HERPETOLOGICAL JOURNAL, Vol. 14, pp. 13-19 (2004)

FEEDING ECOLOGY OF VI PERA LA TA STEI IN NORTHERN PORTUGAL: ONTOGENETIC SHIFTS, PREY SIZE AND SEASONAL VARIATIONS

JOSE C. BRITO

Centro de Biologia Ambiental and Dep. de Biologia Animal, Faculdade de Ciencias da Universidade de Lisboa, Lisbon, Portugal

The diet of Vipera latastei was investigated in northern Portugal from 1998 to 2002. Palpation of stomach contents and forced defaecation from 190 snake specimens resulted in the recovery of 83 identifiableprey items. V. latastei preys on fo ur species of small mammal (76%), two lizard species (14%), three amphibian species (5%) and arthropods s.l. (5%). Estimates of prey availability demonstrated that the most common prey were also the most frequent prey consumed. No differences between the sexes were detected in terms of the proportion of snakes with prey or diet composition. However, there was an ontogenetic shift in diet composition. Juveniles fed mostly on ecthotermic prey (60%), the maj ority of subadults fed on insectivorous mammals and lizards (60%), and adults fed mainly on rodents (88%). This ontogenetic shift is mostly due to the morphological constraints imposed on the juveniles, which cannot swallow large prey items. There is a positive correlation between snake size and prey size. V. latastei is selective in terms of both the species and size of prey ingested, with larger snakes being more selective than smaller snakes. Larger snakes have a narrower food niche breadth than smaller snakes, but their diet composition overlaps 1i1oderately. There is seasonal variation in the diet composition, with snakes taking amphibians mainly in spring and autu mn, lizards in spring, and mammals in summer and autumn. Feeding fr equencies indicate that both males and females - and subadults and adults - consume prey more frequently during summer.

Key words: dietary habits, food selection, prey availability, snakes, Viperidae

INTRODUCTJON Despite its former wide distribution in P01iugal, it pres­ ently only occurs in isolated and small populations, Foraging ecology plays a direct or indirect role in vir­ especially in the south (Godinho et al., 1999). This spe­ tually all aspects of snake life-history (Mushinsky, cies is threatened mainly by habitat loss, as a 1987). For instance, obtaining sufficientenergy is a cru­ consequence of forest fires and intensification of agri­ cial factor in reproduction and it may influence the culture, and direct human persecution (Brito al., frequency of reproduction (Naulleau & Bonnet, 1996), et 200 l ) . Additionally, the Portuguese preliminary con­ reproductive output (Gregory & Skebo, 1998) or the ma­ servation status of "Indete1minate" for V. latastei ternal condition after reproduction (Shine & Madsen, (S.N.P.R.C.N., 1991), demonstrates insufficient I 997). Feeding also affects demographic traits of knowledge concerning tltis species. Thus, a conserva­ populations, especially rates of growth and survival tion programme has been established in n01ihern (Madsen & Shine, 2000). Portugal, which aims to identifypriority areas for con­ Although the diet of western European vipers(V ipera servation and to develop management strategies for spp.) is relatively well documented (Bea et al., 1992), viper populations and their habitats. data on Lataste's viper (Vip era latastei) are scarce and Data on foraging ecology is important because mostly anecdotal (Valverde, 1967; Yericad & Escarre ( l) species with very specialized diets usually are unable to 1976; Lopez Jurado & Caballero, 1981 ), or come from studies carried out in captivity outside their natural dis­ use alternative prey types if their ha bitats are modified or destroyed; and (2) ontogenetic shifts in diet compo­ tribution area (Saint-Girons, 1980). Although general sition may contribute to the endange1ment of species, accounts (Bea & Brana, 1988) state that this species for instance due to declines in a specific prey type feedson lizards and small mammals, no study has spe­ by a particular age/size class of snakes (Webb cifically addressed its foraging ecology. needed & Shine, 1998). Within the framework of the conserva­ Vipera latastei is a small-sized viper, with snout-vent tion programme for latastei, the following questions length usually not larger than 60.0 cm. It is distributed V. were asked: (I) is it a size-selective predator and how fromnorth-west ern Africa (Morocco, Tunisia and Alge­ does prey availability influence the diet composition? ria) in the south, to the Iberian peninsula in the north, (2) with the exception of Cantabrian and Pyrenean regions. Are there ontogenetic shiftsin the diet composition? (3) Are there intersexual differences in the diet compo­ sition? ( 4) What is the frequency of feeding and are Correspondence: J. Brito, Centro de Investigai;:ao em there seasonal changes in diet composition? These Biodiversidade e Recursos Geneticos da Universidade do Porto (CIBIO/UP), Instituto de Ciencias Agrarias de Vairao, questions should help in understanding the foraging R. Padre Armando Quintas, 4485-661 Vairao, Portugal. E­ ecology of V. latastei and allow the identificationof fa­ mail: [email protected] vourable ecological niches for the species. 14 J. C. BRITO

