- Messages
- 3,684
<!-- m --><a class="postlink" href="http://74.6.239.67/search/cache?ei=UTF-8&p=Lacerta+lepida&fr=b1ie7&u=www.ucm.es/info/zoo/bcv/pdf/2006_JZool_268_295.pdf&w=lacerta+lepida&d=TPMtKp2uSYJp&icp=1&.intl=us" onclick="window.open(this.href);return false;">http://74.6.239.67/search/cache?ei=UTF- ... 1&.intl=us</a><!-- m -->
"Lacerta lepida"
Below is a cache of <!-- m --><a class="postlink" href="http://www.ucm.es/info/zoo/bcv/pdf/2006_JZool_268_295.pdf" onclick="window.open(this.href);return false;">http://www.ucm.es/info/zoo/bcv/pdf/2006 ... 68_295.pdf</a><!-- m -->.
Abundance, microhabitat selection and conservation of
eyed lizards (Lacerta lepida): a radiotelemetric studyJ. A. DÃ?â??Ã?´Ã?â??Ã?±az1, C. Monasterio1& A. Salvador21 Dpto. de ZoologÃ?â??Ã?´Ã?â??Ã?±a y AntropologÃ?â??Ã?´Ã?â??Ã?±a FÃ?â??Ã?´Ã?â??Ã?±sica (Vertebrados), Facultad de BiologÃ?â??Ã?´Ã?â??Ã?±a, Universidad Complutense, Madrid, Spain2 Dpto. de EcologÃ?â??Ã?´Ã?â??Ã?±a Evolutiva, Museo Nacional de Ciencias Naturales, C.S.I.C., Jos Ã?â??Ã?´e Guti Ã?â??Ã?´errez Abascal, Madrid, SpainKeywordscensus methods; detectability; lizard;radiotracking; refuge selection.CorrespondenceJos Ã?â??Ã?´e A. DÃ?â??Ã?´Ã?â??Ã?±az, Dpto. de ZoologÃ?â??Ã?´Ã?â??Ã?±a yAntropologÃ?â??Ã?´Ã?â??Ã?±a FÃ?â??Ã?´Ã?â??Ã?±sica (Vertebrados), Facultadde BiologÃ?â??Ã?´Ã?â??Ã?±a, Universidad Complutense,E-28040 Madrid, Spain. Tel:+34 91 394 5136; Fax:+34 91 394 49 47Email: <!-- e --><a href="mailto:[email protected]">[email protected]</a><!-- e --> 25 January 2005; accepted 23 June2005doi:10.1111/j.1469-7998.2005.00031.xAbstractThe utility of radiotelemetry as a tool for estimating the size and microhabitat
requirements of a population of Lacerta lepida, the largest European lacertid, was
investigated in central Spain. Population density estimates based on repeated
marking and recapture (3.2 lizards ha1) were much higher than those based online transects (0.22 lizards ha1). The probability of sighting lizards before theycould retreat into a refuge was largely increased by our ability to radiolocate them.
Rocks were selected as refuges 96% of the times, and the locations of radiotracked
lizards were much closer to rocks than randomly expected. Rocks used as retreat
sites were larger and had more crevices than those available at random, which
suggests that refuge selection was primarily determined by the need to Ã?¯Ã?¬Ã?Ând shelter
from predators. Rockrose patches, which were positively selected, might be used as
refuge-connecting corridors that combine shelter with opportunities to forage and
thermoregulate. Our results emphasize the need for using radiotelemetry to
establish baseline information on abundance and to clarify the actuality, extent
and pattern of the population declines experienced by species that may function as
key links in their ecosystems, but the wariness of which poses a serious problem for
monitoring their conservation status.IntroductionThe ectothermy and insectivorous diet of lizards allow them
to attain high densities in low productivity ecosystems
because they can successfully exploit a large prey base (small
arthropods) that most endothermic predators cannot energe-
tically afford to feed on (Regal, 1983). These high densities,
in turn, provide a resource for a myriad of higher level
predators such as raptors and diurnal carnivores. Thus, over
the long term, lizards are key links between trophic levels in
unpredictable environments, and changes in lizard popula-
tion densities can have cascading effects on other trophic
levels. This high Ã?¢ââ??¬Ã?Å?caloric capacitanceÃ?¢ââ??‰â??¢ may be particularly
important in the case of large-sized species such as the eyed
lizard Lacerta lepida, the largest European lacertid (snoutÃ?¢ââ??‰â?¬Å?
