Catchability in snakes: consequences for estimates of breeding frequency

Catchability in snakes consequences for estimates of breeding frequency

1 Xavier Bonnet and Guy Naulleau

Abstract Estimates of breeding frequencies in wild populations occupy a central place in life-history studies but they remain hard to obtain in snakes Capture-recapture procedures were carried out during 27 periods over 3 years in a large population of female asp vipers Vipera aspis (317 marked individuals) Catchability was greater in reproductive females (264 f 174 (SD) captureslyear) than in nonreproductive females (176 f 102) Three different methods were used to estimate breeding frequency (i) the proportion of reproductive to nonreproductive females for each period (noncorrected estimates NC) was measured (ii) the ratio of the number of reproductive females to the total number of females was estimated using the CAPTURE program (catchability-corrected estimates CE) (iii) breeding frequencies of marked individuals were observed in the field for 3 years (observed breeding frequency OF) The results obtained with the three methods were different The NC procedure indicated that 63 of females breed each year suggesting an annual -biennial cycle The CE (34 reproductive females) and OF (34 reproductive females) procedures both suggested a triennial cycle The OF method was the most accurate breeding frequencies were overestimated with the NC method The NC method is the one most commonly used and estimates of breeding frequency in snakes obtained in these studies may be imprecise We conclude that differential catchability between reproductive and nonreproductive females should be taken into account in estimating breeding frequencies

Rdsumd Lestimation de la friquence de la reproduction chez les populations naturelles occupe une position centrale en ecologie toutefois ce paramktre reste trks difficile a obtenir chez les serpents Vingt-sept sessions de capture-recapture ont CtC rCalisCes sur un grand nombre de femelles de vipkres aspic Vipera aspis (317 individus marquCs) La capturabilitk des femelles reproductrices (264 + 174 (Ccart type) captures par an) est plus ClevCe que celle des femelles non reproductrices (176 f 102) Trois mCthodes diffkrentes ont ete utilisees pour estimer la frequence de la reproduction (i) le rapport entre le nombre de femelles reproductrices et de femelles non reproductrices a CtC calcule pour chaque session de capture (estimation non corrigee NC) (ii) le rapport entre le nombre des femelles reproductrices et le nombre total de femelles a ete estime a laide du programme CAPTURE

(estimation corrigCe en tenant compte de la capturabilitk CE) (iii) la frkquence de reproduction a CtC obtenue chez des individus marquis et suivis sur le terrain pendant 3 annCes (frkquence de reproduction observCe OF) Les rksultats obtenus par les trois mCthodes sont diffkrents La methode NC indique que 63 des femelles se reproduisent chaque annCe suggCrant un cycle moyen annuel - bisannuel Les procedures CE (34 femelles reproductrices) et OF (34 femelles reproductrices) suggkrent toutes les deux lexistence dun cycle moyen triennal ~ t a n t donne que la mCthode OF est la plus prCcise la frkquence de la reproduction est surrestimke avec la methode NC La mCthode NC est la plus utiliske les estimations de la frequence de la reproduction des serpents pourraient Ctre impricises dans beaucoup dCtudes Nous concluons que la capturabilitk differentielle entre les femelles reproductrices et non reproductrices doit Ctre prise en compte pour estimer la frequence de la reproduction

Introduction 1987 Madsen and Shine 1994) Snakes exhibit very secre-

Breeding frequency is one of the major reproductive traits in iteroparous animals which is used to test life-history theories in endeavouring to understand the factors that determine life- history patterns It is particularly important for correlating clutch size with parental survival or current with future reproduction or to test the trade-off between growth and reproduction (Bull and Shine 1979 Stearns 1992) Relation- ships between clutch size breeding frequency and survival have been investigated in many vertebrates (Stearns 1992) but remain poorly documented in snakes (Seigel and Ford

tive habits and long-term field studies are difficult to carry out (Seigel 1993) Fecundity survival and lifetime repro- ductive success are known for only one species Vipera bems and for only one small population (Madsen and Shine 1994)

However some populations in some localities such as a number of viperids of northern Europe (eg Vipera berus and V aspis) offer better conditions for observation (Madsen and Shine 1994 Naulleau and Bonnet 1995) In these popula- tions the mating season occurs in early spring when ambient temperatures are rather low Soon after hibernation both males and females require high body temperatures (Naulleau et al 1987 Bonnet and Naulleau 1993) and bask frequently

Received February 17 1995 Accepted August 24 1995 this behaviour greatly increases their catchability X Bonnet and G Naulleau Centre National de la Basking in sunlight is known to be a thermoregulatory Recherche Scientifique Centre dEtudes Biologiques de behaviour in reptiles that allows the body temperature to rise Chize 79360 Beauvoir sur Niort France and physiological processes to occur at higher rates One

Can J Zool 74 233-239 (1996) Printed in Canada 1 ImprimC au Canada

Can

J Z

ool

Dow

nloa

ded

from

ww

wn

rcre

sear

chpr

ess

com

by

Cou

peri

n on

08

311

5Fo

r pe

rson

al u

se o

nly

Can J Zool Vol 74 1996

Table 1 Numbers of snakes (reproductive females + nonreproductive females) captured during each period in relation to the reproductive cycle

March to early June Summer August to September (vitellogenesis) (pregnancy) (parturition)

1 2 3 4 5 6 7 8 9

1993 Noofcapturesperperiod 14 12 27 63 69 46 34 20 20 No newly marked 14 12 21 50 38 14 3 3 9 Total no marked 14 26 47 97 135 149 152 155 164