METHODS The availability of potential prey was estimated in a smaller area (2.24 ha), and compared with prey data STUDY AREA from thevipers capturedin this area (64% of the total ex­ The study was conducted in the extreme north-west­ amined snakes). Due to logistical reasons, prey ern part of the distribution area of V latastei (Brito & availability fo cuses only on small mammals and lizards. Crespo, 2002), in a National Woodland and Biosphere Small manu11al abundance was estimated using l 00 Reserve, "Mata de Albergaria ", Portugal ( 4 l 049'N, Shen11a1m traps, placed in line at 10 111 intervals, baited 8°07'W; elevation c. 750 m a.s.1.), included in a pro­ and surveyed for six nights, in June 1999, 2000 and tected area, Parque Nacional da Peneda-Geres. The area 2001 . This period should conespond to ilie annual abun­ is a series of granite mountains, witl1 altitudes ranging dance peak of small manmials in this region (Mathias, from 50 to 1500 m a.s.I. The climate is Atlantic Medi­ 1999). Traps were examined in early morning (at about terranean (Goday, 1953), characterized by wet winters 07:00 to 08:00 hr) and late afternoon (at about 19:00 to and hot summers. Mean monthly rainfall ranges from55 20:00 hr). Lizard abundance was estimated through mmin July to 457 nun in January (average annual rain­ visual encounter surveys of 1 OOO m, in the same area at fall above 3200 mm/year), and mean air temperature the same time period. Tluee lacertid lizard species were ranges from 7. 9 °C in January to 2 0 .4 °C in July (aver­ foW1d in this smaller study area (Podarcis bocagei, age annual air temperatme 14 °C) (INAG, 2002). The Lacerta schreiberi and L. lepida), but - due to the large vegetation consists mostly of deciduous oak forests size of adult L. schreiberi and L. lepida - only juveniles (Quercus robur) or mixed deciduous and coniferous were considered as potential prey for V latastei. forests (Pinus sp.) with arbutus trees (Arbutus unedo). The extent of sexual divergence in viper body mor­ Major shrubs include heath ( Calluna vulgaris and Erica phology was quantified according to the methods sp.), gorse (Ulex sp.) and brooms (Cytisus sp.). outlined by Shine (1991 a). For SVL, the quotient be­ tween fe male SVL and male SVL was found. For HL, FIELD AND ANALYTICAL METHODS HW and HH, a regression equation linking female meas­ Data were collected between 1998 and 2002 in an ures to the SVL was calculated. An estimate of the area of around 2.300 ha. 135 snakes were caught by female headmeasure was obtainedusing the mean ma le band, sexed by analysing external tail morphology, SVL. The ratio of female to male head length, at a mean measured for snout-vent length (SVL), head length male SVL, provided an index oftJ1e extent of sexual size (HL), bead width (HD), and head height (HH), and per­ dimorph ism in relative head measmes (Shine, 1991 a). manently marked for future identificationby clipping For analysis purposes, snakes were divided into tluee unique combinations of ventral scales. Snakes were pal­ size categories according to tl1eir sexual maturation sta­ pated to detect prey in the stomach and intestines, but tus: (1) Juveniles and neonates (SVL <30.0 cm); (2) only forced to defecate intestine contents and never subadults (SVL 30.0-40.0 cm); (3) adults (SVL >4 0.0 forcedto disgorge prey; however, prey presence was cm) (Brito & Rebelo, 2003). Diffe rences in diet compo­ recorded in such cases. Additionally, 55 road-killed sition were analysed by x2 tests and contingency tables snakes were dissected to collect stomach and intestine (Siegel & Castellan, 1988). To avoid pseudoreplication prey items. Mammals were identified using identifica­ of data, no snake was included in the analysis more ilian tion keys based on the internal structure of hairs once. Niche breadth was measured using the standard­ (Teerink, 1991). Reptiles, amphibians and invertebrates ized Levin's index (BA) and niche overlap between were identifiedthrough their externalmorphology. Prey ea tegories was measured using Pianka' s symmetrica I items collected from road-killed snakes were measured equation (Ojk) (for details see Ksebs, 1989). The relation (body length), whenever their preservation status al­ between snake size and prey size was detennined by cal­ lowed. culating Kendall's tau (Siegel & Castellan, 1988),