vent length and total length may reach 242 and 754 mm,
respectively) and possibly the �¯�¬â�šag-ship species for the con-
servation of European reptiles and amphibians (Corbett,
1989). This lizard is mainly found in the Iberian Peninsula,
where it is an important part of the diet of several raptors
that could optimize their energy intake by actively selecting
this large reptile prey (Mart Ã?â??Ã?´Ã?â??Ã?±n & L Ã?â??Ã?´opez, 1996). It has even
been argued that the generalized decline of L. lepida, pointed
out by several authors (Allen, 1977; Corbett, 1989; Mateo,
2002), may partly be due to its increased importance in
predatorsÃ?¢ââ??‰â??¢ diets after the dramatic decrease in rabbit Orycto-lagus cuniculuspopulations because of myxomatosis andviral pneumonia epidemics (Mart Ã?â??Ã?´Ã?â??Ã?±n & L Ã?â??Ã?´opez, 1996).However, quantitative data providing support for thesuspected declines of L. lepida populations are very few or
non-existent, whereas data collected during unsystematic
surveys may present a misleading picture of the status of the
species. In some instances, the methodological information
provided may be insufÃ?¯Ã?¬Ã?Âcient to allow reliable repetition of
population counts. Allen (1977) reported signiÃ?¯Ã?¬Ã?Âcant de-
creases in population density between 1969 and 1975 at
degraded habitats subject to human pressure, and Mateo
(2002) reported densities that have fallen dramatically (from
more than 50 lizards ha1to a vestigial presence) in sectorsof the species range, but these authors give no information
about the census methods used. Line transects along a well-
conserved open Mediterranean forest yielded a much lower
density of 1.5 lizards ha1(Cano, 1984). Thus, reliable Ã?¯Ã?¬Ã?Âelddata are needed to establish baseline information on abun-
dance and to clarify the actuality, extent and pattern of
population declines.Habitat requirements of L. lepida are also unclear, despitetheir importance for developing appropriate management
strategies. Although widespread in a variety of both uncul-
tivated and man-made habitats, it prefers sites with a
complex vertical structure of vegetation and with some
rocks (Castilla & Bauwens, 1992), and it avoids denselyJournal of Zoology 268 (2006) 295Ã?¢ââ??‰â?¬Å?301 c 2006 The Zoological Society of London295Journal of Zoology. Print ISSN 0952-8369vegetated forest tracts with no clearings (Llorente et al.,
1995). However, it may be abundant at sites with hardly any
cover of vegetation (Cheylan, Megerle & Resch, 1990) as
long as there are refuge-providing structures such as rocks,
boulders or stone piles. In fact, the removal of stone refuges
in Ã?¯Ã?¬Ã?Âelds during agricultural intensiÃ?¯Ã?¬Ã?Âcation exerts a further
pressure upon the species (Corbett, 1989), suggesting that
the availability of refuges may be an important factor for the
maintenance of viable populations.The objective of this paper is to take advantage of radio-telemetry to accomplish a double goal: (1) to compare
different methods to estimate the abundance of an L. lepida
population; and (2) to examine its microhabitat and refuge
selection patterns. Radiotelemetry provides detailed infor-
mation about habitat use by active individuals (Salvador,
Veiga & Esteban, 2004) and retreat-site selection by hid-
den ones. This, in turn, allows us to compare various census
methods by considering how they are affected by the
reduced detectability of hidden individuals. Our aim was to
identify the most suitable methods to estimate the popula-
tion size and microhabitat requirements of species that may
function as key links in their ecosystems, but the wariness of
which poses a serious problem for the design of monitoring
and management programs.Materials and methodsStudy site, radiotracking procedure and
census methodsOur study site was a deciduous Pyrenean oak Quercus
pyrenaicaforest near Navacerrada (Sierra de Guadarrama,central Spain: 401440N, 41000W; 1250 m altitude), in whichshrub patches (mainly of oak saplings and rockroses Cistus
laurifolius) are interspersed with grasslands and rocky out-crops. We established a 7 ha plot (400175 m) that wevisited 5 days a week, weather permitting, from 15 April
until 15 July 2004.We restricted our study to adult lizards because subadultsand juveniles could not be Ã?¯Ã?¬Ã?Âtted with radiotransmitters, and
because their detectability was extremely low (only two
subadults and one juvenile were observed). Lizards were
noosed using a Ã?¯Ã?¬Ã?Âshing pole between 27 April and 17 June.
Capture efÃ?¯Ã?¬Ã?Âciency was high and we were able to capture
most adult lizards at Ã?¯Ã?¬Ã?Ârst encounter. Only two individuals,
which were occasionally seen on different sampling days,
could not be noosed. Immediately after capture, the lizards
were weighed and measured [snoutÃ?¢ââ??‰â?¬Å?vent length (SVL)]. We
outÃ?¯Ã?¬Ã?Âtted lizards with 2.5 g radiotransmitter collars (Biotrack
Ltd, Wareham, UK) and released them at the site of capture
after ensuring that they had recovered from the stress
because of capture and handling, and that their escape
behavior was normal.Between 3 June (when we had already captured, radio-collared and released 12 individuals) and 17 June, we
performed seven mark-recapture sessions to obtain popula-
tion estimates. We chose the Schnabel method of repeatedmarking and recapture because it is particularly appropriate
when the study animals are scarce, difÃ?¯Ã?¬Ã?Âcult to detect or
capture, and can be obtained only in small numbers (Cox,
1985). On each new sampling day, we walked throughout
