Journal of Uerpellog, Vol. 35, No. 2, pp. 282-2t32, 200 Copyright 2.01 Socity for the Study of Anmpiibians and Reptiles

Ecology and Behavior of Lizards of the Parthenogenetic Cnemidophorus laredoensis Complex and Their Gonochoristic Relative Cnemidophorus gularis: Implications for Coexistence MARK A. PAULISSEN Department of Biological and Environnmental Sciences, MsNeese State University,.Lake Charles, Louisiana 70609, LISA E-mnail: tnpauiiss@ftmail. ncnteese.edu ABSTRACr.-The Cneinidophorus laredoensis complex consists of two, all-female parthenogenetic lizard species, designated LAR-A and LAR-B, that commonly coexist with their gonochoristic (= bisexual) relative Cnernidophorus gularis in southern Texas. The ecology and behavior of these Cuemidophorus lizards was studied in a Texas state park where LAR-A, [AR-B, and C. gularis coexist. I used time of capture records to estimate daily activity periods and used focal animal observations of free-ranging individuals to determine what proportion of time lizards spent using different microhabitats, occupying different exposures (full sun, partial sun, shade), and performing different behaviors during the May-June breeding seasons of 19941996. There was little difference among the three species in daily activity period, microhabitat use, or exposure occupancy; added to previously published data on diet similarity, these results suggest that competition between the two parthenogens and C. ,goLaris potentially is severe. The greatest behavioral differences were between male C. guclaris and female C. gularis, LAR-A, and LAR-B: male C. golarisspent a much greater percentage of their time interacting with other lizards and moved significantly farther per unit time than did females. These results reflect the large proportion of time male C. go laris devoted to trailing and courting female C. guaris as well as parthenogenetic females, especially LAR-A. The only major difference between C. gularis and the two parthenogens was that the latter spent a greater proportion of their time searching for prey. Despite this, LAR-A and LAR-B did not achieve a greater prey capture rate, nor a greater success at locating the preferred prey of termites than did C. ,Sularis,suggesting the two parthenogens are less efficient foragers than their gonochoristic relative. This may account for the reason why the two parthenogens are not competitively excluding C. guiaris from the park despite advantage of a much greater population size. Overall, data from this and other studies lend support to the idea that parthenogenetic Cneanidophorus are ecologically and behaviorally 'inferior' to their gonochoristic congeners and should, therefore, be able to coexist with them only in habitats where selective forces are relaxed.

Nearlv all species of vertebrates are gonochoristic, that is, consist of two sexes that reproduce sexually. However, there are about 50 unisexual, all-female vertebrate species, of whici approximately 30 reproduce by obligate parthenogenesis (Vrijenhoek et al., 1989). Accounting for the existence of these parthenogens and, given that they do exist, determining why theyv are so rare has proved a daunting challenge to evolutionary biology (Bell, 1982; Lynch, 1984; Vrijenhoek, 1989). The lizard genus Caiemidaluhorus (Teiidae) has proven to be a rich source of model systems from which evolutionary hypotheses lhaveIbeen derived and tested. About one-third of the species are parthenogenetic, all originating from hybrids between two or more gonochoristic species (Wright, 1993). These parthenogens often coexist with gonochoristic congeners (Case, 1990; Paulissen et al., 1992), making it possible to gather ecological and behavioral data on parthenogenetic and gonochoristic species living in the same enviroiunental milieu. The first important attempt to distill ecological observations of parthenogenetic and gono-

cshoristic lizards into an explanatory theory was the "weed hypothesis" proposed by Wright and Lowe (1968). It stated that parthenogenetic Ctiemiidophorus could become established only in disturbed or ecotonal habitats that are unsuitable for gonochoristic species. Cuellar (1977) expanded on this idea by stating that parthenogens tend to occupy floodplain, beach, or lakeside habitats because these areas are subject to periodic natural disturbances that eliminate lizard populations and so can be occupied only by species with good colonizing abilities. Cuellar (1977) also emphasized that parthenogens must remain isolated from gonochoristic congeners to ensure they do not lose their genetic identity through hybridization with gonochoristic males, a phenomenon termedl destabilizing hybridization (Lynch, 1984). The general implication is that parthe.nogenetic Cnenuidophorus are susceptible to being displaced or eliminated by contact with gonochoristic congeners and, therefore, should persist only where gonochoristic species are absent. However, the numerous published examples of parthenogenetic Cnemidopihorus co-

CNEMIDOPHORUS ECOLOGY AND BEHAVIOR existing with gonochoristic congeners have cast doubt on the validity of these ideas (Paulissen et al., 1992). Other authors have suggested that the hybrid origin or parthenogenetic mode of reproduction may give parthenogens an ecological or evolutionary advantage. Because all parthenogenetic Cnernidophiorusoriginated from hybrids between gonochoristic species, they all exhibit unusually high levels of genetic heterozygosity (Dessauer and Cole, 1989). Several authors have postulated that such highly heterozygous genotypes confer fitness advantages on parthenogens (Parker and Selander, 1976; Cole, 1984). Furthermore, parthenogenetic reproduction theoretically results in an intrinsic rate of population increase at least twice that of sexually reproducing organisms because parthenogens do not produce males (Williams, 1.975; Maynard Smith, 1978). Based on egg-production data from laboratory populatiorns, Cole (1984) estimated that a parthenogenetic Cnermidophorus exsatnguis would produce a population twice the size of a gonochoristic species within three years. Realization of either or both of th-ese advantages should, theoretically, enable a parthenogenetic species to exclude a gonochoristic one from an area (Bell, 1982; Chaplin, 1993). A different viewpoint has recently been expressed by Price et al. (1993). They point out that, because parthenogenetic Cnemidophorus are derived from hybrids, their genotypes are a comnbination of two (or more) separate genomes that have been shaped by differing suites of selective pressures. A hybrid genotype would not be expected to function as well as either of the original parental genotypes, regardless of any advantage conferred by heterozygositv. Furthermore, parthenogenesis produces populations that lack the genetic diversity necessary to enable them to evolve through natural selection. Therefore, unlike their gonochoristic relatives, parthenogenetic species are not able to become better adapted to their environments over evolutionary time. This, combined with the tendency of parthenogens to accumulate deleterious mutations, a phenomenon known as Muller's ratchet (Vrijenhoek, 1989), should cause parthenogens to be ecologically and behaviorally "inferior" to their to their gonochoristic relatives (Price, 1992; Price et al., 1993). The best that parthenogens can do is to exist in environments where selective forces such as predation or parasitism are relaxed (Price et al., 1993). Deciding among these and various other hypotheses requires detailed study of the ecology and behavior of parthenogenetic and gonochoristic Cnemidaphorus lizards in their natural habitat. The Cnernidophorus laredoensiscomplex comprises two diploid, parthenogenetic species that