1994 No of captures per period 48 45 81 72 26 31 29 23 7 1 No newly marked 48 30 43 27 8 6 7 2 20 Total no marked 48 78 121 148 156 162 169 171 191

Note Periods were designated as follows 1 from the end of hibernation until 15 March 2 16-31 March 3 1 - 15 April 4 16-30 April 5 1 - 15 May 6 16 May to the end of vitellogenesis 7 15 days in July 8 15 days in August 9 25 August to 10 September

may expect differences in frequency of sun exposure related had been killed on the road during the entire study period None of to body size and body temperature requirements Spermio- the 64 adult females monitored by radio tracking during the entire genesis in males and particuarly vitellogenesis in females active season left the study area and none crossed any road (Bonnet and Naulleau 1993 Bonnet et al 1994) may require - - body temperatures higher than ambient air temperatures Consequently in spring the greatest proportion of females easily found in the wild should be vitellogenic ones Preg- nancy in these snakes also requires an apparently narrow range of relatively high body temperatures (Charland and Gregory 1990) so the catchability of pregnant females should also be high later in the activity season although perhaps not as high as in the cold springtime when females are expected to be found in exposed locations more often

To test the influence of reproductive status on basking behaviour (and hence catchability) and consequently on breed- ing frequency estimates we tested three different methods on a large sample of adult female asp vipers (Vipera aspis 3 17 marked individuals) (i) the proportion of reproductive females was calculated for each capture period (ii) an esti- mate of the reproductive female population was made using the computer program CAPTURE (Otis et al 1978 30 Decem- ber 1991 version) (iii) 97 individuals were monitored during 2 or 3 consecutive years in order to describe their actual or realized individual reproductive cycle Breeding fre- quency estimates obtained by means of the three methods were compared

Materials and methods

Study area The asp viper Vipera aspis is a small venomous snake widely distributed throughout Europe The study was carried out in west- central France (47 03 N 02 00W) near the northern distribution limit of the species The area covered approximately 33 ha and was bordered by a large road to the north a large road with a high steep bank ( gt 4 m) to the east a large village (Les Moutiers) and the Atlantic Ocean to the south and an open camping area with several roads to the west Consequently the studied population was rela- tively isolated from neighbouring populations This conclusion was supported by the following observations whilst the total population was estimated to be greater than 700 snakes including immatures adult males and females (unpublished data) we found no viper that

Measurements and individual marking The smallest females (n = 5) that gave birth were 475 cm in total length Consequently females gt 475 cm long were considered sexually mature (and henceforth are referred to as adults) All individuals were caught by hand Snout-vent length (SVL) and total length (to the nearest 05 cm) mass (to the nearest 1 g) and sex were recorded Each viper was individually marked by scale- clipping in 1992 (n = 84) and fitted with an electronic tag (1 1 f 1 mm length 125 kHz sterile transponder TX 1400L RhBne MCrieux DestronJIDI INC) in 1993 and 1994 (n = 287 including 54 previously marked in 1992) the total number of marked adult females was 3 17 Electronic tag identification has several advan- tages it is less traumatic than scale-clipping (healing is slower after clipping particularly during long periods of bad weather unpub- lished data) and individual identification is extremely rapid thus reducing disturbance Finally it decreases the risk of mistakes due to the large sample of marked vipers (n = 551 all adults on the study area) Vipers were immediately released at the place of capture within 15 min after initial capture and more rapidly after recaptures

Reproductive status The reproductive status of females was determined using two methods At the beginning of vitellogenesis (March -April Bonnet et al 1994) a female with body condition greater than the repro- ductive threshold (Bonnet et al 1994) was considered reproductive (reproductive females RF) The body condition designations for all 195 free-ranging females caught in March- April and again during the summer were found to be correct for determining whether the female would reproduce in that year hence body condition can be used as a predictor of early reproductive status in female V aspis For small and presumably young females however the reproduc- tive threshold value of body condition and the associated probability of reproduction are less predictable During this study it was not possible to classify 37 females (particularly small females with body condition equal to or slightly above the threshold value) Second from midvitellogenesis (in May) to the end of gestation (late August to September Saint Girons 1957 Bonnet et al 1994) palpation of the abdomen enabled us to detect growing eggs and developing

Can

J Z

ool

Dow

nloa

ded

from

ww

wn

rcre

sear

chpr

ess

com

by

Cou

peri

n on

08

311

5Fo

r pe

rson

al u

se o

nly

Bonnet and Naulleau

Table 2 Annual mean numbers of captures of reproductive females (RF) and nonreproductive females (NRF) in relation to body length -

Size Mean no (cm) of captures SD n F d f P

RF Small ~ 5 3 5 222 143 60 505 2179 0006 Medium-sized 535 - 565 252 173 61 Large gt 565 318 192 61

NRF Small ~ 5 3 5 156 084 75 547 2175 0004 Medium-sized 535 -565 167 088 48 Large gt 565 213 125 55 -

Note See the text for differences between small medium-sized and la1

embryos At this time the distinction between RF and nonreproduc- tive females (NRF) was easily made

Capture - recapture Each year (1992 - 1994) the population was monitored during the complete reproductive period from the beginning of vitellogenesis to the end of parturition in September Each year nine capture- recapture periods of 2 weeks each were carried out (see Table 1) The vipers were not observed on rainy days in early spring when they remained sheltered At each capture snakes were colour- marked on the back to avoid multiple captures during the same period and to reduce disturbance