the study plot and we noted, for each lizard seen, whether it
was marked or not. If not, we captured and radiocollared it,
so that the total number of marked animals increased from
12 to 18 between 3 and 17 June. Because all marked
individuals could be radiotracked, we were certain that no
death or emigration occurred during the study period, thus
fulÃ?¯Ã?¬Ã?Âlling the main conditions for the population estimate to
be valid.We also considered the number of different individualsthat, according to radiotracking data, were located within a
1 ha section of the plot that had been censused in 2000 and
2001 in the course of a study of the home ranges of
radiocollared lizards (Salvador et al., 2004). This was done
to obtain comparable data about population density that
could be used to assess its temporal variation.As an alternative method of estimating lizard abundance,we also walked a number of parallel transects that were
regularly distributed across the study plot. Transects were
walked on 8 sunny days between 08:00 and 12:00 h (Mean
European Time). Each observer walked slowly, at a con-
stant velocity of c. 1 km h1, counting all adult lizardsdetected within a 10 m wide band. In addition, we also
recorded the lizards seen perching on rocks above the mean
height of vegetation within a 50 m wide band. Overall, we
walked a total amount of 36 km of transects.Microhabitat and refuge selectionWe used an RX-8910HE (Televilt, Lindesberg, Sweden)
radio receptor to locate radiocollared individuals between
08:00 and 15:00 h, and we registered their position with a
Garmins(Garmin Ltd, Romsey, UK) GPS 12 PersonalNavigators. We obtained data about lizard activity byradiotracking marked individuals and noting whether they
were active or hidden. Thus, our estimates are of minimum
activity, because some lizards may have sought refuge
before being detected.Locations of active lizards were used for the analyses ofmicrohabitat selection. When we spotted a radiocollared
lizard, four 10 m lines were laid out radiating from the lizard
location into the four cardinal directions. We registered the
presence or absence of grass, leaf litter, rockrose shrubs, oak
saplings and rocks at 2 m intervals along these lines. We also
used a calibrated stick to note the presence or absence of
vegetation at different heights above the ground. This
procedure allowed us to calculate the per cent cover values
for each habitat variable. Because lizards used rocks as
refuges with few exceptions, we noted the distance to the
nearest rock that had a maximum diameter 440 cm and at
least one crevice that could be entered by an adult lizard. We
also noted the mean distance to the nearest potential refuge
in each of the four quadrants deÃ?¯Ã?¬Ã?Âned by the plant-cover lines.We characterized refuges by noting the maximum length,width and height of each rock or rocky outcrop whereJournal of Zoology 268 (2006) 295Ã?¢ââ??‰â?¬Å?301 c 2006 The Zoological Society of London296Radiotelemetry and lizard conservationJ. A. DÃ?â??Ã?´Ã?â??Ã?±az, C. Monasterio and A. Salvadorhidden lizards were radiotracked. We also noted the number
of crevices that could be entered by an adult lizard. Although
some refuges were used several times by the same or different
individuals, we counted these refuges only once to avoid
pseudoreplication.To measure the availability of microhabitats and refuges,we selected 33 points using a random number table applied
to latitude and longitude values within the ranges set by the
GPS locations of radiotracked lizards. We used the GPS to
determine the exact location of these points, and we mea-
sured all habitat variables as described previously. We
characterized the rocks 440 cm with at least one crevice
that were closest to each random point, to obtain a null
hypothesis of refuge availability.When necessary, habitat and refuge variables were arc-sine- or log-transformed to fulÃ?¯Ã?¬Ã?Âll the requirements of para-
metric tests. We performed a principal components analysis
to reduce the number of habitat variables, rotating the
factors with eigenvalues greater than one (Varimax rota-
tion) to obtain a clearer pattern of loadings.ResultsPopulation density and detectability of
lizardsDuring the study period, we captured 18 adult lizards (eight
males and 10 females); the mean SVL (1SE) was158.33.3 mm (range = 142Ã?¢ââ??‰â?¬Å?190), and the mean body masswas 94.87.0 g (range = 60Ã?¢ââ??‰â?¬Å?155).According to the Schnabel method, the adult populationsize (1SE) was 22.51.3 lizards (Fig. 1). Because theestimate of population size remained relatively constant
after the third mark-recapture day, had a relatively small
standard error and was close to the actual number of lizards
observed, we are conÃ?¯Ã?¬Ã?Âdent that this Ã?¯Ã?¬Ã?Âgure is a robust
approximation to the number of adult lizards present at the
study plot during the census period. Thus, we obtained a
population density of 3.210.19 lizards ha1. The 1 ha sec-tion of the plot previously censused in 2000 and 2001
included radiolocations of seven individuals, i.e. less than
in 2000, when Salvador et al. (2004) captured 10 individualsZ130 mm in SVL at the same plot, but more than in 2001,when only Ã?¯Ã?¬Ã?Âve individuals were captured (partly because of
the difÃ?¯Ã?¬Ã?Âculty in recapturing lizards with the Ã?¯Ã?¬Ã?Âshing pole).These results contrast with the ones obtained by meansof transect counts. Considering the 10-m wide census band,
the mean lizard abundance (1SE) was 0.220.08lizards ha1, whereas considering the 50-m wide censusband, it was 0.120.03 lizards ha1. Thus, although somelizards could be seen perching on rocks, the overall detect-
