Wolfe. D. il. (1973). Notl~~crrnrnetri, srorisiicai nzethods. John Wiley & Sons. New York. Mucy, R . U. (1982). Temperature, physiology, and the ecology of reptiles. Biology ofthe reptilia, vol 12, pp 25-91. Cans C., Pough 1'. H. (Eds). London, Ncw York: Academic Press. Huey, R. B. & Slatkin, M. (1976). Costs and benefits of lizard therrnoregulation. Quart. Rev. Biol. 51, 363384. Lee, J. C. (1980). Comparative thermal ecology of two lizards. Oecologia 44, 171- 176. Keppel, G. (1991). Design and analysis: A researcher's handbook., Prentice Hall. Magnuson, J. J. Crower., L. B. & Medvick, P. A. (1979). TemperatiIre as an ecological resource . Amer. Zbol. 19, 33 1-343. rquet, P. A. Ortiz J. C. Bozinovic F & Jaksic F. M. (1989). Ecological aspects of thermoregulation at high altitudes: the case of the Andean Liolaemus lizards in northern Chile. Oecologia 81, 16-20. Ma:rtinez-Rica, J. P. & Laplaza, E. (1989). Variaciones morfomttricas y de folidosis en poblaciones vecinnc de Podarcis hispanica y Podi ilis (Rept.ilia, ?v. Esp. H erp. Sauria), y su relacidn con la t 3, 159-172. the ermond, T. C. (1979). Habitat cons1 behavior, morphology and communitj e of Anolis lizards. Ecology 60, 152-164. u, Y. P. (1987). Ecolog,ie comparke de rieux populations de lezar*ddes mu).ailles, Podarcis mu]ralis (Laurenti, 1978), el9 France ..- Thtse de Doctora.t de I'Univ. de Paris VI. Neiro, E. Gonnan, G. Soule, M. Yung Yang,, S. Clove]r, R. & Jovancwic, V. (1972). Competitiive exclu;sion La --- :. :-. I@^..-:. --.:A I--\ UGLWGGII Insular Lacerta s ~ e c t JiU U ~I IU. ~.. 1LUVGI LlCJaGJ. Notes on Iexperimental introd1 10, 183-190. Oulboter, P. E. (1981). The ecology or me lslana nzard Podarcis sicula saljii: Correlation microdistribution with vegetation coverage, the1 environment and food-size. Amphibia-Reptilia 2, : 257. ez-Mellado, V. (1983~).Alimentacidn de dos espe:cies simpiltridas de saurios en el Sistema Cc:ntral. -St1udia Oecologica 4, 89- 114. Activity Ptrez-Mellado, V. (19836) and thermoregulation patterns in two species of --L--- . I Lacertidae: Podarcis hispanica (Steinda~nncr, I 670) and Podarcis bocagei (Seoane, 1884).- Cienc. Ecol. Syst. (Portugal) 5, 5-12. rerez-Mellado, V. & Galindo, M. P. (1986). Sistematica de Podarcis (Sauria Lacertidae) ibiricas y norteafricanas mediante ticnicas multivariantes. Edic. Univ. Salamanca. 1-163. 4

..a-0

4--.

Porter, W. P. & Tracy, C. R. (1 983). Biophysical anai!>c:, of encrsetics, time-spacc utilization, ;in3 distributional limits. In Lizard ecology. Stlcdics 0 , ~(; r~tndcir!rgunisn~,pp. 55-83. I l ~ i c yR. D. , I'ianka t