Population estimates were obtained using CARURE (Otis et al 1978) Each 2-week period was considered a capture period The first five periods only from early March to 15 May were used for the computer procedure (see Table 1) The CARURE procedure assumes a closed population ie that no births deaths or migra- tion occur between sampling periods and is generally used for experiments covering short periods of time (Otis et al 1978) Births did not influence our analyses because only adult females were considered and newborns need 3 or 4 years to reach maturity (Saint Girons 1975) We found no evidence of migration from the studied population Mortality may have occurred between periods from March to May but annual survival rates of adult females were apparently high (gt07 unpublished data) and were probably higher if we consider only 25 months of the year Finally CARURE pro- vides the opportunity to test different models including hetero- geneity of capture probabilities in population (M) time-specific variation in probability of capture (M) behavioural response after initial capture (M) and combinations between these models

Total body length rather than SVL was used to allow compari- sons with previously published results However total length and SVL were strongly correlated (r = 098 P lt 00001 n = 78) Data are expressed as means + SD

Results

Capture - recapture differences between RF and NRF The mean number of captures per female each year was 264 f 174 (range 1-8 n = 182) for RF and 176 + 102 (range 1-5 n = 178) for NRF (Fig 1) There was a large difference between RF and NRF (ANOVA F[13591 = 3390 P lt 0001) Adult survival may be different in RF and NRF and this may have affected our results either increasing (greater survival of NRF) or decreasing (greater survival of RF) the observed difference between RF and NRF To test this possibility only females that were known to have sur- vived during the year of capture (recaptured 1 or 2 years

-ge females

Fig 1 Numbers of captures of females per year Open bars indicate nonreproductive females and solid bars reproductive females

100 1 I

No of captures per female each year

later thus excluding 1994 capture periods) were included in the analyses The results were similar the number of cap- tures was 274 f 167 (n = 35 range 1 - 8) for RF and 165 f 086 (n = 71 range 1-4) for NRF (FllMl = 1963 P lt 0001)

There was no difference in body length between RF (5535 f 375 cm) and NRF (5462 + 457 cm) (ANOVA F[13591 = 276 P gt 005) We found a positive correlation between body length and number of captures in both RF (r = 024 P lt 0001 n = 182) and NRF (r = 025 P lt 0001 n = 178) (overall correlation r = 025 P lt 0001 n = 360) To classify snakes as small medium-sized or large female body length was standardized (Z = (x- X)ISD) so that the distributions would have a mean of 0 and SD of 1 The distributions were divided into three classes small (2 lt -043) medium-sized (-043 lt Z lt 043) and large (2 gt 043) females (Marti 1990) Small females had a body length of less than 535 cm large females were greater than 565 cm and medium-sized females were in between There

Can

J Z

ool

Dow

nloa

ded

from

ww

wn

rcre

sear

chpr

ess

com

by

Cou

peri

n on

08

311

5Fo

r pe

rson

al u

se o

nly

236 Can J Zool Vol 74 1996

Table 3 Numbers of reproductive females (RF) and nonreproductive females (NRF) caught during each capture period in 1993 and 1994

March to early June Summer August to September (vitellogenesis) (pregnancy) (parturition)

1 2 3 4 5 6 7 8 9

1993 No of RF per period 8 6 14 24 32 26 23 17 15 No of NRF per period 6 6 13 34 33 20 9 3 5 RFINRF () 57 50 52 41 49 57 72 85 75

1994 NoofRFperperiod 20 29 42 42 16 22 19 18 54 No of NRF per period 24 12 36 28 10 8 9 5 16 RFINRF () 45 71 54 60 62 73 68 78 77

Note Period numbers are the same as in Table 1

were significant differences in the numbers of captures between the three categories of RF and NRF small females tend to be less visible than larger ones (Table 2)

CAPTURE was used in 1993 and 1994 1992 was excluded from this analysis because our selection was biased towards known females during this first year (neglecting several small unmarked females) Of the eight models tested during the CAPTURE procedure goodness-of-fit tests suggested that population size estimated under time variation and indi- vidual heterogeneity in capture probabilities was the best model (Mth Chao et al 1992) Population size estimates made with the Mth model were 269 f 45 (approximate 95 confidence interval 207 - 382) in 1993 and 297 f 35 (approxi- mate 95 confidence interval 242 - 385) in 1994 Individual heterogeneity in capture probabilities was probably due to the differences in sun-exposure behaviour between RF and NRF and to a lesser extent to the difference between small and large females Time variations were linked to changes in climatic conditions during the annual cycle A large pro- portion of captures (534 or 73) were made between the end of hibernation and the end of vitellogenesis (March April and May which involved the first six capture periods Table I) In summer the snakes became less visible but the probability of capture increased again in late August and early September (Table I) The mean number of captures per period was 445 f 231 (n = 12) from March to May 240 f 46 (n = 4) in summer (F131 = 845 P lt 001) and 455 (n = 2) from August to September The snakes were more visible when climatic conditions particularly low ambient temperatures increased basking In summer ambient temperatures were high and snakes reached preferred body temperatures without basking (unpublished radio-tracking data) except after a cool night or during rainy periods