283

commonly coexist with their gonochoristic relative C. gularis in the Rio Grande valley of south Texas and Mexico (Walker, 1987a,b). The two parthenogens, provisionally designated LAR-A and LAR-B following Walker (1986), were derived from separate hybridizations between the gonochoristic species C. gularis and C. sexlineatus (Wright et al., 1983; Abuhteba, 1990). Both LARA and LAR-B as well as C. guiaris are easy to observe in the field, making it possible to collect data on lizard behavior under natural conditions. Herein, I present data on activity cycles, microhabitat use, and behavior for free-ranging LAR-A, LAR-B, and C. gularis and interpret the results in the light of the various hypotheses that have been advanced to explain the existence and raritv of parthenogenetic vertebrates. MATERIALS AND MFnTHODS

Study Anirnials.-The lizards of the C. laredoensis complex occupy disturbed habitats in sandy or sandy-loam soil on either side of the Rio Grande from Del Rio, Texas/Acuna, Mexico, southeast to Brownsville, Texas/Matamoros, Mexico (Walker, 1987a,b). LAR-A, the species originally named C. laredoensis by McKinney et al. (1973), occupies the middle half of this range; LAR-B (which has not been given a formal Linnaean namrie) occurs at the northern and southern ends of the range (Walker, 1987b; Abuhteba, 1990). Ihe entire geographic range of the twvo parthenogens is contained within the range of C. gularis, and they coexist at several sites in Starr and H:idalgo Counties in Texas. The three are readily distinguishable in the field by differences in color and pattern: C. gularis has a reddish-brown tail, numerotus spots between its lateral stripes, and a pink or orange throat that is visible whlen the lizard moves; male C. gularis have a dark blue chest, whereas female C. gularis have a plain white chest. Both parthenogens have a grey tail, few spots, and a white throat. TAR-A is distinguished from LAR-B by the vertebral stripe: narrow and vivid yellow-green in LAR-A, wider, thinly divided and tan in LARB (Walker 1987b). All three species are diurnal, ground-dwelling arthropodivores; their main prey is termites, which they find by moving over a wide area and probing the vegetation and ground with their snouts (Paulissen et al., 1988, 1992). Adult parthenogens and female C. gularis are the sarme size, 65-75 mm snout-vent length (SVL); adult male C. gularis are somewhat larger (8(1-85 mm SVL). When active, all three species maintain a high body temperature of 39-40'C (Paulissen, 1999a). All three species retire to burrows at night and during inclement weather (Walker et al., 1986; Paulissen, 1997); apparently none shares burrows with other individuals. Male C. gularis often court and sometimes cop-

284

MAIRK A. PAULISSEN

ulate with LAR-A and LAR-B femtales (Paulissen, 1995a; pers. obs.); these matings sometimes produce triploid hybrids (Walker et al., 1989, 1991). Study Sife.-This study was conducted along the RZio Grande River Hiking Trail in BentsenRio Grande Valley State Park near Mission, Texas (hereafter Bentsen Park). The 1.8-km. dirt trail runs through a tract of relict, subtropical mesquite-hackberry forest bordered by cultivated fields on one side and an overgrown marshy area on the other. The understory varies from dense clumps of grass in some sections, to sparser vegetation interspersed with patches of dead grass and /or leaf litter, to the open bare ground of the trail itself. Because the tree "canopy" is open along the most of the traii, patches of open sun, partial (filtered) sun, and full shade are available along the entire length of the trail. All three lizard species are found along the entire length of the trail but were most abundant in those areas where the soil is sandy or a sandy-loam mix. When the trail was first surveved for lizards in the 1980s, LAR-B outnumbered LAR-A by more than 4:1, and C. gularis was absent. By the time of this study, LAR-A outnumbered LAR-B by about 4:1, and C. gularis was present in small nunmbers (Walker et al., 1996). Further descriptions of the study site are given in Paulissen (1994, 1999b). Data Collection.-The daily activity periods of L.AR-A, LAR-B, and C. glak-is were estimated from time of capture records taken from a demographic study conducted the same time as the behavioral study (:Paulissen, 1999b). The number of lizards captured per hour for each hourly interval of the dav was plotted for each of the three species (see Fig. 1). Data on lizard microhabitat use and behavior were collected through focal animal observation of individual animals. When I encourntered a lizard along or near the trail, I paused for a few seconds to see whether the lizard stopped to watch me or otherwise reacted to mv presence. If it did, I searched for another lizard. If the lizard was undcisturbed by my presence, I followed it from a distance of at least 5 m and verbally recorded the microhabitat it was in, its exposure to sunlight (open sun, partial sun, shade, or overcast), and the lizard's behavior into a pocket tape record.er. A snmall pair of field binoculars was used to aid observation and to confirm species identification. Eachi time the lizard switched from one microhabitat, exposure, or behavior to another, I recorded it; later I played back the tape and measured how long the lizard spent occupying each microhabitat and exposure and performing each behavior using a stopwatch (accurate to the nearest second). These times were converted to proportion-s by dividing by

B6i 111| 1II.i LAR-A

40

LAR-B 26

a 15

a

S0~ ;

14

-

C. gularis

x12

10

r

2 Hour of Day

FIG. 1. Comparison of daily activity periods of the parthenogens LAR-A and LAR-B and their gonochoristic congener Ceomidopehorus gularis during May-june 1994-4996 in Bentsen Park. Data are number of lizards

captured per hour for each hourly interval taken from records collected during a demographic study of the lizards in the park. Lizards were censused 0830-1830 CDT; the number of lizards/hour for the 800 and 1801 hourly intervals are corrected to reflect the shorter sampling time during these intervals. the total length of the focal animal observation. I also recorded how many arthropods the lizard consumed during the focal animal observation, taking special note of the consumption of ter-

mites. In addition, as I followed a lizard, I placed markers along the path it had traversed; the first marker was placed at the point where the lizard was when I began the observation, the last at the point where the lizard was when the observation ended. Later, I uised a measuring wheel to measure the distance from. the first to the last marker along the path traced by the markers to estimate the distance the lizard moved during the observation. Each lizard was observed for a minimum of 10 min (mean 14.6 min). At the end of each observation, I measured

285

CNEMIDOPHORUS ECOLOGY AND BEHAVIOR

Number of focal animal observations (N) and total time lizards were obsenred (T) by species and

TAB,E 1. year.