TABLE I. Sample sizes, body measures and sexual size dimorphism of Vipera latastei containing prey. SVL , snout-vent length,; HL, head length; HW, head width; HH, head height (all measures in cm); SD, standard deviation; SSD, sexual size dimorphism. See methods for calculation of SSD. ------Mean Mean Mean Mean 11 SVL±SD HL±SD HW±SD HH±SD

Males 62 36.3±9.1 2.02±0.36 1.38±0.26 0.76±0.13 Feniales 39 37.7±8.3 2.12±0.37 1.40±0.30 0.79±0. 17 SSD (f/m) 101 1.04 1.01 0.99 1.00 Juveniles 21 25.0±3.7 1.56±0. 19 1.02±0. 13 0.57±0.06 Subadults 38 34.3±3.2 1.97±0. 16 l.33±0. 15 0.75±0. 11 Adults 42 45.2±4.7 2.37±0.21 1.59±0.22 0.89±0.09

Total 101 36.9±8.8 2.06±0.37 1.39±0.28 0.78±0. 15 FEEDING ECOLOGY OF VJPERA LA TASTEI 15

testing snake SVL against prey length. In the analysis of (5 %) (Table 2), and the differences between the fre­ feeding frequencyand seasonal changes in diet composi­ quencies of prey type consumed were significant tion, data were pooled into three seasons - spring (X,2=116.8, df=3 , P<0.001). The rodent Apodemus (March to May), summer (June to August) and autumn sylvaticus was the most frequentprey, representing 40% (September to November). No data were available be­ of the total prey taken. Analysis by sex shows that fe­ tween December and February due to the low activity males take proportionally more mammals than males, levels of V. latastei during the winter (Brito, 2003). but differences were not significant (all size categories: Feeding frequencies were infened from the proportion x2=1.74, df=3, P=0.628; adults only: x2= 1.94, df=2, of snakes containing detectable prey items (Shine et al., P=0.164). Since there were no inter-sexual differences 1998). Differences in abundance between snake diet and in the proportion of snakes with and without prey, or in prey availability were analysed by x2 tests contingency diet composition, sexes were pooled in the subsequent table tests. A minimum rejection level of a=0.05 was analysis in the three size categories. used in all statistical tests. There were significant differences in the dietary composition of the three size categories x2= 17.5, df=6, RESULTS ( P<0.05) (Table 2). Juveniles fe d essentially on A total of 190 individual snakes were analysed in this ectothermic prey (60%), subadults fe d on small mam­ study, of which I 01 contained prey (Table 1). The rela­ mals (74%), mainly insectivores (43%), but also on tive number of snakes with and without prey did not ectothermic prey (26%), while adults fed almost exclu­ differ significantly between sexes (x2=0.003, df=I, sively on rodents (88%), and only to a small extent on P=0.954). V. latastei preys on four species of small insectivorous (15%) and ectothermic prey (6%). The mammal (76% oftbe total prey items), two lizard species importance of small mamma ls in the diet of V. latastei (14%), three amphibian species (5%) and aiiluopods s.I. increased with snake size, representing 40% of the diet

TABLE 2. Diet composition and niche breadth (BA) from different sexes and size categories of Vipera latastei. Number of prey items and percentage (in parenthesis). See methods for size categories.

