Sexual and age differences in ecological variables of the lizard Microlophus atacamensis ( Tropiduridae ) from northern Chile Diferencias sexuales y etárias en variables ecológicas del lagarto Micr lophus atacamensis ( Tropiduridae ) del norte de Chile

Microlophus is a lizard genus of South-America which has many species with sexual size dimorphism. An ecological study was performed on M. atacamensis , a species inhabiting the intertidal zone of the coast of northern Chile. The following questions were addressed: (1) does M. atacamensis exhibit sexual dimorphism? (2) do individuals of different age and sex segregate spatially? (3) do individuals of different age and sex use different type of microhabitat? and (4) do individuals of different sexes exhibit differences in thermoregulatory characteristics? The study was conducted in different localities from northern Chile, which include two types of rocky systems used by this species, a high and a low type. At the moment of lizard capture, type of microhabitat, height of perch, body, air and substrate temperatures, were recorded. Individuals were then measured and weighted, and separated by age class and sex. Results indicate that this species exhibits sexual dimorphism, males being larger. There were no intersexual differences in microhabitat used and height of perch in the high type system, although adults and subadults were spatially segregated from juveniles. In the low type system adult females shared microhabitat with juveniles, a probable consequence of behaviors related to the burying of eggs. There were no differences between sexes in thermal characteristics, and body temperature of lizards showed independence from external thermal conditions.


INTRODUCTION
In most animal groups, sexual differences in morphological characters (sexual dimorphism), particularly in body size, is a common phenomenon.The direction of this difference, i.e., Sexual and age differences in ecological variables of the lizard Microlophus atacamensis (Tropiduridae) from northern Chile Diferencias sexuales y etárias en variables ecológicas del lagarto Microlophus atacamensis (Tropiduridae) del norte de Chile
whether males or females are bigger, differs between animals groups; in vertebrates males typically constitute the larger sex (Schoener et al. 1982, Shine 1986, Fairbairn 1990, 1997, Andersson 1994).Several proximate mechanisms have been proposed to explain sexual dimor-VIDAL ET AL.
phism, such as differential mortality of sexes (Stamps 1993) and different growth rates of sexes (Watkins 1996).However, natural or sexual selection remain the ultimate mechanisms explaining sexual dimorphism (Shine 1986, Andersson 1994, Watkins 1998).
Sites were continuously walked along parallel 30 m transect lines (perpendicular to the coastal line), during the whole daily active period of lizards, in both months (09:45 to 18:30 h).Transects were walked in opposite directions to minimize the probability of repeatedly observing the same individual.Two types of rocky systems can be distinguished in the habitat of M. atacamensis: (a) high type system, characterized by the presence of rocks with heights from 0.20 to 3 m, in contact with the sea, behind which there is sand with boulders with heights from 0.10 to 0.20 m; (b) low type system, consisting of a stripe of boulders with heights from 0.20 to 0.35 m, in contact with the sea, behind which there is sand (Fig. 1).
For each lizard observed, the type of substrate used (rocks, boulders or sand) and its height (m) were recorded.A total of 86 individuals were collected (high type system n = 54, low type system n = 32).Immediately after capturing a lizard, the following temperatures were recorded: body (cloacal, T b ), air (10 cm above the substrate, T a ), and substrate (in contact with the surface, T s ), with a UNI-T M-890C thermometer (± 0.1 ºC).Lizards were weighed (g) and measured (mm); some individuals were sacrificed, fixed in 70 % alcohol, labeled, and donated to the herpetologi- cal collection of the Museo de Zoología of Universidad de Concepción (MZUC, see Appendix 1).The morphological measurements obtained from each animal were: snout-vent length (SVL), head length (HL), head width (HW), and right hindlimb length (RHL), with a Mitutoyo caliper (± 0.01 mm).Thereafter, and following Flores et al. (1977) and Ortiz (1980b), lizards were classified as adults, subadults or juveniles, according to body size (SVL).
The sexual dimorphism index (SDI) was calculated for adults and subadults (Lovich et al. 1990, Lovich & Gibbons 1992, Watkins 1996), where: SDI = (mean body size of the larger sex / mean body size of the smaller sex) -1 Results were defined as positive when males were larger than females.
Since the final sample size per locality was small, data were pooled.Analyses of the differences between sexes and ages in SVL, weight, microhabitat use, and perch height, were perf o r m e d b y t w o -w a y a n a l y s i s o f v a r i a n c e (ANOVA), followed by Tukey tests (Zar 1996).Comparisons of HL, HW and RHL between sexes were performed using analysis of covariance (ANCOVA), with SVL as covariate (Sokal & Rohlf 1981); data were previously normalized using log 10 .A one-way ANOVA was used to determine differences between sexes in cloacal temperature in relation to air and substrate temperatures.The thermoregulatory ability was estimated from the variances of T b , and from the slopes of the linear regressions (Sokal & Rohlf 1981, Dytham 1999) between T b and T a and between T b and T s (Huey 1982, Pérez-Mellado & de la Riva 1993).A slope equal to one indicates that animals are completely thermodependent (Báez & Cortés 1990).