ability decreased between 5 and 25 m each side of the
progression line. We conclude that the transect method was
able to detect only a small fraction of the lizards known to
be present. This could be because of a high proportion of
inactive individuals, but our data indicate that 69% of the
observations of radiocollared lizards corresponded to active
animals, and that the visual detectability of active lizards
must therefore be regarded as very low.Microhabitat and refuge selectionLizards were selective in their use of microhabitats as most
variables showed signiÃ?¯Ã?¬Ã?Âcant differences between the loca-
tions of active lizards and the sample of randomly selected
sites (Table 1). The principal components analysis with the
habitat variables produced four principal components (PCs)
that accounted for 72.4% of the variance (Table 2). The Ã?¯Ã?¬Ã?Ârst
component (PC-1) was positively correlated with rock cover
and negatively correlated with the distance to the nearest
refuge. The second component (PC-2) showed a positive
correlation with the cover of oak saplings and with plant
cover 25 and 50 cm above the ground, and a negative
correlation with plant cover 5 cm above the ground. The
third component (PC-3) gave high scores to sites with high
values of rockrose cover leaf litter cover and plant cover 75
and 100 cm above the ground. The fourth component (PC-
4) gave high scores to grasslands with a high cover of herbs
5 cm above the ground and with a low cover of rocks.Lizards actively selected sites with high scores in PC-1(t94= 13.22, Po0.001; Fig. 2), and, to a lesser extent, also inPC-3 (t94= 2.37, P = 0.020; Fig. 2). Thus, lizards tended toremain close to refuges and to select sites with a high rock
cover or a high cover of rockrose patches that could also
provide some shelter. This is basically consistent with the
results of univariate tests after applying the sequential
Bonferroni correction (Table 1).Interindividual differences in microhabitat use were sig-niÃ?¯Ã?¬Ã?Âcant along all axes except for PC-1 (Table 3; one female
was excluded because it only had two observations). This
probably re�¯�¬â�šects structural and/or �¯�¬â�šoristic differences
among the home ranges of different individuals. Thus,
interindividual differences were particularly clear for the
gradient associated with rockrose cover (PC-3), which was
negatively correlated with longitude for the randomly se-
lected sites (r = 0.433, n = 33, P = 0.012); the mean PC-3
scores of radiotracked lizards were negatively correlatedPopulation estimate12141618202224263 June4 June8 June9 June 11 June 15 June 17 JuneFigure 1 Variation of Lacerta lepida population estimates (1SE)according to the Schnabel method during seven mark-recapture days.The line shows the curve Ã?¯Ã?¬Ã?Âtted by least squares. The Ã?¯Ã?¬Ã?Ânal (17 June)population estimate has 95% conÃ?¯Ã?¬Ã?Âdence limits of 22.52 individuals.Journal of Zoology 268 (2006) 295Ã?¢ââ??‰â?¬Å?301 c 2006 The Zoological Society of London297Radiotelemetry and lizard conservationJ. A. DÃ?â??Ã?´Ã?â??Ã?±az, C. Monasterio and A. Salvadorwith the mean longitude of their radiolocations (r = 0.654,
n= 17, P = 0.004). Nevertheless, interindividual differenceswere not signiÃ?¯Ã?¬Ã?Âcant for PC-1 (Table 3), showing that despite
signiÃ?¯Ã?¬Ã?Âcant longitudinal variation in rock cover (correlation
between longitude and PC-1 scores for randomly selected
sites: r = 0.419, n = 32, P = 0.017), all lizards were similar in
their tendency to remain close to rocks (F16, 46= 0.99 andP= 0.478 for interindividual differences in the mean dis-tance to the nearest rock).Only three of 78 refuges used by radiocollared lizards(3.8%) were not rocks (one bramble patch and two dense
rockrose patches); these observations were excluded from
the analyses. Rocks used by lizards as retreat sites were
larger and had more crevices than randomly chosen ones
(Table 4; it should be noted that this is a conservative
analysis because some of the larger outcrops, which were
used several times, were counted only once).DiscussionOur data produced two important results. First, the esti-
mates of lizard abundance were strongly dependent on theTable 1 Characteristics of microhabitats used by 18 radiotracked adult Lacerta lepida, and of 33 randomly selected sitesLizardsRandom sitesMeanSDnMeanSDntPDistance to nearest rock (m)0.81.1656.87.73311.23o0.001Ã?Æ?Ã?â??Mean distance to four nearest rocks (m)3.42.06611.78.0329.21o0.001Ã?Æ?Ã?â??Rockrose cover (%)11.313.1675.98.8332.160.033Leaf litter cover (%)25.416.86735.919.7332.630.010Cover of herbs (%)62.214.66768.515.4332.210.029Cover of rocks (%)26.012.7679.510.4337.74o0.001Ã?Æ?Ã?â??Cover of oak saplings (%)34.818.56735.621.4330.110.916Plant cover 5 cm in height (%)38.026.66725.812.4331.850.067Plant cover 25 cm in height (%)42.415.76746.518.9331.040.300Plant cover 50 cm in height (%)29.014.16721.812.0332.520.013Plant cover 75 cm in height (%)14.77.86712.311.9331.760.082Plant cover 100 cm in height (%)11.49.4673.65.5334.36o0.001Ã?Æ?Ã?â??Results of t-tests and associated signiÃ?¯Ã?¬Ã?Âcance levels are also shown; asterisks indicate differences that remain signiÃ?¯Ã?¬Ã?