N = TN T N T N T

LAR-A LAR-B C. gulatis Total

17 263 nin 7 110 min 6 88 min 30

36 N T = 528 min N= 13 185 min T N =8 1l29 min T N= 57

461 min

T -- 842 min

the temperature on the ground in open sun and in full shade with a soil thermom eter. All data were collected between 0830 and 11330 CDT during May and June 1994-1996, the Ibreeding season for Cnemidophorus lizards in scluthern Texas (Paulissen, 1999b). All three spec ies were observed in all three years producir ig 139 obser0.4T,

LAR-A

*

i-I]

+

Krf :~.' _ SB

CG

SS

BC

1

1~

eC Li MlCROHABrrAT

OT

BC

0.40.3 0.2 0.1.

,tIn: Li

SC

Be

OT

MICROHABITAT *

T

gularis C.

. __

W:g SB

BO

Li

SC

MICROHABITAT

BC

,,

l

oT

Comnparison of the mean prioportion of time LAR-A, LAR-B, and Cneoidophorus
BC - bare ground; LI - leaf litter; GC

i grass

clumps; BC = between grass clumps;. )The astedrOishks microhabitats (see text for explanatior indicate the statistically significant dif ference bet-ween LAR-A and C. gularis in proportion )f time spent in the side stLbble habitat (ANCVA: P == 0.027; Tukey's posthoc test: P < 0.05).

Total

1996

1995

1994

N =

27

N

362 min

T

N =17 -r 235 min N =8 T = 135 min 52 N T1 728 min

80 = 1149 min

N =37 Tr 530 min N =22 T = 352 min

N =139 1

2031

vations totaling over 2000 min of observation time (Table 1). To simplify the analysis of thie data, six microhabitats and six categories of behavior were defined. The six microhaibitats and the abbreviations denoting them in Figure 2 were (1) side stubble (SS)-the accumulation of dead grass and other vegetation litter along the sides of the dirt hiking trail; (2) bare ground (BG)-any patch of open ground including the trail; (3) litter (11)-patches of ground with at least some vegetation litter on them other than the side stubble along the trail; (4) grass clumps (GC)-

dense clumps of grass, either living or dead; (5) between grass clumps (BC)--patches of dead grass and other litter between dense clumps of grass, located only in places where expanses of grass were especially extensive; (6) other (OT)two microhabitats that were seldom used by lizards, viz. pricklv pear cactus and piles of sticks. The behavioral categories were (1) searching-the lizard moved slowly along the grounid in no particular direction and probed bits of ]eaf

litter, bases of vegetation, and holes in th-e ground primarily to locate arthropod prey (though potential mates may also be located during bouts of searching behavior); (2) moving-the lizard moved rapidly in a straight line across open ground; this behavior is distinguished from searching in that, while moving, the lizard does not investigate features of tlhe habitat; (3) stopped-this category included all behaviors in which the lizard was active but not moving, including basking in sunlight by flattening out on the ground and tipping the toes up in the stereotypical Cnemidphorus fashion, adopting the basking posture in full shade (presumably to cool off), defecating, and lying in a hole for a few moments before reemerging; (4) social interactions-this categorv included all behaviors in which the focal lizard interacted with another lizard, either a conspecific or a heterospecific, including aggressively chasing or submissively retreating from another lizard, courting or trailing another lizard closely, straddling another lizard., in which one lizard climbs

286

MARK A. PAULISSEN

onto the tail and back of another lizard and "rides" it for a short distance (Paulissen, 1997), and copulation; (5) eating-included timc spent subduing and consuming prey; and finally (6) escaping from a predator [e.g., a whipsnake (Masticophis taeniatus)]. The last two behavioral categories accounited for less than 5% of focal lizard time budgets and, thus, were not subjected to statistical analyses. To facilitate comparison with other studies of Cnemnidophorus lizard behavior, I calculated the following four statistics: (1) proportion of timne in motion-defined as the sum of the proportion of time the lizard spent searching, nmoving, and performing social behaviors in which the lizard was in motion; (2) velocity moving-the number of meters the lizard moved during the focal animal observation divided by the number of minutes the lizard spent in motion during the observation; (3) mean velocity-the number of meters the lizard moved during the observation divided by the number of minutes of the observation (Garland, 1993); and (4) prey capture rate-the number of arthropod prey the lizard captured during the observation divided by the number of minutes of the observation. In additio-n, I noted whetlher the focal lizard located any termites during the observation period. Statistical comparison of the mean proportion of time lizards spent occupying different microhabitats and exposures and performing different behaviors were made using ANOVA and Tukey's posthoc tests on arcsine-transfornmed data. Separate ANOVAs were run on each category of microhabitat, exposure, and behavior; for example, one ANOVA was done to compare proportion of time spent in side stubble, a second on proportion of time spent in bare ground, etc. Temperature, velocity moving, mean velocity, and prev capture rates were analyzed with ANOVAs on untransformed data. A chi-square test was used to evaluate differences in the number of observations in which the lizards did versus did not locate termites. Preliminary analyses showed male and female C. gularis did not differ in microhabitat use, exposure use, or prey captLre rate, but did in behavior. Therefore, statistical comparisons of microhabitat use, exposure, and prey capture rate were made among the three lizard species, (LAR-A, LAR-B, and C. gularis), whereas comparisons of behavioral data were made among the four species/sex groups (LAR-A females, LAR-B fernales, C. gularis females, aind C. gularis males). REsuI. Is Activity Cycles, Microhabitat Use, and Exposure.-The parthenogens LAR-A and LAR-B and C. gularis all had similar daily activity periods during May-June 1994-1996 (Fig. 1). For

LAR-A

.5 04] a 0 0.3 D.2

I

..1 C FullSun

S.~

P.01d Sun

Shade

EXPOSURE

f A-R-

..5.

I

!

1. 1

ril~~1

R-

8£ e0. 2+

Full S.n

ParSul Sun

Shade

EXPOSURE

I0.

l!