Several adult females were not seen during a given year To estimate the proportion of such invisible snakes we considered that a female (with body length 2 55 cm) not seen in a given year but caught 1 year later was present in the population and was an adult at this time (we assumed a maximal growth rate of 75 cmlyear) Forty-nine RF (n = 23 caught in 1993 n = 26 in 1994) and 19 NRF (n = 1 1 caught in 1993 n = 8 in 1994) were invisible There was a differ- ence between RF and NRF ( X 2 = 610 df = 1 P = 0014) Individuals do not reproduce annually in our population (see below) therefore the 49 RF were probably NRF during

the period when they were invisible Assuming a triennial reproductive cycle (see below) the 19 NRF were either NRF (50) or RF (50) Overall NRF represented the greatest proportion (86) of invisible snakes The ratios of total estimated population size and the number of snakes caught showed that the proportion of invisible snakes was about 38 in 1993 and 36 in 1994 Consequently we assume that a large number of NRF have not been seen in 1993 and 1994

Proportion of RF calculated for each capture period (noncorrected estimates NC)

The ratios of the number of RF to the number of NRF calcu- lated for each period showed a positive bias towards RF (Table 3) Calculated means were 598 RF in 1993 653 in 1994 and 626 when 1993 and 1994 were pooled This proportion suggested an annual -biennial reproductive cycle We found 49 RF in 1992 though this sample was biased towards large and known females (see Methods) this proportion supports the assumption of an annual -biennial reproductive cycle

There was a positive correlation between the RFINRF ratio and the timing of the period (1992 excluded) (Spearmans rank correlation r = 075 P lt 0001 n = 16 Fig 2) This indicated that the probability of finding RF rather than NRF increased from the end of hibernation to parturition

Proportion of RF calculated with C A ~ R E

(catchability-corrected estimates CE) Goodness-of-fit tests suggested that population size estimated under time variation and individual heterogeneity in capture probabilities was the best model (Mth Chao et al 1992) The estimated population size of RF was 77 f 13 (95 con- fidence interval 62 - 120) in 1993 and 1 17 + 17 (95 confi- dence interval 95 - 163) in 1994 The respective proportion of RF relative to the total population was 29 in 1993 39 in 1994 and 34 in 1993 + 1994 These percentages sug- gest that the reproductive cycle of female asp vipers was close to a triennial cycle

Reproductive cycle of females monitored during 2 or 3 consecutive years (observed breeding frequency OF)

Thirty-two females marked in 1992 were recaptured in 1994 7 of these were not seen in 1993 hence 25 adult females

Can

J Z

ool

Dow

nloa

ded

from

ww

wn

rcre

sear

chpr

ess

com

by

Cou

peri

n on

08

311

5Fo

r pe

rson

al u

se o

nly

Bonnet and Naulleau

Fig 2 Relationship between capture - recapture period and ratio of reproductive females (RF) to nonreproductive females (NRF) (Spearmans rank correlation r = 075 df = 16 P lt 0001 For an explanation of period numbers see Table 1

Capture period

Table 4 Sets of capture-recapture data providing information on breeding frequencies of females monitored over 3 years

RF RF RF NRF NRF NRF

NRF RF O RF NRF NRF NRF RF NRF NRF NRF NRF NRF

Breeding frequency n

Biennial 3 Biennial 2 Triennial 4 Triennial 3 Triennial 1 Triennial 1 Quadrennial 4

Note RF reproductive female NRF nonreproductive female reproductive status not known 0 female not caught

Body length 1475 cm

were caught in 3 successive years It was not possible to determine the reproductive status of each female every year (see Methods) The sample size was small (27) and only data where the reproductive status has been determined defini- tively (detection of well-developed eggs or embryos for RF) were retained for analyses Seventy-five different females were captured -recaptured in 2 consecutive years (1 992 - 1993 and 1993-1994) Overall reproductive status was deter- mined on 99 occasions during 2 consecutive years We never observed any adult female to be reproductive in 2 consecu- tive years annual reproduction would be exceptional in our study area We therefore assumed that a female identified as RF in a given year was NRF the next year and also the previous year (if body length was 1 5 5 cm during the year of reproduction)

Different sets of capture - recapture data provided specific information about individuals reproductive cycles (Table 4)

The other data provided no information about breeding fre- quency and were excluded from the analyses (n = 14) Of females monitored over 3 years we found that 28 exhibited a biennial pattern (n = 5) 50 a triennial pattern (n = 9) and 22 a quadrennial pattern (n = 4) This suggests a mixed or opportunistic strategy situation with a 2- 3- or 4-year reproductive cycle Overall the proportion of repro- ductive females calculated for 1 year was 34 suggesting on average a triennial reproductive cycle pattern in the population studied

Discussion

Capture-recapture differences between RF and NRF Among V aspis RF exhibited much more conspicuous behaviour than NRF This difference was not related to average body size which was similar in the two groups In V berus Madsen and Shine (1994) observed that juveniles and NRF were more secretive than RF and were not caught in a particular year however their paper provided no catch- ability values Specific thermal requirements in relation to vitellogenesis and pregnancy probably explain the differ- ences in basking behaviour RF need to bask in the sun much more often and regularly than NRF probably to provide optimal temperatures for growing follicles and embryos (Naulleau 1986 Shine 1991) After winter emergence both RF and NRF (as well as males immatures and newborns) need to bask to warm their body (Naulleau et al 1987 unpublished data) Later ambient temperatures increase with day length and sun exposure so basking becomes less necessary for individuals that are not involved in such pro- cesses as vitellogenesis pregnancy sloughing or digestion (Naulleau 1983 unpublished data) As a consequence from the end of hibernation until parturition the difference in catchability between RF and NRF increases