Males Females Juveniles Subadults Adults Total

No. snakes examined I 17 73 50 70 70 190 % of snakes with prey 53.0 53.4 42.0 58.6 55.7 53.2

Coleopteran 3 (6.3) I (2.9) 2 (13.7) 2 (5.7) 4 (4.8) Total Arthropod 3 (6.3) I (2.9) 2 (13.3) 2 (5.7) 4 ( 4.8)

A ly tes obstetricans I (2.1) I (6.7) I (1.2) Ch ioglossa lusitanica I (2.1) 1 (6.7) I ( 1.2) Triturus boscai 1 (2.9) 1 (3.0) 1 (J .2) Unidentified amphibians 1(2. l) I (2.9) l (1.2)

Total Amphibia 3 (6.3) l (2.9) 2(1 3.3) l (2.9) 1 (3.0) 4 (4.8)

Podarcis bocagei 4 (8.3) 3 (8.6) 1 (6.7) 6(17.1) 7 (8.4) Podarcis hispanica 1 (2.1) 1 (6.7) I (1.2) Un.identifiedLacertidae 3 (6.3) I (2.9) 3 (20.0) 1 (3.0) 4 ( 4.8)

Total Reptilia 8 (16.7) 4 (J 1.4) 5 (33.3) 6 (17.1) 1 (3.0) 12 (14.5)

Crocidura russula 7 (14.6) 5 (14.3) 2 (13.3) 10 (28.6) 12 (14.5) Sorex granarius 2 (4.2) 4(11.4) 1 (6. 7) 5 (14.3) 6 (7.2) Microtus lusitanicus 4 (8.3) 3 (8.6) 2 (13.3) 5 (J 5.2) 7 (8.4) Apodemus sylvaticus 19 (39.6) 14 (40.0) l (6.7) 8 (22.9) 24 (72.7) 33 (39.8) Unidentifiedmamma ls 2 ( 4.2) 3 (8.6) 3 (8.6) 2 (14.3) 5 (6.0)

Total Mammals 34 (70.8) 29 (82.9 ) 6 (40.0) 26 (74.3) 31 (93 .9) 63 (75.9)

STANDARDIZED LEVJN's B A 0.287 0. 142 0.754 0.236 0.044 0.221 16 1. C. BRITO

8 Invertebrates •Amphibians �Reptiles 0 Mammals of juveniles and 94% of adults, while the importance of 100 0 ectothennic prey decreases with snake size, representing ., "6 60% of the diet of juveniles and 6% of adults. -5 75.0 The food niche (B ) varied between 0.044 for adult <.... A 0 " snakes to 0.754 for juveniles (Table 2), indicating that

..g ., 50.0 adults were more specialized than subadults and juve­ 0 Cl. niles. Food niche differences between all size categories s 8 25.0 of males and females indicate a greater specialization in � fema les, which consumed proportionally more mammals 00 than the maks (Table 2). Niche overlap (Oik) was high be­ Spring Summer Autumn tween subadults and adults (Ok=0.978), and moderate i FIG. I. Seasonal variation in the diet composition of Vipera between juveniles and sub-adults (Ok=0.863), and be­ i lalastei from northern Portugal. tween juveniles and adults (Ok =O. 749). Overlap between i all size categories of ma les and fe males was very high 80.0 - A 0.993 ). OMales (Oik = There was seasonal variation in diet composition, al­ �' 60.0 Cl. though this was not significant (x2=8.15, df=6, P=0.228) -.5 (Fig. I). Snakes fed on an increasing proporiion of mam­ � 40 0 mals, representing 50% of the diet in spring, but 72% in � autunm. Reptiles were mostly preyed upon in the spring 5i (25%), while in autunm they were only marginally uti­ '$ 20 0 lized as prey items (9%). Amphibians were mostly preyed upon in the spring (25%) and autunm (14%). 00 Analysis of feeding frequency demonstrated signi fi­ Summer Auhmm Spring cant diffe rences in the proportion of males and fe males 80.0 with and without prey from spring to autu111J1 (males: D Juveniles x2=6.49, df=2, P=0.039; females: x.2=6.66, df=2, m Sub-adults P=0 .036) (Fig. 2A), with both sexes consuming prey 17 60 0 o. •Adults more frequently in the su111J11er. There were also differ­ -.5 ences between the size categories (Fig. 2B), but these � 'iO 0 were only significant for subadults (x.2=6.71, df=2, .E"' 5i P=0 .035) and adults (x2=7 .97, df=2, P=0.01 9). These ::!Cc 20.0 size categories of snakes consumed prey more frequently in the su111J11er. There was a positive correlation between snake SVL 0 0 and prey size (Fig. 3) (Kendall's tau = 0.562, P<0.00 1). Spring Sununer Autumn Small snakes ate smaller prey, mostly lizards and insec­

FIG. 2. Seasonal changes in the proportion of Vipera tivorous manuna ls, but mean prey size shifted at an latastei from northern Portugal containing prey, by sex (A) intermediate body size (SYLs around 40.0 cm), with and by size category (B). See text for size categories. larger snakes feeding almost exclusively on larger prey.