Âcant after applying thesequential Bonferroni correction.Table 2 Principal component (PC) analysis with the microhabitat datadescribed in Table 1PC-1PC-2PC-3PC-4Distance to nearest rock (m)0.8900.1510.1670.046Mean distance to fournearest rocks (m)0.8550.0950.2100.128Rockrose cover (%)0.0760.3840.7200.019Leaf litter cover (%)0.4970.0700.5710.180Cover of herbs (%)0.1970.1510.1070.868Cover of rocks (%)0.7600.0590.1800.526Cover of oak saplings (%)0.1340.7120.0150.052Plant cover 5 cm in height (%)0.2100.5670.2200.604Plant cover 25 cm in height (%)0.1510.6540.3120.387Plant cover 50 cm in height (%)0.2560.7930.0390.069Plant cover 75 cm in height (%)0.0820.3680.6520.194Plant cover 100 cm in height (%)0.2140.1450.7920.192Eigenvalue2.592.252.161.68Explained variance0.2160.1880.1800.140Loadings with absolute values greater than 0.5 are shown in bold.PC1 scores (proximity to rocks)Ã?¢Ã?â? ââ?¬â?¢1.6Ã?¢Ã?â? ââ?¬â?¢1.2Ã?¢Ã?â? ââ?¬â?¢0.8Ã?¢Ã?â? ââ?¬â?¢0.40.00.40.81.2Lizard locationsRandom sitesPC3 scores (cover of rockrose shrubs) Ã?¢Ã?â? ââ?¬â?¢1.4Ã?¢Ã?â? ââ?¬â?¢1.0Ã?¢Ã?â? ââ?¬â?¢0.6Ã?¢Ã?â? ââ?¬â?¢0.20.20.61.0Lizard locationsRandom sitesFigure 2 Mean scores (95% conÃ?¯Ã?¬Ã?Âdence interval) of lizard locationsand randomly selected sites on principal component (PC)-1 (a gradientof proximity to rock cover) and PC-3 (a gradient of development of theshrub layer, mainly of rockroses Cistus laurifolius, 75Ã?¢ââ??‰â?¬Å?100 cm abovethe ground). Factor loadings are given in Table 2.Journal of Zoology 268 (2006) 295Ã?¢ââ??‰â?¬Å?301 c 2006 The Zoological Society of London298Radiotelemetry and lizard conservationJ. A. DÃ?â??Ã?´Ã?â??Ã?±az, C. Monasterio and A. Salvadorcensus method used. Second, microhabitat selection was
dominated by the importance of rocks as retreat sites. In this
section, we discuss these two aspects separately, we empha-
size the utility of radiotelemetry for achieving reliable
estimates of abundance and habitat requirements and we
consider the implications of our results for assessing popula-
tion trends and providing possible management strategies.Estimates of abundanceTransect counts are easy, inexpensive and not very time-
consuming (we spent 36 person-hours walking transects,
and more effort would not have increased the effectiveness
of our censuses). This method may be useful for obtaining
indexes that allow to compare the population densities of
different species in open habitats (Germaine & Wakeling,
2001; Garc Ã?â??Ã?´Ã?â??Ã?±a & Whalen, 2003) or to estimate the abundance
of common species across different habitats (D Ã?â??Ã?´Ã?â??Ã?±az & Car-
rascal, 1991) or microhabitats (Mart Ã?â??Ã?´Ã?â??Ã?±n & Salvador, 1997),
but it is of limited utility for censusing scarce, elusive
animals such as eyed lizards. Thus, abundance estimates
based on line transects were much lower than those based on
the Schnabel method. Transects carried out during occa-
sional visits, such as the ones that could take place in the
context of a larger scale survey, could even suggest that
lizards were absent from the study area (no lizards were
detected within the 10-m wide census band in 3 of the
8 census days).Capturing and radiotagging lizards were expensive andtime-consuming (c. 80 person-hours distributed over 20
days, with a cost of 150h per radiotransmitter) because of
the complexity of the habitat and the wariness of the
animals. However, mark-recapture statistical methods were
necessary to obtain a reliable estimate of population size.This can be attributed to an extremely low detectability,
shown by the fact that only radiotagged lizards could be
located with some ease, and by the decrease in the number of
sightings between the 10 and 25-m wide census bands.
Radiotracking may be replaced by other marking proce-
dures, but at the cost of losing information on retreat sites
and of not assessing the fulÃ?¯Ã?¬Ã?Âllment of the assumption that
the population remains constant throughout the study
period. We therefore encourage the use of the Schnabel
method, combined with radiotelemetry, as a reliable proce-
dure to measure population size. We also emphasize several
precautions. First, data should be collected until standard
errors are approximately coincident with the number of
lizards seen but not captured. Second, population counts
should be carried out on areas large enough (e.g. Z5 ha) to
counteract the large size of lizard home ranges (Salvador
et al., 2004) and the patchy distribution of lizard locations(with sectors where no observations were made). Finally,
repeated censuses aimed to estimate population trends
should not be undertaken at very short intervals, because
lizards are very difÃ?¯Ã?¬Ã?Âcult to noose the second time, and 3Ã?¢ââ??‰â?¬Å?4
years are approximately the time needed to have most of the
population replaced (only a 7-year-old male captured in
2004 was a 2000/2001 survivor).The population density obtained in this study was lowerthan those reported for Mediterranean dehesas with a well-
developed undergrowth of shrubs (Mart Ã?â??Ã?´Ã?â??Ã?±n & L Ã?â??Ã?´opez, 2002;
Mateo, 2002). This may be because of the lack of tree
management and scarce livestock grazing supported by
Pyrenean oaks at our study site, whereas habitat suitability
for L. lepida is expected to be highest at intermediate stages
of forest degradation (Santos & Teller Ã?â??Ã?´Ã?â??Ã?±a, 1989; Llorente
et al., 1995). In agreement with this hypothesis, radiotrackedindividuals selected sites with a high cover of rockrose
shrubs, which are dominant in south-facing forest clearings.
Thus, although agricultural intensiÃ?¯Ã?¬Ã?Âcation is deleterious for
L. lepidapopulations (Corbett, 1989; Cheylan & Grillet,2005), the increase in forest cover associated with the
abandonment of traditional agro-silvo-pastoral techniques
may also have negative effects (Cheylan & Grillet, 2005).