61

C. gularis

S 24~

01 1.+

] FullSun

r P.01.1 Sun

Shade

EXPOSURE

FiG. 3. Comparison of the mean proportion of time LAR-A, LAR-B, and Cnemidoplhorus gularis spent in each of the three exposures; vertical bars are standard errors. ANOVAs showed no statistically significant differences among the three species for any of the three exposures (all P - 0.05). all three, activity seldom began before 090(0 h, peaked at about noon, then declined somewhat before rallving to a second smaller peak at about 160() h; this late afternoon peak was weakler for LAR-B. Lizards were rarelv seen after 1800 h. Analysis of the proportion of time lizards spent in each of the six microhabitat categories showed very little difference among the three species (Fig. 2). Individuals of all three species spent over 7(0% of their time in one of three microhabitats: side stubble (SS), bare ground (BG), or litter (LI). The only statistically significant difference was that C. aularis spent a significantly greater percentage of its time in side stubble than did LAR-A. Similarly, the three species showed no significant differences in the amrount of time spent exposed to full suI, partial sun, or shade (Fig. 3). However, temperature data obtained at the end of each time budget observation suggest LAR-B was active at slightly cooler temperatures than was LAR-A; the difference between the two species was statistically significant for the temperature on the ground in shade measure (Table 2). Lizar,d Befurvior.-There were no significant

287

CNEMIDOPHORUS ECOLOGY AND BEHAVIOR

TABLD 2. Mean ground temperatures measured at the end of the focal animal observation periods of the two parthenogenetic lizard species lAR-A and lAAR-B and their goniochoristic congener Cnemindophlcrus gularis. Means superscripted with the same letter are not significantly different from one another (ANOVA, Tukey's posthoc test; alpha = (1.05). Standard error in parentheses.

Temperature on the ground in sun (C)

'Iemperature on the grontd in shade (IC)

LAR-A

LAR-B

C. gularis

(N =- 75)

(N = 35)

(N = 21)

P of ANOVA

44.5d

42.6"

43.0a

0.165

(0.616)

(0.901)

(1.192)

36.8, (0.359)

34.7' (0.534)

36.4AA (0.696)

differences in the proportion of time LAR-A and LAR-B spent performing the four major behaviors (TFable 3). However, there were significant differences between the two parthenogens and C. gularis females, C. gularis males, or both, and between C. gularns females and C. gularis males. For example, C. guiarisfemales and males spent a smaller proportion of their time engaged in searching than did LAR-A and TAR-B; Tukey's posthoc test showed the difference between LAR-A and both females and males of C. guiaris was statistically significant (Table 3). Female C. gularis spent a significantly greater proportion of their time stopped than did male C. gularis, although there was no significant difference betwveen either sex of C. gularis and either LAR-A or LAR-B. Btut the greatest difference was that male C. gulagis spent nearly one-third of their time engaged in social interactions, whereas female C. gularis, LAR-A, and LAR-B spent 1.5% or less of their time engaged in these behaviors (Table 3). The difference between male C. gularis and the females of all three species was significant and reflects the preoccupation of male C. gularis with courting and trailing females of all three species during the breeding season. Male C. gularis also spent a greater proportion

0.006

of their time in motion (Table 4); the difference between male C. gularis and female C. gularis and between male C. gularis and EAR-B females was statistically significant. Male C. guZaris tended to move faster than females of all three species, although this tendency narrowly failed to reach statistical significance (Table 4; velocity moving). However, male C. gularis did move significantly farther per unit time than did females of all three species (mean velocity; I'able 4). Again, this reflected the tendency of male C. gularis to spend large amounts of time in trailing and courting females. The only major behavioral difference between C. gularis and the two parthenogens was in time devoted to finding food. Cntnaidophorus gularis spent a significantly smaller proportion of its time engaged in searching behavior than did LAR-A; the difference between C. gularis and LAR-B narrowly failed to reach statistical significance (Table 5). Because searching behavior primarily involved locating arthropod prey, it was surprising that, despite their comparative lack of effort, C. gularis did not suffer a significantly lower rate of prey capture than the parthenogens (Table 5). There also was no differ-

TABLE 3. Comparison of the mean proportion of time spent by LAR-A females, LAR-B females, Cnemidophares gularis females, and C. gularis males performing the four major behaviors during focal animal observations (the other two behaviors, eating and escaping from predators, accounted for less than 5% of the focal aniimal's time and did not differ among the four lizard groups). Means superscripted with the same letter are nlot significantly different from one another (A NOVA and lukeys posthoc tests on arcsine transforrmed data).

Standard error in parentheses. LAR-A

Seardin-ig Moving Stoppedi Social interactions

females (N - 80) 0.514" (0.022) 0.070) ((1.008) 0.379,<) (0.023) 0.005a (0.009)

C. gularis

(N -- 37)

C. gularis females (N ---13)

0.468W" (0.032) 0.069" ((1.011) 0.410"" (0.033) 0.002a (0.013)

0.358b (0.054) 0.074" (0.019) 01.526, (0.056) 0.014, (021)

0.3291, (0.065) 0.121, (0.023) 0.223b (0.067) 0.31616 (0.025)

LAR-B females

males (N 7 9)

P of ANTOVA

0.007

0.216 0.01)7 <0.001

288

2MARK A. PAUL ISSEN

lAsBLE 4. C'omparison of the mean proportion of time in motion, velocity moving, and mean velocity of I AR-A females, LAR-B females, Cneomidophorus gistris females, and C. gularis males (see text for explanation). Means superscripted with the same letter are not significantly different frotm one another (ANOVA and Tukevs posthoc tests). Standard error in parentheses.

Proportion of timne in motion Velocity moving (m/min) Mean velocity (ni/min)

LAR-A females (N = 80)

LAR-B females (N - 37)

C. gdlaris females (N - 13)

C. giaris males (N - 9)

0.5387^' (0.022) 5.5d (l).293) 3.2

(1.537' ((0.033) 6.2' (0.431) 2.9'

0.445a ().055) 6.21 (0.727) 2.6,

0.727b ((0.067) 7.95 (0.874) 5.X7

(0.20)9)

(0.308)

(0.519)

((0.624)

ence among the three species in1 the success of focal animals in locating termites (Table 5). DiSCUSSION

Activity and Microhabitat.-The daily activity periods of LAR-A, LAR-B, and C. gullaris are extremely similar during the May-June breeding season in Bentsen Park. The only difference is a minor one: LAR-B is active less frequently during the late afternooni than the other two species. This trend has been documented before at another site approximately 30 km from Bentsen Park (Paulissen et al., 1988) and resulted in few focal animal observations being made of LARB after 1300 h. 'I'his probably accounts for why the ground temperatures measured at the end of LAR-B observations were cooler than those for LAR-A (Table 2). Detailed study of the thermal biology of these three species revealed no significant differences in lizard body temperature or environmental temperatures at points of capture in Bentsen Park (Paulissen, 1999a), nor in body temperatures selected during 24-h cvcles in a laboratory thermal gradient (Sievert and Paulissen, 1996). I'he general pattern is for LAR-A, L.AR-B, and C. gularis to be active at the same timne, a pattern that is typical for coexisting Cniemidiophiorislizards of the sarne size (Medica,