The assumption that the thermal requirements of RF and NRF are different has been tested in a viviparous snake Crotalus viridis (Charland and Gregory 1990) It was shown that the body temperature (Tb) of RF was less variable than that of NRF and that for gravid females the mean plateau Tb increased across the season

Like C viridis V aspis is a viviparous viperid snake One may argue that differences in catchability among females due to reproduction may not occur in oviparous snakes However we suggest that the difference between RF and NRF is widespread in snakes First overrepresentation of NRF during capture periods was observed during vitello- genesis and not only during pregnancy Second there is a continuum between viviparous and oviparous species (Shine 1983) Numerous oviparous snakes keep their eggs in genital ducts after ovulation and may adopt a gravid pattern of thermal behaviour Third in several species females seek nesting sites before laying and may become much more conspicuous when searching for such nesting sites (Swain and Smith 1978 Shine 1991)

Influence of body length on capture - recaptures Large females tended to be more conspicuous than small ones Because small vipers should be more vulnerable to predators than larger ones the possibly more secretive behaviour of the smaller individuals may be related to preda- tor avoidance (Houston and Shine 1994) Several predators

Can

J Z

ool

Dow

nloa

ded

from

ww

wn

rcre

sear

chpr

ess

com

by

Cou

peri

n on

08

311

5Fo

r pe

rson

al u

se o

nly

Can J Zool Vol 74 1996

(crows pheasants magpies and buzzards) are likely to prey upon asp vipers and all were present at the study site (Cramp and Simmons 1980 Madsen and Shine 1993 T Madsen and R Shine unpublished data)

Data gathered on immatures (body length lt475 cm in females and lt 365 cm in males (the smallest size where mating has been observed)) support this hypothesis Imma- tures are generally 1 - 3 years old and their body mass ranges between 10 and 35 g Such small snakes should be extremely vulnerable to bird predators In the study area the likely total number of immatures was greater than 200 although imma- tures were rarely observed Apparently small and hence vul- nerable snakes adopted very secretive behaviour (Madsen and Shine 1994) An alternative but not exclusive hypothesis assumes that small snakes need less sun exposure to reach a high Tb owing to their small body mass or that the optimal Tb that allows physiological processes to occur is lower than in adults

Estimated breeding frequency Despite a large sample size (3 17 adult females) and a 3-year study the data presented in this paper show that the three methods for estimating breeding frequency lead to substan- tially different conclusions The NC method (noncorrected estimates 63 RF) estimated an annual -biennial repro- ductive cycle whereas both the CE (catchability-corrected estimates) and OF (observed breeding frequency) methods (34 RF) estimated a triennial reproductive cycle Capture - recapture of females monitored over several years (OF method) provides the most accurate data and we are confi- dent that a triennial reproductive cycle is most likely in the studied population This latter result was close to those obtained using very different approaches in previous studies ie a histological description of ovarian stages (Saint Girons 1957) and estimating rates of recovery of body fat for repro- duction (Saint Girons 1957 Bonnet and Naulleau 1994 1995) It is encouraging to see that the use of CAPTURE to control for differential catchability between RF and NRF gave similar results Thus in some species field studies made over a few weeks may provide satisfactory breeding frequency estimates We have identified a major problem with the NC method (the RFINRF ratio calculated at a given point in time) which is the most commonly used method in the literature (see below) The NC method greatly overestimated the number of RF This problem has recently been pointed out by Martin (1993) in a 19-year field study carried out on 927 adult female Crotalus horridus He showed that 33 was the most likely average proportion of RF in the studied population how- ever the NC method gave an overestimate of 60 probably because of the greater catchability of pregnant females

In a recent review Seigel and Ford (1987) suggested that breeding frequency was both the most important and least understood aspect of snake reproductive biology The use of inappropriate methods to determine frequency of reproduc- tion could explain why there have been few credible attempts to summarize the patterns of reproductive strategies in this group (Dunham et al 1988) In numerous studies breeding frequencies have been calculated using the proportion of RF at a given time (RFINRF ratio (NC estimate in this study) for a review see Table 82 in Seigel and Ford 1987) In addi- tion the sample size was often not large enough (52 studies

among 85 reviewed by Seigel and Ford 1987) to detect the difference between RFINRF ratios ranging from 35 to 65 (Blem 1982) and the sample size was greater than 100 in only 9 studies Several techniques are particularly sensitive to potential bias introduced by the differences in catchability and activity patterns between RF and NRF (Gibbons and Sernlitsch 1987) in for example the examination of collec- tions in museums and specimens obtained by road sampling both of which have been the basis of many field studies and cannot provide capture - recapture information (Gibbons 1972 Shine 1977 1987 Fitch 1987) Harlow and Shine (1992) suggested that breeding frequencies may be under- estimated when gravid female Pacific Island boa (genus Candoia) were more secretive than NRF Overall sampling bias can strongly affect breeding frequency estimates in snakes and until now this bias has not been included as a correction factor

Differential catchability between RF and NRF can also influence survival estimates Postparturient female V aspis adopt very secretive behaviour the year(s) after reproduction and may became invisible to snake catchers Such indi- viduals could be considered dead too soon after reproduc- tion whilst remaining alive for several years Consequently in some snake species a decrease in maternal survival as a cost of reproduction cannot be correctly correlated with fecundity growth and other trade-offs between life-history variables (Shine and Schwarzkopf 1992 Niewiarowki and Dunham 1994)