RESULTS
The morphometric measurements obtained for the different age classes (excluding juveniles) and sexes are shown in Table 1.A two-way ANOVA, considering age and sex as factors on SVL, indicated that males were significantly longer than females (F 1,56 = 30.81,P = 0.001).Significant differences in SVL were observed between adults and subadults (F 1,56 = 59.23,P = 0.001).The interaction of both factors was not significant (F 1,56 = 2.49, P = 0.12), although the dimorphism index in adults (SDI = 0.18) was higher than in subadults (SDI = 0.12).There were significant differences in weight between sexes (F 1,56 = 14.53,P = 0.001), males being heavier than females.There were differences in weight between age classes (F 1,56 = 59.19,P = 0.001), although the interaction between factors was not significant (F 1,56 = 3.18, P = 0.08).
Males had longer hindlimbs and heads, and wider heads than females, either at the adult or subadults stage (Table 1 and 2).HL and WH, but not RHL, of males and females correlated positively with SVL (Table 2, Fig. 2).However, the interaction with the covariate does not show a difference of slopes for HL and HW, but it does for RHL (Table 3).
Table 4 shows the use of microhabitat by different age classes and sexes, in both rocky systems.A two-way ANOVA considering age and sex as factors, indicated that only age affected the substrate used, in the high system (F 2,48 = 44.60,P = 0.001); juveniles were observed only in the sand, while adults and subadults of both sexes basically did not use this substrate (P < 0.001).Table 5 shows the height of the perch used by the five agesex classes defined.A two-way ANOVA indicated that only age in the high system, had an effect on the height of the perch; juveniles used significantly lower perches than the other categories (P < 0.01).In the low system, the significance of the effect of age was marginal (F 2,25 = 3.00, P = 0.068), because perches of similar height were used by juveniles and females (P = 0.095).
The mean body temperature of M. atacamensis was 24 ºC, and sexes had similar temperatures (F 1,60 = 0.18, P = 0.65).Body temperature showed similar ranges and standard deviations in both sexes (Table 6).There were no significant differences between males and females in the relationships T b versus T a (F 1,60 = 2.84, P = 0.09) and T b versus T s (F 1,60 =1.37, P = 0.24).The slopes of the regressions of T b versus T a and T b versus T s , were not different from zero, in both sexes (Fig. 3).