P of ANOVA 0.004 0.056 (1.00I

1967; Mitclhell, 1979; Schall, 1993; but see Creusere and Whitford, 1982). Simnilarly, there was little difference in the proportion of time the three species spent in different microhabitats and in different exposures. The extremely similar microhabitat use of LARA and LAR-B in Bentsen Park has been documented previously (Paulissen, 1994). The only significant difference found in the present study was that C. giilaris spent more time in side stubble than LAR-A. This is partly a reflection of the large amouint of time male C. gularis spent in this microhabitat trailing or courting females of all three species which were searching side stubble for arthropod prey. Microhabitat use similarity is often observed in areas where parthenrogenetic and gonochoristic Cnemidophorus coexist (Serena, 1984; Price et at., 1993; Schall, 1993). Behavior.-The greatest behavioral difference was between males of C. gularis and females of C. gularis, LAR-A, and LAR-B. Male C. gl:daris spend nearly one-third of their time engaged in social interactions during the May-June breeding season. Most of these interactions were trailing and courting females, not just of their own species but of LAR-A and LAR-B (although

TABI.i 5. Comparison of foraging success among lAR-A females, I.AR-B females, and Cncmitvhqdshforus guttii-s. The means of proportion of time spent searching and priey capture rate were compared using ANOVA and TIukey's posthoc tests; means superacripted with the same letter are not significantly different from one another, standard error in parentheses. The number of observations in which focal lizards did locate termites (number preceding slash) versus did not locate termites (number after the slash) was statistically evaluated using a dcisquare test. I Tukey's posthoc test for LAR-B vs. C. polaris: P 0.066.

Proportion of titne spent searching Prey capture rate (number/mnin) L.ocating termites/not locating termites

LAR-A temales (N = 80)

LAR-B females (N - 37)

males at7t females (N = 22)

P of Test

0.514a

0.468's'

0.346')

ANOVA: 0.0)03

C. gelaris

(0.022)

((0.032)

((1.041)

0 561' (0.161) 22/58

0.736 (0.236) 9/28

0.492 (0.30)6) 4/18

a

ANOVA: 0.774 chi-squareo 0.67

CNEMIDOPHORLUS ECOLOGY AND BEHIAVIOR male C. gula ris courted and trailed LAR-A far more often than LAR-B). The density of LAR-A in Bentsen Park is extremely high (Paulissen, 1.999b); thus, a male C. gularis often encountered several LAR-A during a 10-min focal animial observation period. When a male C. gvlaris encountered either an LAR-A or C. gularis female, he usually changed his search path to intercept her, tongue flicking rapidly as he approached. If the female did not run aw-ay, the male usually began trailin-g the female to the exclusion of all other behaviors, resulting in the high proportion of time spent in social interactions seen in this studv (Table 3). Male C. gularis also tended to move faster than females of all three species and spent a greater proportion of their time in motion than did females, in part because male C. gularis spend so much of their time courting and trailing females. Together, these two factors account for why male C. gularis move significantly farther per unit time than females of all three species (Table 4; mean velocity). The tendency for male Cnemnidophorns lizards to move faster than females has been noted for C. tigris (Anderson, 1993) and C. IWperythrus (Karasov and Anderson, 1984). Furthermore, male C. tigris show significantly greater mean velocities than females during the reproductive season but not during the postreproductive season, presumably because movement by males during the reproductive season was for locating mates as well as arthropod prey (Anderson and Karasov, 1]988). Unlike the male C. gularis in this study, male C. tigris and C. 1/Iperythrus did not spend a greater proportion of their time in motion than did females (Karasov and Anderson, 1984; Anderson and Karasov, 1988). Furthermore, the proportion of time male C. gularis spent engaged in social interactions in this study is an order of magnitude greater than for anv other Cnemidophorus lizards studied. Among gonochoristic species, the proportion of time spent interacting socially is 0.009-0.051 for C. leImo11iscatus (Vitt et al., 1997), < 0.050 for C. sex1ineatus (Paulissen, 1987), and 0.0 for C. deppii (Vitt et al., 1993). Among parthenogenetic species, results are 0.004 for C. unipvareais (Eifler and Eifer, 1998), and 0.0 for C. cryptus (Vitt et al., 1997). Sonie of the difference may be a result of methodological differences among studies; for example, I recorded a male trailing a female as a social interaction, whereas other workers nmay have recorded this behavior as simply being in motion. However, I found that the proportion of time female C, gularis, EAR-A, and lAR-B spent engaged in social interactions was the same as that for Cnetnidophorus lizards described in the studies cited above, suggesting the methodological differences are not very important. This

289

leaves the social environment at Bentsen Park as the best explanation for why male C. gularis spend so much of their time interacting with other lizards: the density of LAR-A is so high that there is simply a very large number of females with which to interact (Paulissen, 1999b). Consideration of the number of potential nmates a liza rd encounters during its daily routine clearly is necessary in interpreting differences among studies. Implications fior Coexistenc.-When Bentsen Park was first surveyed for Cneinidophorus lizards in the 1980s, E-AR-B was conmmon, LAR-A was rare, and C. gularis was absent. By the early 1990s, LAR-A had become the dominant parthenogen, and C. gal/ris had become established (Walker et al., 1996). The density of LAR-A rose dramatically from 1994 throLigh 1996, reaching a peak of 770 lizards/ha, the highest density ever recorded for a CDenmidophorus population (P aulissen, 1999b). During this time, LAR-B density remained level at about 70 lizards/ha, whereas the C. gularis population grew slightly but remained low (pers. obs.). The activity cycle and rnicrohabitat use data reported in this sttudy, plus the diet data reported by Paulissen et al. (1992), demonstrated that C. gularis overlaps extensively with the two parthenogens in all three important niche dimensions: time, space, and food. This suggests that conmpetition between C. guaaris and the parthenogens in Bentsen Park is potentially severe. Furtherrnore, given the overwhelming numerical superiority that the parthenogens (especially LAR-A) have over C. g/laris, plus the presumed reproductive advantage conferred by parthenogenesis (Williams, 1975; Maynard Smith, 1978), it is reasonable to predict that C. gularis will be excluded from Bentsen Park by LAR-A and LAR-B. This does not appear to be happening, however, raising the question of why not. There are two plausible explanations. First, it is possible that none of the resources for which these species comipete is in short enough supply to produce intense competition. The most obvious resource for which lizards might compete is food, especially termites. Yet the Cnemidophoruis lizards in Bentsen Park capture prey at a rate several tinmes greater than has been recorded for any other Cnenitdophorats species. Prey capture rates that have been measured for other species include 0.172 prey/min for C. sexlineafts (Paulissen, 1987), 0.070 for C. tigris (Anderson, 1993); 0.032 and 0.045 for two populations of C. ayperythrus (Karasov and Anderson, 1984); and 0.032 for C. aniparens (Eifler and Eifler, 1998). These values are onlv 5-30% of what I recorded in Bentsen Park (Table 5), suggesting food availability in Bentsen Park is so high that it is uinlikely to be the source of intense competition.