TA BLE 3. Percent frequency of prey types consumed by Vipera /atastei compared to prey availability.* Indicates significant diffe rences between availability and proportion of prey in the diet (P=0.01 ). ** Indicates prey types wh ich varied significantly in abundance over the study period (P<0.00 I).

Consumed Available

1999 2000 2001 1999 2000 200 1

Podarcis bocagei 100.0 iOO.O 100.0 93.8 94 .3 97.5 Lacerta schreiberi 6.3 3.8 2.5 Lacerta lepida 0.0 1.9 0.0

Total reptiles 2 5 16 53 40

Crocidura russula 50.0 15.4 30.0 20.0 2.0 11.1 Sorex granarius 15.4 Microtus lusitanicus 15.4 2.0 Apodemus sylvaticus 50.0 53.8* 70.0 80.0 96. 1 ** 88.9

Total mamma ls 4 13 10 15 51 9 FEEDING ECOLOGY OF VIPERA LA TASTEI 17

25.0 fluctuationsin the availability of prey. Lizards are espe­ • cially abundant during their mating season, in spring, 20.0 and afterhatching, ins ummer, and are thus susceptible • '§ •• to predators (unpublished data). Amphibians should be :B15 0 O!J . -----­ more available during the spring and autumn, when the 5 -;. 10 0 �... climatic conditions enhance their movements (personal J: � observation). The two species of mammal most fre­ 5.0 quently taken ( Crocidura russula and Apodemus •�. . • + sylvaticus) have their offspring in spring and autumn 0.0 -1----.------.------.----,----i (Mathias, 1999), thus being more available during these 10.0 20.0 30.0 40.0 50.0 60. 0 seasons. Snake snout-vent length (cm) Intersexual differences in diet composition have FIG. 3. Relationship between Vipera latastei snout- vent been reported for several snake species, such as the (r'=0.424, n= P=0.003), length and prey length I 9, in northern acrochordid Acrochordus arafurae (Houston & Shine, Portugal. 1993), the pythonid Morelia spilota (Pearson et al., Estimates of availability of the two main prey types 2002), and the colubrid Natrix maura (Santos & (lizards and small mammals; Table 3), and diet composi­ Llorente, 1998). These diffe rences are mainly caused tion analysis indicate tha1 the most common prey by sexual size dimorphism, females being larger than (Podarcis bocagei, Crocidura russula and Apodemus males, and thus able to feed on larger prey. Such sexual sy lvaticus) were also the species most frequentlypreyed differences in diet composition could also correspond upon. Combinil1g data for the three years of study, these to a higher energetic demand for females, wluch must tlu·ee prey species accounted for 88.6% of the diet of the devote more resources to their reproductive output snakes, and their availability represents 97.2% of the to­ (Shine, 1989). Females of V. latastei are larger than tal sample. Although available, juveniles of lacerta males in SVL and relative head length, bu1 the extent of schreiberi and L. lepida were apparently no1 preyed on, sexual size dimorphism is small compared to other and Sorex granarius only marginally so. snake species (Slune, 1991 a). As expected, there are no significant intersexual differences in the diet composi­ DISCUSSION tion. The high value of diet overlap between all size The diet of V. latastei in the study area follows the categories of males and females further demonstrates general diet pattern of the European Vipera (Bea et al., the absence of intersexual differences in die1 composi­ 1992), with the main prey taken being small mammals. tion. Nevertheless, it is significantly diffe rent (x2=J3.4l, Overall, V. latastei exhibits a main foraging period df=4, P=0 .009) from the diet ofthis species reported in a in the summer, which should be related to environmen­ study of museum specimens from Spain (Bea & Brana, tal temperat·ure. Winter in this montane study area is 1988). For instance, amphibians have never previously rainy and long, thus snakes forage mainly in summer been reported as prey for V latastei. However, in the (Brito, 2003), when temperatures should be optimal to study area they represent 13% of diet of the juveniles and enhance digestion (Naulleau, 1982). Juvenile V. 3% of the diet of subadults and adults, respectively. Bea latastei had a similar feeding frequency compared to & Brana (1988) suggested that V. latastei had a low fre­ the other size categories, while juveniles of other Euro­ quency of small mammals in the diet (58%) compared to pean viperids have been reported to feed less frequently other European vipers, while in northern P011ugal they than sub-adults and adults (Prestt, 1971; Luiselli, play an important role, representing up to 76% oftbe diet 1996). It has been suggested that this can be a direct of this viper. These differences in diet composition be­ consequence of the generalized feeding activity of ju­ tween s�dies are probably due to annual and/or seasonal veniles and/or to gape linutation: juvenile snakes are differences in prey availability between regions. For in­ too small to ingest adult rodents, and thus have a stance, mammal abundance in the smaller study area smaller range of prey types which can be ingested fluctuatedsignifica ntly during the three years of study. (Shine & Madsen, 1997). However, since juveniles are Seasonal shifts in dietary composition of snakes are not involved in mating activities, they should maximize expected due to spatial and temporal changes in prey the rate of prey intake in order to enhance their growth availability throughout the year, and examples include rates and survival during the hibernation period (e.g. the viper Vipera ursinii (Agrimi & Luiselli, 1992), the Bonnet, 1997). python liasis fu scus (Madsen & Shine, 1996), and the Ontogenetic shiftsin diet composition are a comn1on colubrid Coluber constrictor (Shewchuk & Austin, pattern in snakes, mainly because they are gape-limited 2001). V. latastei exhibited seasonal changes in diet predators (Mushinsky, 1987), and the European Vipera composition, with snakes taking an increasing propor­ usually exhibit a shift from lizards to small mammals tion of mammals from spring to autumn. Conversely, (Saint-Girons, 1980). Juvenile V. latastei feed essen­ there is a decrease inthe proportion oflizards taken. Am­ tially on ectothennicprey, while adults prey on rodents, phibians are mainly preyed upon during spring and and there is a significant correlation between snake size autumn. These seasonal shifts are most likely related to and prey size. This ontogenetic shiftin the diet compo- 18 J. C. BRITO