We suggest that the differential catchability between RF and NRF (which also exists between males and females) at least in species of snakes inhabiting temperate climatic zones should be regarded as an important factor influencing breeding frequency estimates

Acknowledgements

We thank L Schmidlin for his participation in the data analyses We are grateful to M Vacher L Patard and S Morley who all spent several weeks (months) in the field and who provided valuable help during capture- recapture periods We also thank C McArthur and M Hewison for help with the English and the Ministh-e Fran~ais de 17Environnement which provided official authorization (autorisations de captures et de transport danimaux d7esp2ces protegees no 8909591095 and 922 1 1)

References

Blem CR 1982 Biennal reproduction in snakes an alternative hypothesis Copeia 1982 96 1 - 963

Bonnet X and Naulleau G 1993 Relations entre la glycCmie et lactivitk saisonnibre chez Vipera aspis Amphib-Reptilia 14 295 - 306

Bonnet X and Naulleau G 1994 Utilisation dun indice de condition corporelle (BCI) pour lktude de la reproduction chez les serpents C R Acad Sci Ser I11 Sci Vie 317 34-41

Bonnet X and Naulleau G 1995 Estimation of body reserves in living snakes using a body condition index (BCI) I n Scienta Herpetologica Proceedings of the Seventh Ordinary General Meeting of Societas Europaea Herpetologica Barcelona Spain September 15 - 19 1993 Edited by G A Llorente A Montori X Santos and MA Carretero pp 237 -240

Bonnet X Naulleau G and Mauget R 1994 The influence of

Can

J Z

ool

Dow

nloa

ded

from

ww

wn

rcre

sear

chpr

ess

com

by

Cou

peri

n on

08

311

5Fo

r pe

rson

al u

se o

nly

Bonnet and Naulleau

body condition on 17-0 estradiol levels in relation to vitellogenesis in female Vipera aspis (Reptilia Viperidae) Gen Comp Endocrinol 93 424 -437

Bull JJ and Shine R 1979 Iteroparous animals that skip opportunities for reproduction Am Nat 114 296 - 303

Chao A Lee SM and Jeng SL 1992 Estimating population size for capture - recapture data when capture probabilities vary by time and individual animal Biometrics 48 201 -2 16

Charland MB and Gregory PT 1990 The influence of female reproductive status on thermoregulation in a viviparous snake Crotalus viridis Copeia 1990 1089 - 1098

Cramp S and Simmons KEL 1980 Handbook of the birds of Europe the Middle East and North Africa Oxford University Press London

Dunham AE Miles DB and Reznick DN 1988 Life history patterns in squamate reptiles In Biology of the Reptilia Vol 16 Edited by C Gans and RB Huey Alan R Liss Inc New York pp 441 -522

Fitch HS 1987 Collecting and life-history techniques In Snakes ecology and evolutionary biology Edited by R A Seigel J T Collins and SS Novak Macmillan Publishing Co New York pp 143- 164

Gibbons JW 1972 Reproduction growth and sexual dimorphism in the canebrake rattlesnake (Crotalus horridus atricaudatus) Copeia 1972 222 -226

Gibbons J W and Sernlitsch R D 1987 Activity patterns In Snakes ecology and evolutionary biology Edited by R A Seigel J T Collins and S S Novak Macmillan Publishing Co New York pp 396-421

Harlow P and Shine R 1992 Food habits and reproductive biology of the islands boas (Candoia) J Herpetol 26 60 -66

Houston D and Shine R 1994 Movements and activity pattern of Arafura Filesnakes (Serpentes Achrochordidae) in tropical Australia Herpetologica 50 349 - 357

Madsen T and Shine R 1993 Costs of reproduction in a population of European adders Oecologia 94 488 -495

Madsen T and Shine R 1994 Components of lifetime rerpoductive success in adders Vipera berus J Anim Ecol 63 56 1 - 568

Marti DM 1990 Sex and dimorphism in the barn owl and a test of mate choice Auk 107 246-254

Martin W H 1993 Reproduction of the timber rattlesnake (Crotalus horridus) in the Appalachian Mountains J Herpetol 27 133- 143

Naulleau G 1983 The effects of temperature on digestion in Vipera aspis J Herpetol 17 166 - 170

Naulleau G 1986 Effects of temperature on gestation in Vipera aspis and Vipera berus (Reptilia Serpentes) In Proceedings of the Third Ordinary General Meeting of the Societas Europea Herpetologica Prague 1986 Edited by Z Rocek pp 489 -494

Naulleau G and Bonnet X 1995 Structure of a wild population of Vipera aspis L investigated using a body condition index (BCI) In Scienta Herpetologica Proceedings of the Seventh Ordinary General Meeting of Societas Europaea Herpetologica Barcelona Spain September 15 - 19 1993 Edited by GA Llorente A Montori X Santos and MA Carretero pp 255-258

Naulleau G Fleury F and Boissin J 1987 Annual cycles in plasma testosterone and thyroxine for the male aspic viper Vipera aspis L (Reptilia Viperidae) in relation to the sexual cycle and hibernation Gen Comp Endocrinol 65 254-263

Niewiarowski PH and Dunham AE 1994 The evolution of reproductive effort in squamate reptiles costs trade-offs and assumptions reconsidered Evolution 48 137 - 145

Otis DL Burnham KP White CG and Anderson DR 1978 Statistical inference for capture data on closed animal populations Wildl Monogr No 62

