POLISH JOURNAL OF ECOLOGY 56 2 289–297 2008 (Pol. J. Ecol.) Regular research paper Katarina LJUBISAVLJEVIĆ1, Georg DŽUKIĆ1, Miloš L. KALEZIĆ1,2 1 Department of Evolutionary Biology, Institute for Biological Research “Siniša Stanković”, Bulevar Despota Stefana 142, 11060 Belgrade, Serbia e-mails: [email protected]; [email protected] 2 Institute of Zoology, Faculty of Biology, Studentski trg 16, 11000 Belgrade, Serbia e-mail: [email protected] FEMALE REPRODUCTIVE LIFE HISTORY TRAITS OF THE MEADOW LIZARD, DAREVSKIA PRATICOLA EVERSMANN, 1834 SQUAMATA: LACERTIDAE FROM THE WESTERNMOST BOUNDARY OF THE SPECIES RANGE ABSTRACT: We presented the first data on and Porter 1993, Du et al. 2005, Ji and female reproductive traits of the meadow lizard Wang 2005). The meadow lizard (Darevskia Darevskia praticola from the westernmost bound- praticola Eversmann, 1834) (Fig. 1) appears ary of the species range (Peridanubian Serbia). to be a particularly suitable object for this Mating occurs during April and May, oviposition kind of study on the basis of its two in many period occurs in June, and hatching takes place in respects different characteristics. Firstly, as July. Females mature at body sizes of 49.5 mm of snout-vent length (SVL). One clutch is produced the meadow lizard is a species of the forest annually. The average clutch size of 5.7 eggs (range lizards group of Darevskia genus (Arribas 4 – 8) represents the largest recorded for this spe- 1999), the knowledge of reproductive life cies. Clutch size and clutch mass increased signif- history traits of this species is of general in- icantly with the mother’s SVL. There was no evi- terest, considering the paucity of such stud- dence of the predicted trade-off between egg size ies dealing with small forest lacertids (e.g. and clutch size, as well as of variation in egg size Orlova 1969). Secondly, the meadow lizard associated with maternal SVL. The relative clutch provides a specific opportunity to study the mass (RCM) was rather high (0.60) for lacertid possible interrelation between complex dis- species and was not correlated with female size. junctive distribution patterns with different evolutionary lineages and differences in life KEY WORDS: oviposition period, clutch history traits. Namely, the current taxonomi- characteristics, Darevskia praticola, Serbia cal structure of the meadow lizard includes structuring at the subspecies level (e.g. 1. INTRODUCTION Darevsky 1997, Ljubisavljević et al. 2006, but see Arnold et al. 2007). A nomi- Recent studies of reptiles have accentuat- notypical subspecies (D. p. praticola) occurs ed the necessity of having independent data in northeastern Caucasus and Transcauca- sets from different populations and reptilian sus, while the D. p. pontica is distributed in taxa in order to understand both the general northwestern Caucasus and in southeastern patterns of covariation among life-history Europe. In southeastern Europe the meadow traits, and the relationships between these lizard inhabits southwestern and southeast- and environmental factors (e.g. Adolph ern Romania, Serbia, Bulgaria and Thrace in jjournalournal 114.indb4.indb 228989 22008-06-21008-06-21 112:17:362:17:36 290 Katarina Ljubisavljević et al. Fig. 1. Female Darevskia praticola from Vršački breg, Serbia. Turkey and Greece (Arnold and Ovenden 2. MATERIAL AND METHODS 2002). There it presents a patchy distribution connected to broad-leaved woodlands ex- 2.1. Localities and sampling posed to the influence of Submediterranean climate, restricted mainly to middle altitudes The analysis of female reproductive traits rarely exceeding 600 m (Stugren 1984). was carried out on samples collected from To date, there are several studies which three different localities in Peridanubian have yielded information on reproduction Serbia in 2002 and 2003: Vršački breg (340 of the meadow lizard in the eastern portion m above sea level, 45°08’N, 21°21’E); Avala of its range (Zhukov 1941, Orlova 1969, (250 m, 44°41’N, 20°31’E); Trešnja (250 m, Mushelishvili 1970, Orlova and Ter- 44°36’N, 20°34’E). tyshnikov 1979). Reproductive life history Because the preliminary analysis showed traits of southeastern European populations, no statistical differences in female snout-vent apart from the limited descriptive data for the length (SVL), clutch size and clutch and egg Romanian population (Méhely 1895, Fuhn characteristics between localities (ANOVA and Vancea 1961), are poorly known. for female SVL and ANCOVA with SVL as Here, we present the first basic study the covariate, F 2,7 = 0.05 – 16.74, P> 0.05 for on female reproductive characteristics of all variables), we pooled the data in order to the meadow lizard from the westernmost achieve a reasonable sample size. In the study boundary of its distribution range. Knowl- area, the meadow lizard occurs in hilly ter- edge of the reproductive potential of these rain in and around the open broad-leaved edge populations with a severely fragmented woods of the Hungarian and Turkey oak for- distribution is valuable for their conserva- est association Quercetum-frainetto-cerris tion and