sition could be explained in two ways. First, diffe rences Bea, A., Brana, F., Baron, J. -P. and Saint-Girons, H. in energetic content of prey types. Optimal foraging ( 1992). Regimes et cycles alimentaires des viperes theory predicts that a predator will ignore small prey europeennes (Reptilia, Viperidae). Annee Biologique when densities of large prey are high enough for preda­ 31, 25-44. tor survival (Schoener, 1971 ). Additionally, the Bonnet, X. ( 1997). Influence of size on survival in ingestion of numerous small prey would increase cap­ newborn asp vipers ( Vipera aspis): preliminary resu lts. ture and handling time compared to the ingestion of Amphibia-Reptilia 18, l 07-1 12. fewer large prey. Thus, adult snakes should prefer mam­ Brana, F., Bea, A. & Saint-Girons, H. (1988). mals to lizards, due to the larger size and higher Composici6n de la dieta y ciclos de alimentaci6n en energetic content of manmrnls. Second, morphology Vipera seoanei Lataste, 1879. Variaciones en relaci6n may impose constraints on ingestion capacity in juvenile con la edad y el ciclo reproductor. Munibe 40, 19-27. snakes. Snakes are gape-size limited predators, there­ Brito, J. C. (2003). Seasonal and daily activity patterns of fore juveniles are forced to eat small prey. With an Vipera latastei in northern Portugal. Amphibia­ increase in snake body size the number and size of in­ Reptilia 24, 497-508. gestible prey types increase, allmving larger snakes to Brito, J. C. & Crespo, E. G. (2002). Distributional analysis utilize a greater range of prey sizes and taxa (Shine, of two vipers (Vip era latastei and JI. seoanei) in a 1991b). potential area of sympatry in the Northwestern Iberian Nevertheless, lizards are not entirely excluded from Peninsula. Jn Biology of the Vipers, 129-138. Schuett, the diet of adult V. latastei, representing about 6% of G. W., 1-l oggren, M., Douglas, M. E. and Greene, H. W. prey taken, which is a common pattern in other Euro­ (Eds.). Eagle Mountain Publishing. pean viperids (Brana et al., 1988; Luiselli & Agrim.i, Brito, J. C. & Rebelo, R. (2003). Diffe rential growth and I 991 ). Estimates of prey availability demonstrate that mortality affe ct sexual size dimorphism in Vipera both small prey (lizards and insectivorous mammals) latastei. Copeia 2003, 865-871. and large prey (rodent mammals) are readily available. Brito, .I. C., Rebelo, A. & Crespo, E. G. (2001). Viper Faced with this availability of prey, adult snakes should k illings for superstitious reasons in Portugal. Boletin not be entirely forced to shift from lizards to rodent de la Asociaci6n Herpetol6gica Espaliola J 2, l 01-1 04 . ma mmals. Goday, S. R. (1953). Apreciaci6n sintetica de Jos grados In conclusion, V. latastei