Natural Selection on Behavioral Phenotypes of the Lizard Uta ...

Natural Selection on Behavioral Phenotypes of the Lizard Uta Stansburiana @

Stanley F. Fox Ecology, Vol. 59, No. 4. (Summer, 1978), pp. 834-847. Stable URL:

http://links.jstor.org/sici?sici=0O12-9658%28 197822%2959%3A4%3C834%3ANSOBP0%3E2.O.C0%3B2-X Ecology is currently published by The Ecological Society of America.

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/esa.html. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

JSTOR is an independent not-for-profit organization dedicated to creating and preserving a digital archive of scholarly journals. For more information regarding JSTOR, please contact [email protected].

http://www.jstor.org/ Wed Jul 12 17:33:20 2006

Lt i

9 1 4 ) . 1978. pp X3&847 1978 b) thc t i u l u g ~ c a lS u ~ ~ r ot t y -imerlca

niiii'\.

NATURAL SELECTION ON BEHAVIORAL PHENOTYPES O F THE LIZARD UTA STANSBURIANA1 S I , A ~ I . FF.. \FOX Dc~pcirtn?c,nr($Biology, Bosto?~U n i i ~ c ~ r sBoston, it~. .Mu.r.ruc~hiisrrt.,0221.5 USA' Ab,truc,t. Behavioral differences of resource utilization between juvenile survivors and nonsur.stqjr~rgrriwere measured. Marked indivivors of the desert side-blotched lizard Uru ~tti,~sburiuntr viduals were followed through time in enclosed field plots s o disappearance indicated death. In the 1st yr of the study. one characterized by light selection pressure, survival was independent of most of the behavioral variables. In the 2nd yr of the study. one characterized by heavy selection pressure, oversummer survivors were active under conditions of clearer skies. warmer temperatures. later morning hours. earlier evening hours, and were associated with vegetation furnishing more shade and shelter than nonsurvivors. Two-way analysis of variance demonstrated behavioral differences between survivors and nonsurvivors, independent of age. Survivors were seen less frequently during the summer of heavy selection pressure, and were more specialized in their activity (smaller phenotypic breadth). The home ranges of survivors contained significantly greater densities of yucca and shelter holes. and lesser densities of Croron. The diversity of plants and other elements on the home ranges of survivors was significantly greater than on those of nonsurvivors. Older. larger juvenile lizards moved to the unmodified sector of an experimental plot in which vegetation was cut away. Those older juveniles which showed some activity on the modified sector experienced heavy mortality. No genetic bask for the behjvioral differences between survivors and nonsurvivors was obtained. The behavioral variation in the population, if directly heritable. may be maintained by opposing. fluctuating selection pressures: or. if a consequence of microhabitat differences. by the limiting amount of superior habitat used somewhat exclusively by surviving lizards. The quality of the home range an individual inhabits probably influences his behavior and ultimately his survival. Aggressive behavior may influence the quality of home range an individual obtains.

Kr.i ~ t , o r c l ~Brhui,ior: : h r h u ~ , i o r uc'cology: / dc,sc,rr.\; d$ferrr~tiul t t ~ o r r u l i t ~1izcrrd.s: ,: ~ ~ c i t ~ i .trlrc,rcil rion: p o p u l u t i o ~vuritrtior~: ~ 1r.rus; Uta stansburiana.

Natural selection is the sustained differential production of offspring by various genotypes. Such differential production of offspring may be generated through differential fecundity and/or differential viability of these various genotypes. This study is an attempt, at the behavioral level, to quantify natural selection acting through differential survival (viability) of juveniles of the desert side-blotched lizard. Utrr s~tut~.shtrriiinii .stc.jnc,geri. Differential survival of prereproductive juveniles ultimately leads to differential production of offspring because only those that live to maturity may contribute offspring to the next generation. Natural selection has been studied among a variety of organisms. Reptiles. probably because of their easily countable scales, have often been used to demonstrate the effects of natural selection on morphological phenotypes. Most of these studies have compared adult and juvenile distributions of specific meristic characters and inferred differential mortality from observed differences between the 2 samples (Dunn 1915. Inger 1943. Mertens 1947, Hecht 1952, Fox 1975). Of those studying reptiles only Fox (1975) directly observed natural selection on a cohort of marked individuals followed through time.

' Manuscript received 7 February 1977: accepted I2 January 1978. ' Present address: School of Biological Sciences. Oklahoma State University Stillwater. Oklahoma 74074 USA.