protection. (Radovanović 1951, Matvejev 1961). jjournalournal 114.indb4.indb 229090 22008-06-21008-06-21 112:17:362:17:36 Life history traits of Darevskia praticola 291 2.2. Data collection in the laboratory laboratory and additional oviductal clutches from five autopsied females. These individu- In the first half of June of 2002 and 2003, als were part of a larger sample of 41 females twelve late-pregnant females were collected which were examined for estimation of size from the study area and transported to the at sexual maturity based on the smallest in- laboratory. Individual females were housed dividual containing vitellogenic follicles or in individual terraria (30 × 20 cm) with a oviductal eggs. The mean maternal SVL was substrate of sand, while stones, pieces of calculated on 28 specimens from this sample bark and leaf litter were provided as shel- determined to be mature. The sample came tering sites. Food (mealworms and other in- from Georg Džukić’s herpetological collec- sects collected in the field) and water were tion of the Institute for Biological Research, available ad libitum. The lizards were main- Belgrade. tained under natural light and photoperiod We used multiple regression, analyses of conditions. The females were inspected al- variance (ANOVA) and covariance (ANCO- most every hour during daytime and once VA) to analyse the corresponding data. The during the night. Following oviposition, the analyses were carried out using the computer females were measured (snout-vent length, package Statistica® (STATISTICA for Win- SVL) and weighed. The eggs were dug up dows. StatSoft, Inc., Tulsa, OK). and carefully removed from the terraria, weighed and measured (maximum length 3. RESULTS and width), and their viability judged by the external characteristics of the eggshell. A 3.1. Sexual maturity digital calliper (0.01 mm precision) was used for the linear measurements, while mass Ovaries of immature females of the measurements were taken with an electronic meadow lizard consisted of transparent fol- balance (accuracy 0.001 g). The characteris- licles up to 1.6 mm in diameter. Somewhat tics of the eggs were, in all cases, determined larger follicles (>2 mm) showed signs of vi- within 8 h of laying. The estimate of egg vol- tellogenesis and were used as an indication of ume (V=4/3π a2b, a and b being half of the the first sign of reproductive activity of the fe- width and length of the egg, respectively) males. According to this criterion and on the was taken as an overall measure of egg size. basis of 41 collection specimens examined, In all cases each clutch was unequivocally the minimum size at which the females from assigned to an individual female, allowing us Serbia attained sexual maturity was 49.5 mm to calculate the relative clutch mass (RCM) in SVL, and all females larger than this size as the ratio of clutch mass to postparturition reproduced. Percentage of mature females in body mass. the analysed sample was 68%. 2.3. Statistical procedures 3.2. Oviposition and clutch frequency Descriptive statistics (mean, standard er- The data obtained through fieldwork and ror, standard deviation, range) for all traits from specimens already preserved in collec- were calculated. For subsequent analyses all tion gathered in different years were pooled variables were log-transformed, to ensure together in order to give an overall descrip- data normality and to generate homoge- tion of the oviposition period in this species. neous variances (Sokal and Rohlf 1981). At the end of April the females bore small Since some previous studies (see e.g. Galán to medium vitellogenic follicles. Specimens 1997) revealed that the relationships between from all three study localities carried ovi- clutch characteristics and maternal SVL dif- ductal eggs during May, while in all cases the fer depending on whether oviductal or vi- clutches were laid during June in the labora- varium-laid eggs are taken into account, here tory. None of the examined females exhibited we considered only the vivarium-laid eggs. a simultaneous presence of enlarged vitello- Only the mean clutch size was estimated on genic follicles and oviductal eggs or corpora the basis of data for 12 clutches laid in the lutea. We neither found other indicators for jjournalournal 114.indb4.indb 229191 22008-06-21008-06-21 112:17:372:17:37 292 Katarina Ljubisavljević et al. Table 1. Egg and clutch characteristics of female Darevskia praticola from Serbia based on average values for each clutch. The mean ± SE, SD, range and number of adult females or clutches analysed (N) are shown. mean ± SE ± SD range N Female SVL (mm) 54.74 ± 0.60 3.16 49.49 – 60.51 28 Clutch size 5.71 ± 0.29 1.21 4 – 8 17 Clutch mass (g) 1.508 ± 0.090 0.311 1.184 – 2.220 12 Relative clutch mass (RCM) 0.601 ± 0.028 0.097 0.454 – 0.771 12 Egg mass (g) 0.264 ± 0.008 0.029 0.215 – 0.332 12 Egg length (mm) 10.67 ± 0.17 0.59 9.62 – 11.69 12 Egg width (mm) 6.61 ± 0.07 0.23 6.21 – 6.94 12 Egg volume (mm3) 244.76 ± 7.40 25.65 194.08 – 282.87 12 the production of more than one clutch per P> 0.05) and egg volume (r = –0.37, F 1,10 = season.