290)

MARK A. PAULISSEN

The only other resource for which lizards might compete is burrows. Although LAR-A and LAR-B generallv do not share burrows with each other, they do not defend their burrows from each other either (unpulbl. data). IPresumably neither parthenogen defends its burrow from C. gularis. The second explanation for why C. g/uiris is not being excluded from Bentsen Park is that the parthenogens are ecologically and behaviorally "inferior." Price et al. (1993) suggested that the hybrid origin, reproductive mode, and accumulation of deleterious mutations by parthenogenetic species such as LAR-A and IAR-Bi should cause them to perform poorly relative to their gonochoristic relatives. A few studies support this idea. Price (1992) found that parthenogenetic C. tesse/latus are more approachable., less cautious, and presumably more vulnerable to predators than are C. tigris, and Cullum (1997) found that parthenogenetic Cnenidop/lorus generally have lower endurance than their gonochoristic congeners. Comparison of LAR-A and LAIN-B to C. g/lturis suggest a similar trend of parthenogen inferiority. Although neither LAN-A nor LAR-B are more approachable than C. gularis (Paulissen, 1995b), LAAR-A females are slower and more easilv captured during flight than are C. gularis females (Paulissen, 1998). Furthermore, LAR-A do not live as long as C. gularis in Bentsen Park and so do not realize a greater lifetime reproductive success (Paulissen, 1999b). Also, this stLdy suggests C. gularis is a more efficient forager than is LAR-A: the parthenogen spends about half its time searching for arthropod prey, whereas C. gularis spends only about one-third of its time doing so, yet LAR-A does not capture more prey (Table 5). There are few data that directly compare LARB to C. gularis, and this study suggests LAR-B is at worst only a marginally less efficient forager than its gonochoristic relative. Howe.ver, the decline in relative abundance of LAR-B since the 1980s in Bentsen Park (Walker et al., 1996) and elsewhere in the Rio Grande Valley of 'Texas (unpubl. data) suggests it is also ecologicallv inferior to C. gularis. This inferiority may render these parthenogens incapable of competitively excluding their gonochoristic relative, even with the advantage of nuimerical dominance. If LAR-A and LAR-B are in fact ecologically inferior to C. gularis, then it is possible that C. gularis eventually may exclude the two parthenogens from Bentsen Park. Replacement of parthenogenetic species by gonochoristic ones has recentlv been documented in nocturnal house geckos. The gonochoristic species I-/emidactyl/us frel-atus is displacing the parthenogenetic iH.garnutii and Lepidodactylus lugubris from islands in the tropical Pacific (Bolger and Case, 1992; Case

et al., 1994). Behavioral studies show H. frenatus is behaviorally dominant to both parthenogens (Bolger and Case, 1992), and elegant experimental studies conduicted on Hawaii show that, when1 H. frenatus is present, the smaller L. lingubris avoids areas around insect-attracting lights and is less successful at capturing prey (Petren et al., 1993; Petren and Case, 1996). The result is that body condition and fecundity of L. legubris decline in the presence of H. frenatus, presumably leading to the replacement of L. 1ug7bris (Case et al., 1994). Cnenmidophorus gularis probably cannot displace LAR-A and LAR-B in this way. Male C. gue/ris are larger than parthenogenetic females, but neither thev nor female C. gularis act aggressivelv toward LAR-A or LAR-B. Unlike the situation with house geckos, the arthropod prey upon which Cnemidouphorus lizards feed are not concentrated in small areas; thus, even if the parthenogens avoided C. gularis, they would not necessarily suffer diminished foraging success. Regardless, food appears to be abtndant in Bentsen Park; thus, parthenogens can probably find enough food despite their foraging inefficiency. This makes it unlikely C. gularis will ever competitively exclude LAR-A and LAR-B from Bentsen Park. Given that other possible ways by which C. gularis could displace parthenogens, such as destabilizing hybridization, are not occurring to any extent in this system (Paulissen et al., 1992), it appears LAR-A and LAR-B will coexist with C. gularis indefinitely. It is notewvorthy that, although the parthenogenetic gecko L. hulgbris is being displaced bv the gonochoristic species H. frelnntus, it is stably coexisting with a congeneric gonochoristic species on at least one atoll (Hanley et al., 1994). Coexistence of parthenogenetic and gonochoristic Cnemidophorus lizards is common (Case, 1990; Paulissen et al., 1992). Most sites where parthenogens and gonochoristic species coexist are disturbed, transitional, or ecotonal habitats (WVright and Lowe, 1968; Paulissen et al., 1992). Such habitats are characterized by relaxed selective pressures in the form of greater food availability or reduced predator or parasite abundance. Price et al. (1993) suggested these conditions may be necessary to enable ecologically inferior parthenogens to persist over ecological time scales and to coexist with gonocho7ristic congeners. The existence of parthenogens, and coexistence of parthenogens with gonochoristic congeners, would not be expected in undisturbed, climax habitats where selective forces are stronger (Glesener and Tilman, 1978). Acknacwl/dQments.-I thank the staff of the Texas Parks and Wildlife Department for permission t) conduct this study (state park study permit 96-95) and the staff of Bentsen-Rio Grande