Saint Girons H 1957 Le cycle sexuel chez Vipera aspis (L) dans lOuest de la France Bull Biol Fr Belg 91 284-350

Saint Girons H 1975 Critkres dige structure et dynamique des populations de reptiles In Problkmes decologie Edited by F Bourlikre and M Lamotte Masson et cie Editeurs Paris pp 233-252

Seigel RA 1993 Summary future research on snakes or how to combat lizard envy In Snakes ecology and behavior Edited by RA Seigel and JT Collins McGraw-Hill Inc New York pp 395-402

Seigel RA and Ford NB 1987 Reproductive ecology In Snakes ecology and evolutionary biology Edited by R A Seigel J T Collins and SS Novak Macmillan Publishing Co New York pp 210-252

Shine R 1977 Reproduction in australian elapid snakes 11 Female reproductive cycles Aust J Zool 25 655 -666

Shine R 1983 Reptilian reproductive modes the oviparity - viviparity continuum Herpetologica 39 1-8

Shine R 1987 Ecological ramifications of prey size food habits and reproductive biology of Australian copperhead snakes (Autrelaps Elapidae) J Herpetol 21 2 1 - 28

Shine R 1991 Australian snakes a natural history Reed Books Pty Ltd Sydney

Shine R and Schwarzkopf L 1992 The evolution of reproductive effort in lizards and snakes Evolution 46 62-75

Stearns SC 1992 The evolution of life histories Oxford University Press New York

Swain TA and Smith HM 1978 Communal nesting in Coluber constrictor in Colorado (Reptilia Serpentes) Herpetologica 34 175 - 177

Can

J Z

ool

Dow

nloa

ded

from

ww

wn

rcre

sear

chpr

ess

com

by

Cou

peri

n on

08

311

5Fo

r pe

rson

al u

se o

nly

This article has been cited by

1 Daniele Muri Johan Schuerch Naiumlke Trim Joaquim Golay Alexandre Baillifard Athimed El Taher Sylvain Dubey2015 Thermoregulation and microhabitat choice in the polymorphic asp viper (Vipera aspis) Journal of ThermalBiology [CrossRef]

2 B Castella J Golay J-C Monney P Golay K Mebert S Dubey 2013 Melanism body condition and elevationaldistribution in the asp viper Journal of Zoology 290101111jzo2013290issue-4 273-280 [CrossRef]

3 AUREacuteLIE AIumlDAM CATHERINE LOUISE MICHEL XAVIER BONNET 2013 Effect of Ambient Temperaturein Neonate Aspic Vipers Growth Locomotor Performance and Defensive Behaviors Journal of Experimental ZoologyPart A Ecological Genetics and Physiology 319101002jezv3196 310-318 [CrossRef]

4 Catherine Louise Michel Jean-Henri Pastore Xavier Bonnet 2013 Impact of cool versus warm temperatures ongestation in the aspic viper (Vipera aspis) Comparative Biochemistry and Physiology Part A Molecular amp IntegrativePhysiology [CrossRef]

5 Xavier BonnetThe Evolution of Semelparity 645-672 [CrossRef]6 John D Willson Christopher T Winne Brian D Todd 2011 Ecological and methodological factors affecting

detectability and population estimation in elusive species The Journal of Wildlife Management 75101002jwmgv75136-45 [CrossRef]

7 Hartwell H Welsh Clara A Wheeler Amy J Lind 2010 Spatial Ecology of the Oregon Gartersnake Thamnophisatratus hydrophilus in a Free-Flowing Stream Environment Copeia 2010 75-85 [CrossRef]

8 XAVIER BONNET FRANCcedilOIS BRISCHOUX DAVID PEARSON PHILIPPE RIVALAN 2009 Beach rock asa keystone habitat for amphibious sea snakes Environmental Conservation 36 62 [CrossRef]

9 MARCO A L ZUFFI AUGUSTO GENTILLI ELENA CECCHINELLI FABIO PUPIN XAVIER BONNETERNESTO FILIPPI LUCA M LUISELLI FILIPPO BARBANERA FERNANDO DINI MAURO FASOLA2009 Geographic variation of body size and reproductive patterns in Continental versus Mediterranean asp vipersVipera aspis Biological Journal of the Linnean Society 96 383-391 [CrossRef]

10 Franccedilois Brischoux Xavier Bonnet 2008 Estimating the impact of sea kraits on the anguilliform fish community(Congridae Muraenidae Ophichthidae) of New Caledonia Aquatic Living Resources 21 395-399 [CrossRef]

11 Marieke Lettink Philip J Seddon 2007 Influence of Microhabitat Factors on Capture Rates of Lizards in a CoastalNew Zealand Environment Journal of Herpetology 41 187-196 [CrossRef]

12 Xavier Santos Conchita Arenas Gustavo A Llorente Xavier Ruiz 2007 Exploring the origin of egg protein inan oviparous water snake (Natrix maura) Comparative Biochemistry and Physiology Part A Molecular amp IntegrativePhysiology 147 165-172 [CrossRef]

13 I Ineich X Bonnet F Brischoux M Kulbicki B Seacuteret R Shine 2007 Anguilliform fishes and sea kraits neglectedpredators in coral-reef ecosystems Marine Biology 151 793-802 [CrossRef]