is sel ective in tenns of the de vegetaci6n de la sierra de Geres. In A Flora e a species and the size of prey i11gested, with larger snakes Vegeta9iio da Serra do Geres, 449-480. Lisboa: being more selective than smaller snakes. In addition, lnstituto de Alta Cultura. larger snakes have a narrower food niche breadth than Godinho, M. R., Tei xeira, J., Rebelo, R., Segurado, P., do smaller snakes. The ontogenetic shift in the dietary Loureiro , A., Alvares, F., Gomes, N., Cardoso, P., habits is mostly due to the. morphological constrnintsim­ Camilo-Alves, C. & Brito, J. C. (1999). Atlas of the posed on juveniles, which ca1mot swallow large prey continental Portuguese herpetofauna: an assemblage of items. pu blished and new data. Revis/a EspaJiola de Herpetologia 13, 61-82. ACKNOWLEDGMENTS Gregory, P. T. & Skebo, K. M. ( 1998). Trade-offs between This work was financially supported by a PhD grant reproductive traits and the influence of food intake from Fundar;:ao para a Ciencia e Tecnologia (PRAXIS during pregnancy in the garter snake, Th amnophis XX I/BD/ I 6093/98) and by Parque Nacional da Peneda­ elegans. American Naturalist 151, 477-486. Geres. lnstituto da Conservar;:ao da Natureza (l CN) Houston, D. & Shine, R. (1993). Sexual dimorphism and granted the permissions for capture of snakes. We ac­ niche divergence: feeding habits of the Arafura knowledge the Banco de Tecidos de Vertebrados fi lesnake. Journal of Animal Ecology 62, 737-748. Selvagens (ICN) for providing some of studied speci­ I. N. A. G. (2002). Dados de Base - Sistema Nacional mens and all the people that collected road-killed vipers, Informas:ao de Recursos Hidricos. lnstituto Nacional especially F. Alvares and A. Rebelo. Thanks to E. da Agua. Available on-line: http://snirh.inag.pt/snirh/ Crespo, T. Madsen, R. Rebelo, M. Carretero, S. dados_ base/main_nav _fr. html Thorsteinsd6ttir and A. Loureiro for connnents on the Krebs, C. J. (1989). Ecological Methodology. New York : manuscript. Harper Collins. Lopez Jurado, L. F. & Caballero, M. R. (1981). Predaci6n REFERENCES de Vipera latastei sobre Mustela nivalis. Doliana Acta Vertebrata 8, 298"299. Agrimi, U. & Luiselli, L. M. (1992). Feeding strategies of the viper Vipera ursinii ursinii (Reptilia: Viperidae) in Luiselli, L. (1996). Food habits of an alpine population of viper ( Vip era ammodytes). Journal of the Apennines. Herpetological Jo urnal 2, 37-42. the sand Bea, A. & Brana, F. ( 1988). Nota sabre la alimentaci6n de Herpetology 30, 92-94. ,Vip era lataslei, Bosca, 1878 (Reptilia, Viperidae). Luiselli, L. M. & Agrimi, U. (1991 ) . Composition and variat ion of the diet of Vipera aspis francisciredi in Munibe 40, 121-124. FEEDING ECOLOGY OF VJPERA LA TA STEI 19