Similar direct relationships between anthropogenic forest
clearance and population density or habitat use have been
reported for other large-bodied heliothermic lizards (Vitt
et al., 1998; Sartorius, Vitt & Colli, 1999).Microhabitat and retreat-site selectionOur data suggest that the availability of refuges was the
most important single factor determining the quality of a
habitat for L. lepida. Thus, rocks were selected as refuges by
a vast majority (495%) of hidden lizards, and lizards
remained on average 8.5 times closer to rocks than expected
at random. The importance of retreat-site selection for
ectotherms has been emphasized by several studies (Chris-
tian, Tracy & Porter, 1984; Huey et al., 1989; Huey, 1991;
Schlesinger & Shine, 1994; Webb & Shine, 2000; Sabo, 2003)
that have pointed out that most ectotherms actually spend
longer periods in retreats than above ground (Huey, 1982).Table 3 ANOVAs with interindividual differences in microhabitat use,as measured by the scores of radiotracked lizards on the principalcomponents shown in Table 2SS effect(among individuals)SS error(within individuals)F16, 46PPC-17.6612.581.750.070PC-229.6942.062.030.031PC-330.7032.922.680.005PC-429.1835.052.390.011PC, principal component.Table 4 Characteristics of rocks used as hiding refuges by radio-tracked lizards, and of 32 randomly selected rocksLizardsÃ?¢ââ??‰â??¢ rocksRandom rocksMeanSDnMeanSDntPNumber of crevices2.91.5741.80.832 4.39o0.001Height (cm)1268975583032 4.99o0.001Width (cm)1239175763832 3.250.002Length (cm)202156751145032 4.11o0.001Results of t-tests and associated signiÃ?¯Ã?¬Ã?Âcance levels are also shown.Journal of Zoology 268 (2006) 295Ã?¢ââ??‰â?¬Å?301 c 2006 The Zoological Society of London299Radiotelemetry and lizard conservationJ. A. DÃ?â??Ã?´Ã?â??Ã?±az, C. Monasterio and A. SalvadorThus, refuge selection can have a major impact on the
thermal physiology and ecology of ectotherms (Huey et al.,
1989; Huey, 1991; Goldsbrough, Hochuli & Shine, 2003).Nevertheless, we suspect that refuge selection by L. lepidawas not primarily related to the thermal properties of
retreats. Huey et al. (1989), in a detailed study of the thermal
consequences of retreat site selection by garter snakes
(Thamnophis sirtalis), carried out during midsummer at a
latitude and altitude similar to the ones reported here, found
that rock height, rather than shape or mass, was the primary
determinant of the daily thermal cycles under rocks. More-
over, thick boulders (443 cm) offered temperatures that
never reached the lower limit of the snakesÃ?¢ââ??‰â??¢ preferred range
(28 1C; Scott, Tracy & Pettus, 1982), and were used much
less frequently than warmer rocks of intermediate thickness
(20Ã?¢ââ??‰â?¬Å?40 cm). This is in contrast with our results, because 88%
of the rocks used as retreat sites were Z45 cm thick. Not
unexpectedly, animals captured soon after emergence from
their refuges were fairly cool (personal observation), despite
the high body temperatures (30Ã?¢ââ??‰â?¬Å?35 1C) exhibited by Ã?¯Ã?¬Ã?Âeld-
active lizards (Busack & Visnaw, 1989). We therefore
hypothesize that the selection of large rocks (length, width
and thickness were highly correlated, all P0.001) wasprimarily determined by the need to Ã?¯Ã?¬Ã?Ând shelter from
predators. This would be consistent with the high predation
pressure to which L. lepida is exposed (Mart Ã?â??Ã?´Ã?â??Ã?±n & L Ã?â??Ã?´opez,
1996; see Salvador et al., 2004 for data on predation at the
study site), and that has been hypothesized as one of the
causes of its suspected decline. The importance of refuges
has also been noted at the Crau steppe in southern France, a
hard soil plain where eyed lizards were restricted to speciÃ?¯Ã?¬Ã?Âc
areas with stone piles (built during the Second World War to
impede the landing of Allied aircraft), which were the only
retreat sites available for adult lizards (Mateo, 2004). The
preference of lizards for rocks with several crevices should
also be useful for eluding predators, because it could facil-
itate both entry and exit from retreat sites. Larger rocks
provide deeper crevices, which may offer a more stable
microenvironment (Huey et al., 1989; Kearney, 2002; Beck
& Jennings, 2003). Also, the positive selection of rockrose
patches with a high plant cover 75Ã?¢ââ??‰â?¬Å?100 cm above the ground
suggests that these patches may be used as refuge-connect-
ing corridors that provide shelter to foraging and thermo-
regulating lizards.Concluding remarksRadiotelemetry was an invaluable tool to assess the popula-
tion density and spatial ecology of L. lepida. Radiotransmit-
ters allowed us to conÃ?¯Ã?¬Ã?Ârm the assumptions of most mark-
recapture methods, to achieve reliable estimates of popula-
tion size, and to obtain data on retreat-site selection (Huey
et al., 1989; Beck & Jennings, 2003; Whitaker & Shine,2003). Moreover, radiotelemetry was also essential for an
adequate characterization of microhabitat preferences, be-
cause the probability of sighting marked lizards before they
could seek shelter in a nearby refuge was largely increased
by our ability to radiolocate them. An additional advantageof radiotracking is that it facilitates assessment of interindi-
vidual differences in microhabitat or retreat-site selection,
which may be useful for understanding the patterns of space
use (Whitaker & Shine, 2003; Salvador et al., 2004). In our
study, for instance, interindividual differences in microhabi-
tat use, which were signiÃ?¯Ã?¬Ã?Âcant for most variables, were
seemingly related to differences in habitat structure within
the study area. However, there were no signiÃ?¯Ã?¬Ã?Âcant inter-
individual differences in PC-1, the habitat axis on which
selection was the strongest. This suggests that all indivi-
duals, independent of other characteristics of their home
ranges, remained as close as possible to rocks, which
reinforces the role of retreat-sites as a limiting factor for
eyed lizards (Mateo, 2004).We can therefore conclude that adequate management ofretreat sites is crucial for the conservation of L. lepida, and
that the removal of stone and boulder refuges in Ã?¯Ã?¬Ã?Âelds
associated with agricultural intensiÃ?¯Ã?¬Ã?Âcation should be avoided
(Corbett, 1989). We suggest adding artiÃ?¯Ã?¬Ã?Âcial shelter sites to
restore degraded habitats, a procedure that has been success-
ful for several species of endangered reptiles (Hecnar &
McCloskey, 1998; Webb & Shine, 2000; Mateo, 2004; Sou-
ter, Bull & Hutchinson, 2004). We also encourage the use of
radiotelemetry to monitor the demographic trends of the
largest European lacertid, because its wariness, rapidity, and
close association with retreat sites make its visual detectabil-
ity unexpectedly low. Although we found no clear evidence
of decline at our relatively well-conserved study area, it is
important to conÃ?¯Ã?¬Ã?Ârm the extent to which human pressure has
led to a substantial decrease of its populations over most of
its distribution range (Allen, 1977; Mateo, 2002).AcknowledgementsThis research was funded by DGESIC projects BOS2001-
0533 and CGL2004-01151/BOS. Permission for the capture
and radiotracking of lizards was provided by Agencia de
Medio Ambiente, Comunidad de Madrid. The Ã?¢ââ??¬Ã?Å?El Ventor-
rilloÃ?¢ââ??‰â??¢ Field Station (Museo Nacional de Ciencias Naturales,
CSIC) supplied logistic support. We thank J. P. Veiga for
the helpful comments on a previous draft.ReferencesAllen, A. (1977). Changes in population density of the Eyedlizard, Lacerta lepida, at three localities in Portugal
between 1969 and 1975. Br. J. Herpetol. 5, 661Ã?¢ââ??‰â?¬Å?662.Beck, D. & Jennings, R. (2003). Habitat use by Gila Mon-sters: the importance of shelters. Herpetol. Monogr. 17,
111Ã?¢ââ??‰â?¬Å?129.Busack, S. & Visnaw, J.A. (1989). Observations on the naturalhistory of Lacerta lepida in C Ã?â??Ã?´adiz province, Spain.