14 Juan Manuel Pleguezuelos Xavier Santos Joseacute Carlos Brito Xavier Parellada Gustavo Adolfo Llorente SoumiaFahd 2007 Reproductive ecology of Vipera latastei in the Iberian Peninsula Implications for the conservation of aMediterranean viper Zoology 110 9-19 [CrossRef]

15 Otavio A V Marques Selma M Almeida-Santos Murilo G Rodrigues 2006 ACTIVITY PATTERNS IN CORALSNAKES GENUS MICRURUS (ELAPIDAE) IN SOUTH AND SOUTHEASTERN BRAZIL South AmericanJournal of Herpetology 1 114-120 [CrossRef]

16 Amy J Lind Hartwell H Welsh David A Tallmon 2005 GARTER SNAKE POPULATION DYNAMICSFROM A 16-YEAR STUDY CONSIDERATIONS FOR ECOLOGICAL MONITORING Ecological Applications15 294-303 [CrossRef]

17 Neil B Ford Franccedilois Brischoux Debra Lancaster 2004 REPRODUCTION IN THE WESTERNCOTTONMOUTH AGKISTRODON PISCIVORUS LEUCOSTOMA IN A FLOODPLAIN FOREST TheSouthwestern Naturalist 49 465-471 [CrossRef]

18 Luca Luiselli 2004 Meadow viper Vipera ursinii in Italy aspects of conservation ecology (Reptilia Viperidae)Italian Journal of Zoology 71 167-172 [CrossRef]

Can

J Z

ool

Dow

nloa

ded

from

ww

wn

rcre

sear

chpr

ess

com

by

Cou

peri

n on

08

311

5Fo

r pe

rson

al u

se o

nly

19 Olivier Lourdais Richard Shine Xavier Bonnet Michaeumll Guillon Guy Naulleau 2004 Climate affects embryonicdevelopment in a viviparous snake Vipera aspis Oikos 104101111oik2004104issue-3 551-560 [CrossRef]

20 Luca Luiselli 2004 Meadow viper Vipera ursinii in Italy aspects of conservation ecology (Reptilia Viperidae)Italian Journal of Zoology 71 167-172 [CrossRef]

21 Richard Shine Xavier Bonnet Harold G Cogger 2003 Antipredator Tactics of Amphibious Sea-Snakes (SerpentesLaticaudidae) Ethology 109 533-542 [CrossRef]

22 X Bonnet R Shine O Lourdais G Naulleau 2003 Measures of reproductive allometry are sensitive to samplingbias Functional Ecology 17 39-49 [CrossRef]

23 Xavier Bonnet Olivier Lourdais Richard Shine Guy Naulleau 2002 REPRODUCTION IN A TYPICAL CAPITALBREEDER COSTS CURRENCIES AND COMPLICATIONS IN THE ASPIC VIPER Ecology 83 2124-2135[CrossRef]

24 Xavier Bonnet David Pearson Mitchell Ladyman Olivier Lourdais Don Bradshaw 2002 lsquoHeavenrsquo for serpents Amark-recapture study of tiger snakes (Notechis scutatus) on Carnac Island Western Australia Austral Ecology 27442-450 [CrossRef]

25 Olivier Lourdais Xavier Bonnet Richard Shine Dale DeNardo Guy Naulleau Michael Guillon 2002 Capital-breedingand reproductive effort in a variable environment a longitudinal study of a viviparous snake Journal of Animal Ecology71 470-479 [CrossRef]

26 Robert N Reed Richard Shine 2002 Lying in Wait for Extinction Ecological Correlates of Conservation Statusamong Australian Elapid Snakes Conservation Biology 16 451-461 [CrossRef]

27 Xavier Bonnet David Pearson Mitchell Ladyman Olivier Lourdais Don Bradshaw 2002 Heaven forserpents A mark-recapture study of tiger snakes (Notechis scutatus) on Carnac Island Western Australia AustralEcology 274 442 [CrossRef]

28 Xavier Bonnet Guy Naulleau Don Bradshaw Richard Shine 2001 Changes in Plasma Progesterone in Relation toVitellogenesis and Gestation in the Viviparous Snake Vipera aspis General and Comparative Endocrinology 121 84-94[CrossRef]

29 Thomas Madsen Richard Shine 2000 Silver spoons and snake body sizes prey availability early in life influenceslong-term growth rates of free-ranging pythons Journal of Animal Ecology 69101111jae200069issue-6 952-958[CrossRef]

30 Thomas Madsen Richard Shine 2000 Silver spoons and snake body sizes prey availability early in life influenceslong-term growth rates of free-ranging pythons Journal of Animal Ecology 696 952 [CrossRef]

31 Marco AL Zuffi Federico Giudici Paolo Ioalegrave 1999 Frequency and effort of reproduction in female Vipera aspisfrom a southern population Acta Oecologica 20 633-638 [CrossRef]

32 Patrick T Gregory Lisa H Crampton Kristina M Skebo 1999 Conflicts and interactions among reproductionthermoregulation and feeding in viviparous reptiles are gravid snakes anorexic Journal of Zoology 248101111jzo1999248issue-2 231-241 [CrossRef]

33 Marco A L Zuffi Xavier Bonnet 1999 Italian subspecies of the asp viper Vipera aspis Patterns of variability anddistribution Italian Journal of Zoology 66 87-95 [CrossRef]

Can

J Z

ool

Dow

nloa

ded

from

ww

wn

rcre

sear

chpr

ess

com

by

Cou

peri

n on

08

311

5Fo

r pe

rson

al u

se o

nly