DISCUSSION
Results indicate that M. atacamensis exhibit sexual size dimorphism.The snout-vent length, the length of hindlimb and head, and the width of the head, were larger in males.The existence of species with sexual dimorphism has been documented in many lizard families, such as Tropiduridae (Schoener et al. 1982, Pérez-Mellado & de la Riva 1993, Duarte 1996) and Iguanidae (Schoener 1967, Cooper & Vitt 1989, Anderson & Vitt 1990).The fact that Microlophus and species of other genus of the family Tropiduridae show sexual Three covariance analyses for head length (HL), head width (HW), and right hindlimb length (RHL), using snout-vent length (SVL) as covariate and sex as factor.Data came from adults and subadults of Microlophus atacamensis Tres análisis de covarianza para la longitud de la cabeza (HL), ancho de la cabeza (HW) y la longitud de la extremidad posterior derecha (RHL), utilizando la longitud hocico-cloaca (SVL) como covariable y el sexo como el factor.Los datos son de adultos y subadultos de Microlophus atacamensis  size dimorphism (Schoener et al. 1982, Snell et al. 1988, Pérez-Mellado & de la Riva 1993, Duarte 1996, Watkins 1996, 1998), strongly suggests that this character may be ancestral in the family.
In Microlophus the degree of sexual dimorphism shows intrageneric variation, and M. atacamensis is one of the species with the smallest dimorphism indexes, even if only the values obtained for adults are considered (Watkins 1996).This suggests that the selective pressures that maintain  sexual dimorphism in M. atacamensis are weaker than for other Microlophus species.Additionally, the differences in SDI observed between adults and subadults of M. atacamensis, suggests that the proximal mechanism of this sexual size dimorphism is a differential growth rate between sexes, as was proposed for M. occipitalis (Watkins 1996).
The ultimate mechanisms proposed to explain sexual dimorphism are natural and sexual selection (Darwin 1871), and the degree of elaboration of a trait would be determined by a balance between the costs and benefits associated with these two mechanisms (Shine 1989, Pérez-Mellado & de la Riva 1993).Natural selection involves an ecological segregation of sexes in at least one niche axes, in which resources are limited, thus reducing intraspecific competition.Dietary niche partitioning between sexes has been discussed as a possible selective force underlying evolution of sexual size dimorphism in lizards, in cases where food resources are limited (Shine 1986, Camilleri & Shine 1990, Fairbain 1990, Andersson 1994).
Because head size is related to the size of prey consumed (Schoener 1967, Schoener et al. 1982), morphometric results of M. atacamensis may support the hypothesis of trophic segregation between sexes.To date, there is no information about the nature and availability of the prey consumed by this species, or whether sexes consume different prey.
On the other hand, sexual selection proposes that females select larger males given the strong relationship between body size and competitive efficiency, which leads to larger males having larger or higher quality territories (Manzur & Fuentes 1979, Heisig 1993).Moreover, males with larger head show higher success during intrasexual agonistic interactions (Cooper & Vitt 1989, Anderson & Vitt 1990, Watkins 1998).Considering that M. atacamensis has a polygynic social system (Heisig 1993), sexual selection seems to be the most likely hypothesis to explain sexual dimorphism in this species (Martins 1994).
The sexual size dimorphism observed in adults and subadults of M. atacamensis, was not corre- lated with a differential use of substrate.This result is interesting because different studies indicate a strong relationship between morphological characteristics and type of substrate used (i.e., Vitt et al. 1997).Nevertheless, there is a clear spatial segregation between adults and subadults in relation to juveniles, mainly in the high system, as was previously suggested (Donoso- Barros 1966, Flores et al. 1977, Ortiz 1980a, Heisig 1993).Our present results clearly indicate that adults and subadults of both sexes use higher perches which are closer to the sea than do juveniles, and these latter were observed frequently in the sand, far from the sea.However, during the oviposition period (November-December) females were found using sand, as juveniles did, particularly in the low system.This switch in substrate use by females might reflect the fact that pregnant females bury their eggs in the sand (Donoso-Barros 1966, Ortiz 1980a, Olivares et al. 1987).Adults and subadults of M. atacamensis actively forage