CNEMIDOPHORUS ECOLOGY AND BEHAVIOR Valley State Park for the courtesies extended me during the stLidy. Thanks to M. Mayon and J. Paulissen for help with the data analysis and B. Boothe for providing lodging during parts of this study. Drafts of this manuscript benefited from the constructive criticisms of my colleagues G. Hartman and H. Meyer. Financial support was provided by the Franklin L. and Laura Chavanne Miller Endowed Professorship in Science and by a Shearman Research Fellowship aiwarded by McNeese State University. LITFRAT'URE CLTED AO3uiYTEBA, R. M. 1990. Clonal diversity in the par-

thenogenetic whiptail lizard, Cewniduph.orus "laredoensis" complex (Sauria: Teiidae), as determined by skin transplantation and karyological techniques. Unpubl. Ph.D. diss., Univ. of Arkansas, Favetteville. ANDERSON, R. A. 1993. Analysis of foraging in a liz-

ard, Cnemidophiorus tharis: salient features and environmental effects. In J. W. Wright and L. J. Vitt (eds.), Biology of Whiptail Lizards (Genus Coem'?idphorus), pp. 83-116. Oklahoma Museum of Natural History, Norman. ANDERSON, R. A., AND W. H. KIARASOV. 1988. Energetics of the lizard Cnernidophiorus tigris and life history consequences of food-acquisition mode. Ecol, Monogr. 58:79-110. BLri., C. 1982. 'The Masterpiece of Nature: The Evolution and Genetics of Sexuality. Univ. California Press, Berkeley. BOLGER, D. T., AND T. J. CASE. 1992. Intra- and interspecific interference behaviour among sexual and asexual geckos. Anim. Behav. 44:21-30, CASE, T, J. 1990. Patterns of coexistence in sexual and asexual species of Cnenkidophuorus lizards. Oecologia 83:220- 227. CASE, r. J., D. I BOLGER, AND K. PETREN. 1994. Invasions and competitive displacement among house geckos in the tropical Pacific. Ecology 75: 464-477. CHIAPLIN, J. A. 1993. The local displacement of a sexually reproducing ostracod by a conspecific parthenogen. HI:eredity 71:259-268. Co:Fl., C. J. 1984. Unisexual lizards. Sci. Am. 250:94100. CREUSERE, E M., AND W G. WHITFORD. 1982. Temporal and spatial resource partitioning in a Chihuahuan desert lizard community. In G. W Salt (ed.), Ecology and Evolutionary Biology: A Round Table on Research, pp. 121-127. Univ. of Chicago Press, Chicago. CUELLARz, 0. 1977. Animal parthenogenesis. Science 197:837-843. CuiIUM, A. J. 1997. Comparison of physiological performance in sexual and asexual whiptail lizards

ETFLER, D. A., AND) M. A. EIFILER. 1998. Foraging behavior and spacing patterns of the lizard Cnemnidophorus uniparens. J. Hferpetol. 32:24-33. GARLAND, T. 1993. Locomotor performance and activity metabolism of Cneinidophorus tigris in relation to natLral behaviors. In J. W Wright and L. J. Vitt (eds.), Biology of Whiptail Lizards (Genus Cnernidophorus), pp. 163-210. Oklahoma Museum of Natural History, Norman. GLESENER, R. R., AND D. TILMAN. 1978. Sexuality and the components of environmental uncertainty: clues from geographic parthenogenesis in terrestTial animals. Am. Nat. 112:659-673. HANLEY, K. A., D. T. BOLGER, AND '. J. CASE. 1994. Comparative ecology of sexual and asexual gecko species (Lepidodactylus) in French Polynesia. Evol. Ecol. 8:438-454. KARAS-OV, WV. H., AND R. A. ANDERSON. 1984. Interhabitat differences in energy acquisition and expenditure in a lizard. Ecology 65:23.5-247. LYNCH, M. 1984. Destabilizing hybridization, general purpose genotypes and geographic parthenogenesis. Q. Rev. Biol. 59:257-290. MAYNARD SMiTH, J. 1978. The Evolution of Sex. Canmbridge Univ. Press, Cambridge. MCKINNEY, C. O., E R. KAY, AND R. A. ANDERSON.

1973. A new all-female species of the genus Cnen7idophiorus. Herpetologica 29:361-366. MEDICA, P A. 1967. Food habits, habitat preference, reproduction, and diurnal activity in four sympatric species of whiptail lizards (Crnemidophorus) in south central New Mexico. Bull. South. Calif. Acad. Sci. 66:251-276. MITCHIlI, J. C. 1979. Ecology of southeastern Arizona whiptail lizards (Cnemidoplorus: Teiidae): population densities, resource partitioning, and niche overlap. Can. J. Zool. 57:1487-1499. PARKER, E. D., AND R. K. SELANDER. 1976. The organization of genetic diversity in the parthenogenetic lizard Cnecidophorus tesselatus. Genetics 81:791805. PA,ULISSE7N, M. A. 1987. Optimal foraging and intra-

specific diet differences in the lizard Cnemidophorus sexlineatus. Oecologia 71:439-446. -. 1994. Microhabitat use and escape behaviors of syntopic clonal complexes of the parthenogenetic whiptail lizard Cnemidophorus laredoensis. Am. Midl. Nat. 132:10-18. . 1995a. Sexual and pseudosexual behaviors of the unisexual lizard Cnentidophorus laredoensis in

nature. Herpetol. Nat. Hist. 3:165-170. ______. 1995b. Comparative escape behavior of parthenogenetic and gonochoristic Cnemidophlorus in southern Texas. Copeia 1995:223-226. -.____

(genus Cnemidophorus): implications for the role of heterozygosity. Am. Nat. 150:24-47. Ji COE.E. 1989. Diversity between and within nominal forms of unisexual teiid lizards. In R. M. Dawley and J. P' Bogart (eds.), Evolution and Ecology of Unisexual Vertebrates, pp. 49-71. Museum Bull. 466, New York State Mu-

DESSAUER, H. C., ANI) C.

seum, Albany.

291.