relation to age and reproductive stage. Amphibia­ Shewchuck, C. H. & Austin, J. D. (2001). Food habits of Reptilia 12, 137-144. the racer (Coluber constrictor mormon) in the northern Madsen, T. & Shine, R. (I 996). Seasonal migration of part of its range. Herpetological Journal 11, 151-155. predators and prey - a study of pythons and rats in Shine, R. (1989). Ecological causes for the evolution of tropical Australia. Ecology 77, 149-156. sexual dimorphism: a review of the evidence. Th e Madsen, T. & Shine, R. (2000). Silver spoons and snake Quarterly Review of Biology 64, 4 I 9-461 . body sizes: prey availability early in life influences Shine, R. ( 1991 a). Jntersexual dietary divergence and the long-term growth rates of free-ranging pythons. evolution of sexual dimorphism in snakes. American Journal of Animal Ecology 69, 952-958. Naturalist 138, 103-122. Mathias, M. L. (Coard.). (1999). Guia dos Maniife ros Shine, R. (I 991 b ). Why do larger snakes eat larger prey Te rrestres de Portugal Continental, A{:ores e Madeira. items ? Functional Ecology 5, 493-502. Lisboa: Jnstituto da Conservayao da Natureza and Shine, R. & Madsen, T. (I 997). Prey abundance and Centro de Biologia Ambiental da Faculdade de predator reproduction: rats and pythons on a tropical Ciencias da Universidade de Lisboa. Australian floodplain.Eco logy 78, I 078-I 086. Mushinsky, H. R. (I 987). Foraging ecology. In Snakes - Shine, R., Harlow, P. S., Keogh, J. S. & Boeadi. (1998). Ecology and Evolutionary Biology, 302-334. Seigel, The influence of sex and body size on food habits of a R. A., Collins, J. T. and Novak, S. S. (Eds). New York: giant tropical snake, Py thon reticulatus. Functional McGraw Hill. Ecology 12, 248-258. Naulleau, G. ( 1982). Action de la temperature sur la Siegel, S. & Castellan Jr., N. J. (1988). Nonparametric digestion chez Jes viperes espagnoles du genre Vipera. Statistics fo r the Behavioral Sciences. New York: Publicaciones de/ Centro Pirenaico de Biologia McGraw Hill. Experimental 13, 89-94. S. N. P. R. C. N. (eds.). (1991). Livro Vermelho dos Naulleau, G. & Bonnet, X. (1996). Body condition Vertebrados de Portugal. Vol. I - Mamiferos, Aves, threshold for breeding in a viviparous snake. Anfibios e Repteis. Lisboa: Servivo Nacional de Oecologia 107, 301-306. Parques, Reservas e Conservayao da Natureza. Pearson, D., Shine, R. & How, R. (2002). Sex-specific Teerink, B. J. (I 991 ). Hair of West-Europ ean Mammals. niche partitioning and sex ual size dimorphism in Cambridge: Cambridge University Press. Australian pythons (Morelia sp ilota imbricata). Valverde, J. A. (1967). Estructura de una comunidad Biological Journal of the Linnean Society 77, I I 3- mediterranea de vertebrados terrestres. Monografias 125. de Ciencia Moderna, 1. CSIC. Prestt, I. (I97 I). An ecological study of the viper Vipera Vericad, J. R. & Escarre, A. (1976). Datos de alimentaci6n berus in southern Britain. Journal of Zoology J 64, de ofidios en el Levante sur iberico. Mediterranea 1, 5- 373-4 I 8. 32. Saint-Girons, H. ( 1980). Modifications selectives du Webb, J. K. & Shine, R. (1998). Ecological characteristics regime des viperes (Reptilia: Viperidae) !ors de la of a threatened snake species, Hoplocephalus croissance. Amphibia-Reptilia 1, 127-I 36. bungaroides (Serpentes, Elapidae). Animal Santos, X. & Llorente, G. A. ( 1998). Sexual and size­ Conservation I, 185-193. related diffe rences in the diet of the snake Natrix maura from the Ebro delta, Spain. Herpetological Jo urnal8, 16 I -165. Schoener, T. W. (1971). Theory of feeding strategies. Annual Review of Ecology and Systematics 11, 369- 404. Accepted: 20.2.03