in the intertidal zone, and adults and subadults use perches more exposed to insolation than juveniles (Yáñez 1951).Therefore, these two age groups use areas with better thermal and trophic resources, than do juveniles.This would explain the highly territorial behavior reported at least for adult males (Heisig 1993).If food and thermal resources are limited, probably adults and subadults exhibit aggressive behavior towards juveniles; hence, juveniles with lower competitive abilities use suboptimal substrates, thus showing a spatial segregation in relation to the other age classes (Ruby & Baird 1993).However, Donoso- Barros (1948Barros ( , 1966) ) and Ortiz (1980a) proposed that this spatial segregation could be a consequence of cannibalism by adults and subadults on juveniles.
Males and females of M. atacamensis had similar body temperature during activity in the field, around 24 ºC.Interestingly, Heisig (1993) indicated that individuals of M. atacamensis are active only when the substrate temperature exceeds 22 ºC.Considering that males and females did not show differences in the use of the substrate, probably they had the same availability of thermal resources, and therefore similar body temperatures are to be expected.On the other hand, body temperatures showed independence from the thermal environmental condition, since they did not correlate either with air or with substrate temperatures; therefore, this species can be considered a thermoregulator (Huey 1982, Báez & Cortés 1990).This thermal independence, also reported for M. quadrivittatus, another species inhabiting the intertidal zones of the north of Chile (Báez & Cortés 1990), should allow those species foraging actively in the intertidal zone or close to it, not to loose heat fast by conduction or convection.This thermal independence acquires even more relevance considering that a wet lizard will loose heat even faster than a dry one by convection (O'Connor 1999).On the other hand, in other M i c r o l o p h u s s p e c i e s i t w a s s h o w n t h a t antipredator responses depend on body temperature (Watkins-Colwell 1997); thus, in the case of M. atacamensis, predation would be another selective factor favoring thermal independence.
Microlophus atacamensis had body temperatures significantly lower than those reported for Chilean Liolaemus lizards, which are around 35 ºC and range from 31 to 37 ºC (Fuentes & Jaksic 1976, Marquet et al. 1989, Carothers et al. 1998, Labra 1998, Labra et al. 2001).Since there are no Liolaemus species using microhabitats similar to those of M. atacamensis (Donoso-Barros 1966), it is not possible to determine if this difference in body temperatures is a consequence of M. atacamensis being exposed to constraints for thermoregulation (Huey & Slatkin 1976), or that there is a phylogenetic component determining this difference between genera.However, considering that the selected body temperatures (sensu Gans & Pough 1982) of M. heterolepis are 32.4 ± 1.23 ºC (mean ± SE, n = 6, unpublished results), it is possible to suggest that the low body temperature of M. atacamensis is a consequence of constraints to thermoregulation.In summary, M. atacamensis exhibts sexual size dimorphism, although this is not reflected in the type of substrate used.There is spatial segregation in the type of substrate used in adults and subadults, in relation to juveniles.Finally, males and females showed no differences in thermoregulation.

Fig. 3 :
Fig. 3: Relationship between (A) body and air temperature (T b -T a ) and (B) body and substrate temperatures (T b -T S ) in ( … ) males and ( _ ) females of Microlophus atacamensis.The slopes of regression (b) and r values are indicated.Relación entre (A) temperatura corporal y ambiental (T b -T a ) y (B) temperatura corporal y del sustrato (T b -T S ) en ( … ) machos y ( _ ) hembras de Microlophus atacamensis.Se indica el valor de la pendiente de la ecuación de regresión (b) y su respectivo valor de r.

TABLE 3
Test of parallelism of interaction of the factor (sex) with the covariate (SVL) for head length (HL), head width (HW), and right hindlimb length (RHL).Data came from adults and subadults of Microlophus atacamensis

TABLE 4
Microhabitats used (%) by Microlophus atacamensis in two rocky systems, high and low type; n = sample size Microhabitats utilizados (%) por Microlophus atacamensis en dos sistemas rocosos, alto y bajo; n = tamaño de la muestra Perch height (m) used by the different age classes and sexes of Microlophus atacamensis in two rocky systems, high and low type.Data are given as mean ± standard deviation (mean ± SD) and range (R); n = sample size