Like morphology. behavior is subject to natural selection and marked by various levels of heritability (Ehrman and Parsons 1976). Evidence for natural selection on nonreproductive behavior is rare. although some studies suggest that the behavior of prey may be molded by predation pressure (e.g.. Goss-Custard 1970. Pianka 1970, Levinton 1971. Charnov et al. 1976. Stein and Magnuson 1976). Watson (1967) has shown that territorial behavior influences survivorship of Red Grouse. Other studies of birds have demonstrated the adaptive significance of territorial behavior (reviewed by Brown and Orians 1970). Since behavior varies among individuals in a population, i t seemed reasonable to expect individual differences in the manner of gathering resources. I hoped to relate these differences to survival of known individuals followed through their juvenile summer. I chose to study Utu st~tlshrlriiit~ii because it is sedentary, diurnal and observable in the field. abundant, nearly annual, subject to heavy predation. and well studied by other workers (Soule 1 9 6 7 ~ .1967h. Hoddenbach and Turner 1968. Tinkle 1965. 1 9 6 7 ~ .1967h. 1970, Turner et al. 1970. Parker and Pianka 1975, Medica and Turner 1976. and others). Utcl (and indeed lizards in general) is an ideal subject for evolutionary1 experimental field studies. It can be easily marked for individual field recognition. large numbers of indiciduals (particularly juveniles) can be found across small areas. recapture rates are extremely high. i t is ame-

Summer I978

NATURAL S E L E C T I O N O N U7A

nable to field experimentation. and. for t h ~ sstudy in particular. experiences heavy mortality. Generation times are short so that evolutionary change over successive generations can be studied over several years. This paper describes the behavioral aspect of a larger study which relates differential survival of juvenile lizards on the biochemical. morphological. and behavioral levels (Fox 1973. 1975). M A IF.RI,AI.S A N D M ~ T H O D S Sprc.ios u t ~ dstirdy .sire

835

It~dii,idirtllidetlrrfictlrrotl

When first captured. each lizard was individually marked with a combination of clipped toes and paint spots which allowed continuous recognition throughout the season. The toe clipping was used as a permanent method of identification. In 1971. 2 toes. and in 1972. 3 toes were removed from each lizard. T o minimize possible handicap. only the extreme tip. just basal to the nail was removed. and no more than 1 toe per foot was removed. No toe regeneration was observed. Each lizard was also marked with 4 dots of quickdrying enamel applied with a disposable syringe to the dorsal surface. Five colors were used but no more than 2 were used on any lizard. The paint had to be reapplied following each molt. but otherwise allowed individual field recognition.

A population of Urti sttrtl.c.h~rritr~ltrwas studied in its natural habitat in western Texas. Most of the data were obtained on first-summer juvenile lizards, although adults were also observed. The observations were made on an area of 95%. When a color-marked maining plot was unenclosed. individual was seen, its position and other data were In this area of its range Uru reaches adult densities recorded without capture. of -33iha and juveniles are often as dense as 2251 Each time a lizard was observed, the following data ha. Average clutch size is 3 eggs and average number were recorded: grid position. date. time, cloud cover. of broods per qear is 3-4. Adults are basically semel- temperature. wind speed. structural assocition, molt parous, experiencing 90% turnover from year to year. condition, and (if captured) snout-vent length, wholeAdults are territorial, males against males and females body weight. evidence of tail loss. and length of tail against females. Usually territories of I male and 1 stump. Some variables were measured in a slightly female overlap to the extent that the social system is different manner each season (Table I). In 1971. temeffectivelq monogamous. Lizards hatched in one sum- perature was ranked from hot to cold ( 1 4 ) . whereas mer reproduce in the next. Uttr is a sit-and-wait pred- in 1972, temperature was measured with a Celsius ator on a diverse collection of arthropods: the juve- thermometer in a standard white shade box 10 centiniles re14 heavily on ants. (I>etails of life history from metres above open sand. In 1971. wind speed was es.I inkle 1967b.). timated in only 2 categories. still and windy (1-2). and

S T A N L E Y F. FOX

836

' T A B L F I. Ob,er\ational variable, and categories. C'ategories of structural association are plant species and other habitat ,trLlctures rank ordered by the e,timnted amount of shade and shelter each provided -

Variable name - -

I97 1 cntegorie, -

~

1972 categories -

'Time Continuous As S t r l ~ c t ~ ~assoc~ation rnl I Mesquite As 7- Broomweed As 3 Yucca As 4 I'ack rut neat As 5 Bluebush A\ 6 Grass A, 7 Beneath b o a ~ dA\ 8 A b o \ e board A, A, 9 Open sand