Amphibia-Reptilia 10, 201Ã?¢ââ??‰â?¬Å?213.Cano, C. (1984) La comunidad de lac Ã?â??Ã?´ertidos de un encinarcontinental. Ciclo anual de actividad Graduate thesis,Universidad Complutense, Madrid.Journal of Zoology 268 (2006) 295Ã?¢ââ??‰â?¬Å?301 c 2006 The Zoological Society of London300Radiotelemetry and lizard conservationJ. A. DÃ?â??Ã?´Ã?â??Ã?±az, C. Monasterio and A. SalvadorCastilla, A. & Bauwens, D. (1992). Habitat selection by thelizard Lacerta lepida in a Mediterranean oak forest.
Herpetol. J. 2, 27Ã?¢ââ??‰â?¬Å?30.Cheylan, G., Megerle, A. & Resch, J. (1990). La Crau, steppevivante. Guide du naturaliste dans le d Ã?â??Ã?´esert provencÃ?â??Ã?¸al.Radolfzell: Ed. J Ã?â??Ã?¨urgen Resch.Cheylan, M. & Grillet, P. (2005) Statut pass Ã?â??Ã?´e et actuel dul Ã?â??Ã?´ezard ocell Ã?â??Ã?´e (Lacerta lepida, sauriens, lacertid Ã?â??Ã?´es) en
France. Implication en termes de conservation. Vie et
Milieu, 55, 15Ã?¢ââ??‰â?¬Å?30.Christian, K.A., Tracy, C.R. & Porter, W.P. (1984). Physio-logical and ecological consequences of sleeping-site
selection by the Galapagos land iguana (Conolopus
pallidus). Ecology 65, 752Ã?¢ââ??‰â?¬Å?758.Corbett, K. (1989). The conservation of European reptiles andamphibians. London: Christopher Helm.Cox, G.W. (1985). Laboratory manual of general ecology. 5thedn. Dubuque, IA: William C. Brown.D Ã?â??Ã?´Ã?â??Ã?±az, J.A. & Carrascal, L.M. (1991). Regional distribution ofa mediterranean lizard: inÃ?¯Ã?‰â?¬Å¡uence of habitat cues and prey
abundance. J. Biogeogr. 18, 291Ã?¢ââ??‰â?¬Å?297.Garc Ã?â??Ã?´Ã?â??Ã?±a, A. & Whalen, D.M. (2003). Lizard communityresponse to a desert shrubland-intertidal transition zone on
the coast of Sonora, Mexico. J. Herpetol. 37, 378Ã?¢ââ??‰â?¬Å?382.Germaine, S.S. & Wakeling, B.F. (2001). Lizard speciesdistributions and habitat occupation along an urban gra-
dient in Tucson, Arizona, USA. Biol. Conserv. 97, 229Ã?¢ââ??‰â?¬Å?237.Goldsbrough, C.L., Hochuli, D.F. & Shine, R. (2003). In-vertebrate biodiversity under hot rocks: habitat use by the
fauna of sandstone outcrops in the Sydney region. Biol.
Conserv. 109, 85Ã?¢ââ??‰â?¬Å?103.Hecnar, S.J. & McCloskey, R.T. (1998). Effects of humandisturbance on Five-Lined Skink, Eumeces fasciatus,
abundance and distribution. Biol. Conserv. 85,
213Ã?¢ââ??‰â?¬Å?222.Huey, R.B. (1982). Temperature, physiology, and the ecologyof reptiles. In Biology of the Reptilia, Vol. 12, Physiology
(C): 25Ã?¢ââ??‰â?¬Å?29. Gans, C. & Pough, F.H. (Eds). London:Academic Press.Huey, R.B. (1991). Physiological consequences of habitatselection. Am. Nat. 137, S91Ã?¢ââ??‰â?¬Å?S115.Huey, R.B., Peterson C, .R., Arnold, S.J. & Porter, W.P.(1989). Hot rocks and not-so-hot rocks: retreat site selec-
tion by garter snakes and its thermal consequences. Ecol-
ogy 70, 931Ã?¢ââ??‰â?¬Å?944.Kearney, M. (2002). Hot rocks and much-too-hot rocks:seasonal patterns of retreat-site selection by a nocturnal
ectotherm. J. Therm. Biol. 27, 205Ã?¢ââ??‰â?¬Å?218.Llorente, G.A., Montori, A., Santos, X. & Carretero, M.A.(1995). Atlas de distribuci Ã?â??Ã?´o dels AnÃ?¯Ã?¬Ã?Âbis y Re`ptils de
Catalunya y Andorra. Figueres: El Grau.Mateo, J.A. (2002). Lacerta lepida Daudin, 1802. Lagartoocelado. In Atlas y Libro Rojo de los AnÃ?¯Ã?¬Ã?Âbios y Reptiles de
Espa Ã?â?¹Ã?â??na: 225Ã?¢ââ??‰â?¬Å?227. Pleguezuelos, J.M., M Ã?â??Ã?´arquez, R. &
Lizana, M. (Eds). Madrid: DGCN-Asociaci Ã?â??Ã?´on Herpe-
tol Ã?â??Ã?´ogica Espa Ã?â?¹Ã?â??nola.Mateo, J.A. (2004). Lagarto ocelado Ã?¢ââ??‰â?¬Å?Lacerta lepida.In Enciclopedia Virtual de los Vertebrados Espa Ã?â?¹Ã?â??noles.