_-

1997. Aggressive behaviors of lizards of the

parthenogenetic Cnemidophorus laredoensis complex (Sauria: Teiidae) in southern Texas. Tex. J. Sd. 49: 49-56. _ 1998. Laboratory study of escape tactics of parthenogenetic and gonochoristic Cneniidophorus from southern Texas. Copeia 1998:240-243. 1999a. T'lhermal biology of the parthenogenetic whiptail lizards of the Cneridophorus loredoensis complex (Sauria: Teiidae) in southern Texas. Tex. J. Sci. 51:37--48. 1999b. Iife history and drought tolerance of

292

MARK A. PAULIISSEN

the parthenogenetic whiptail lizard Octunidopbuorus laredoeeosis (Telidae). I- erpetol. Nat. Hist. 7:41-57. PAUISSE T,,M. A., J. M. WALKER. AND J. E. CORDES. 1988. Ecology of svntopic clones of the parthenogenetic whiptail lizard, Cnreinidophorus 'laredoensis.' J. Herpetol. 22:331-342. - 1992. Can parthenogenetic Cneomidopphors lar

edoensis (Teiidac) coexist with its bisexual congener? T.Herpetol. 26:153-158. PETREN, K., AND T J. CASE. 19)96. An experimental demonstration of exploitation competition in an ongoing invasion. Ecology 77:118-132. PETREN, K., D. '. BCOLGER, AN) T J. CASFe. 1993. Mechanisms in the competitive success of an invading sexual gecko over an asexual native. Science 259: 354--357. PRICE, A. H, 1992. Comparative behavior in lizzards of the genus Cneiofdophloros (Teiidae), with com-

brates, pp. 24-31. Museum Bull. 466, New York State Museum, Albany. VIJENIHOIK, R. C., R, M. DAWLEY, C. J. COic, AN) J. P. BoGAcTr. 1989. A list of the known unisexual vertebrates. In R. M. Dawley and J. 1 Bogart (eds.), Evolution and Ecology of Unisexual Vertebrates, pp. 19-23. Museum Bull. 466, New York State Museum, Albany. WAIKER, J. M. 1986. I'he taxonomy of parthenogenetic species of hybrid origin: cloned hybrid populations of C'nemnidgphorus (Sauria: 'Teiidae). Svst. Zool. 35:427-440. 1987a. Distribution and habitat of the parthenogenetic whiptail lizard, Cnemfdoephorus laredoe-nsis (Sauria: Teiidae). Am. Midl. Nat. 117:3109-332. -.____ 1987b. Distribution and habitat of a new major clone of partheniogenetic whiptail lizard (genus Cnemidopherus) in Texas and Mexico. 'Fex. J. Sci. 39: 313--334. WALKER, J. M., S. E. TRALUTI 1, J. M. BRITbON, AND J. E. CoRDEs. I 986. Burrows of the parthenogenetic

ments on the evolution of parthenogenesis in reptiles. Copeia 1992:323-331. PRIcE, A. fHi.,J. L. ILAPOINTE, AND J. W ATMAR. 1993. whiptail lizard Cnemridophuorus laredlensis (ieiidae) The ecology and evolutionary implications of comin Webb Co., Texas. Southwest. Nat. 31:408-410. petition and parthenogenesis in Coecoidophorus. In WAL KER, f. M., W C. GUEST, J. E. CORDES, AND M. A. J. W Wright and L. J. Vitt (eds.), Biology of WhipPAUI ISSEN. 1989. Morphological and chromosomtail ILizards (Genus Cnenifdophorus), pp. 371-410. al evidence of hybridization between all-female Oklahoma Museum of Natural FEistory, Norman. Ctiemindphorus laredoensis and gon)ochoristic Cnemidophorus gularis. Copeia 1989:1059-1064. SC-TALL, J. J. 1993. Commnity ecology of CnetnidoWALKER, J. M., R. M. AB3uTE1THA, AND J. E. CORDES. plJtous lizards in southwestern Texas: a test of the 1991. Morphological and experimental verification weed hypothesis. In J. W. Wright and L. J. Vitt of hvbridization between all-female Cnemnidophorus (eds.), Biology of Whiptail Lizards (Genus Co;niilareentsis B and Cmcmiidophzorus gularis (Squamata: dophorus), pp. 31C9-343. Oklahoma Museum of NatTeiidae). Herpetologica 47:152-164. ural H4istory, Norman. WALKER, J. M., M. A. PAULISSEN, AND J. E. CORDES. SFRENA, NJ. 1984. Distribution and habitats af par1996. Apparent dcanges in the composition of a thenogenctic and sexual Coenfidophorus ler;niscatus community of cnemidophorine lizards (Sauria: Tei(Sauria: Teiidae) in Surinam. Copeia 1984:713-719. idae) in a subtropical Texas forest. Southwest. Nat. STUVERT, L. M., AND M. A, PAn1.1553 N. 1996. Temper41 :64-67. ature selection and thermoregulatory precision of WILLIAMS, G. C. 1975. Sex and Evolution. Princeton bisexual and parthenogenetic Coemndopeporus lizUniv, IPress, Princeton, NJ. ards frorm southern Texas, USA. J. 'fl-erim. Biol. 21: WRIGHi, J WV. 1993. Evolution of whiptail lizards (ge15-20). nus CoetniAophorus). In J. W. Wright and T.. J. Vitt VITI, L. J., P. A. ZANI, J. P CATDjWELL, ANO R. U. DuRi(eds.), Biology of Whiptail lizards (Genus CnemiscIIE. 1993. Ecology of the whiptail lizard CiLenidophorus), pp. 27-82 Oklahoma Museum of Natural idophorus deppli, on a tropical beach. Can. J. Zool. History, Norman. 71:2391-2400. WRIGHI, J W, ANT) C. 1I. I.OWE. 1968. Weeds, polyploids, parthenogenesis, and the geographical and VrrT, L. J., P A. ZANI, J, P CATDWELL, M. C. DE ARecological distribution of all-female species of AUtO, AND W. M. MAGNU5CSON. 1997. Ecology of Cnicnidophorus. Copeia 1968:128-138. whiptail lizards (Cneouidophorus) in the Amazon reWRIGTIi, J. W, C. ST'OLSKY, AND W. M. BROWN. 1983. gion of Brazil. Copeia 1997:745-757. The origin of the parthenogenetic lizard CneeidoVRIJENHOLK, R. C. 1989. Genetic and ecological conphorus /aredoensis inferred from mitochondrial straints on the origins and establishment of uniDNA analysis. I1-lerpetologica 39:41.0-416. sexual vertebrates. InIR. M. Dawley and 1. P. Bogart (eds.), Evohition and Ecology of Unisexual VerteAccepted: 28 AuguLst 2000.

COPYRIGHT INFORMATION

TITLE: Ecology and behavior of lizards of the parthenogenetic Cnemidophorus laredoensis complex and their gonochoristic relative Cnemidophorus gularis: implications for coexistence SOURCE: Journal of Herpetology 35 no2 Je 2001 WN: 0115205199014 The magazine publisher is the copyright holder of this article and it is reproduced with permission. Further reproduction of this article in violation of the copyright is prohibited.

Copyright 1982-2001 The H.W. Wilson Company.

All rights reserved.