Carrascal, L.M. & Salvador, A. (Eds). Madrid:
Museo Nacional de Ciencias Naturales. http://
<!-- m --><a class="postlink" href="http://www.vertebradosibericos.org/reptiles/laclep.htmlMart" onclick="window.open(this.href);return false;">http://www.vertebradosibericos.org/rept ... p.htmlMart</a><!-- m --> Ã?â??Ã?´Ã?â??Ã?±n, J. & L Ã?â??Ã?´opez, P. (1996). Avian predation on a largelizard (Lacerta lepida) found at low population densities in
Mediterranean habitats: an analysis of bird diets. Copeia 3,
722Ã?¢ââ??‰â?¬Å?726.Mart Ã?â??Ã?´Ã?â??Ã?±n, J. & L Ã?â??Ã?´opez, P. (2002). The effect of Mediterraneandehesa management on lizard distribution and
conservation. Biol. Conserv. 108, 213Ã?¢ââ??‰â?¬Å?219.Mart Ã?â??Ã?´Ã?â??Ã?±n, J. & Salvador, A. (1997). Microhabitat selection bythe Iberian rock lizard Lacerta monticola: effects on density
and spatial distribution of individuals. Biol. Conserv. 79,
303Ã?¢ââ??‰â?¬Å?307.Regal, P.J. (1983). The adaptive zone and behaviour oflizards. In Lizard ecology: studies of a model organism:
105Ã?¢ââ??‰â?¬Å?118. Huey, R.B., Pianka, E.R. & Schoener, T.W.
(Eds). Cambridge, MA: Harvard University Press.Sabo, J.L. (2003). Hot rocks or no hot rocks: overnight retreatavailability and selection by a diurnal lizard. Oecologia
136, 329Ã?¢ââ??‰â?¬Å?335.Salvador, A., Veiga, J.P. & Esteban, M. (2004). Prelimi-nary data on reproductive ecology of Lacerta lepida
at a mountain site in Central Spain. Herpetol. J. 14,
47Ã?¢ââ??‰â?¬Å?49.Santos, T. & Teller Ã?â??Ã?´Ã?â??Ã?±a, J.L. (1989). Preferencias de h Ã?â??Ã?´abitat yperspectivas de conservaci Ã?â??Ã?´on en una comunidad de
lac Ã?â??Ã?´ertidos en medios cerealistas del centro de Espa Ã?â?¹Ã?â??na. Rev.
Esp. Herpetol. 3, 259Ã?¢ââ??‰â?¬Å?273.Sartorius, S.S., Vitt, .L.J. & Colli, G.R. (1999). Use ofnaturally and athropogenically disturbed habitats in
Amazonian rainforests by the teiid lizard Ameiva ameiva.
Biol. Conserv. 90, 91Ã?¢ââ??‰â?¬Å?101.Schlesinger, C.A. & Shine, R. (1994). Choosing a rockerspectives of a bush-rock collector and a saxicolous
lizard. Biol. Conserv. 67, 49Ã?¢ââ??‰â?¬Å?56.Scott, J.R., Tracy, C.R. & Pettus, D. (1982). A biophysicalanalysis of daily and seasonal utilization of climate space
by a montane snake. Ecology 63, 482Ã?¢ââ??‰â?¬Å?493.Souter, N.J., Bull, C.M. & Hutchinson, M.N. (2004). Addingburrows to enhance a population of the endangered pygmy
blue tongue lizard Tiliqua adelaidensis. Biol. Conserv. 116,
403Ã?¢ââ??‰â?¬Å?408.Vitt, L.J., Avila-Pires, T.C.S., Caldwell, J.P. & Oliveira,V.R.L. (1998). The impact of individual tree harvesting on
thermal environments of lizards in Amazonian rain forests.
Conserv. Biol. 12, 654Ã?¢ââ??‰â?¬Å?664.Webb, J.K. & Shine, R. (2000). Paving the way for habitatrestoration: can artiÃ?¯Ã?¬Ã?Âcial rocks restore degraded habitat of
endangered reptiles? Biol. Conserv. 92, 93Ã?¢ââ??‰â?¬Å?99.Whitaker, P.B. & Shine, R. (2003). A radiotelemetric study ofmovements and shelter-site selection by free-ranging
brownsnakes (Pseudonaja textilis, Elapidae). Herpetol.
Monogr. 17, 130Ã?¢ââ??‰â?¬Å?144.Journal of Zoology 268 (2006) 295Ã?¢ââ??‰â?¬Å?301 c 2006 The Zoological Society of London301Radiotelemetry and lizard conservationJ. A. DÃ?â??Ã?´Ã?â??Ã?±az, C. Monasterio and A. Salvador