I971 I971 I971 I971 I971 1971 I971 I971 I971 I971

As As As As

1971 I971 1971 1971

I 2 3 4

Sky condition\

Clear Partly cloudy Cloudy Overcast

--

Ecology. Vol. 59. No. 4

ground). The original experimental design of various predation pressures was not further ~ltilizedafter preliminary analyses showed no promise. Data from all plots were combined for the analqses reported here. Avian and mammalian predation. thought to be relatively in5ignificant (Tinkle 1967h). was ~ ~ n c o n t r o l l e d . All other animals on the plots were left undisturbed. The on14 other abundant lizard on the study area was h o ~ ' r ~ .T\ h e T e x a s the Whiptail ( C ~ ~ c ~ t ~ ~ i a ' o p tigri.5). ~tr and the skink horned lizard ( P h r ~ t ~ o s o t , c,or~lrrtrrt~l) (Err~,~rc,c~.r ob.rolctr~.c)were present in lower numbers.

Repeated behavioral observations on identifiable individuals permitted a precise description of their resource utilization in terms of several ecological pal-amContinuou, Celsius I Hot 'Ternpel.at~~re eters. As an individual was 'ecologically described,' 2 Warm its fitness was determined by monitoring its survival. 3 Cool Fitness was then compared with ecological behavior 4 Cold in order to reveal their relationship. Over several obI Still Wind speed 7- Windy servations. every lizard accumulated a series of values for each parameter presented in Table 1 . Because behavioral traits may change with age, and because the age (size) distribution of survival classes were different (Table 2). comparison of the behavior in 1972, was measured with a Dwyer Wind Meter1'and of surviving and nonsurviving lizards was made at seval (4) from still to -32 kml eral age intervals over their juvenile summer. l'wothen grouped into e q ~ ~ranks way analysis of variance (Model I , nonorthogonal mulh. tiple analysis of variance exact computation) segmented the observed variation within each behavioral Because predation is thought to be a major source parameter between the factors survival and age class. of mortality in this species (Tinkle et al. 1962. Tinkle Survivors were older and larger on the average. pri1967h). and because the enclosure walls excluded rep- marily because they were seen over the whole sumtilian predators. predators were introduced into 3 of mer. while observations of nonsurvivors naturally the enclosed plots. Because of its known predation on ceased after their death. My analysis removes this arU . .ctirt1.chlirirrt1tr (McKinney and Ballinger 1966) and tifact from the comparison of survivors and nonsurits small size and abundance. Arizot~ii c4cgtrnc was vivors. Age was estimated from snout-vent length used as the primary predator in the enclosed plots. (SVL) (Tinkle 1967h: p. 137) at each lizard's first capThe leopard lizard (Croriiph~,rlr.c1i~islizc,t7ii)was capture and subsequently computed from the number of tured locally and used as another predator species. No days elapsed since the first capture date. The values predators were introduced into 1 of the enclosed plots. of each behavioral trait were gathered for each year Predation pressures ranged from no introductions to in 10-day age classes. Each lizard was often repre7 C. \t,i,slizc,t1iiper plot (which soon dwindled to =2 or sented in several age classes (if i t survived > 10 days) 3 due to escape. death, or summer retreat ~ ~ n d e r -and sometimes more than once in any particular age T A B L I2.

Age and sno~lt-ventlength ( S V L ) differences between s u r \ i \ o r s and n o n s ~ ~ r v i v o r ,

- -

- -

Year

i

-+ S D

S V L at first capture

1971 1972

31.55 -+ 5.11 3 0 . 1 -+ 5.95

Age in days at first capture"

I971 I972

32.7 30.2

Mean S V L

I972

33.88

~-

N o n s u n i \ ors

S u n i \ ors L'

I82 27 1

.i2

SD

29.59 i 4.86 26.56 -+ 3.85

.-

L'

P

60 III

30 daqs) across the plot halves did. however, indicate habitat selection ancllor differential rnortalitq . Signiticantlq fewer observations of these lizarcls were made on the moclitied side ~c~r.51i.5 the con5.50. /) < .05). Data on home trol (82 vs. 116. "X ranges before the habitat niodification h e r e not gathered. s o a direct comparison of lizard movements foll o ~ i n gmoclification cannot be made. H o ~ e v e r ,the distribution of all lizarcls over the 1st 5 days of the experiment suggestecl the ovet-all distribution of older juveniles to be due primaril) to habitat selection and not mortalit) (no deaths occurred over the 1st 5 daqs).

STANLEY F. FOX


FIG.3 . Scatterplot and least-5quares regre\,ion lines for total fecal production against live lizard weight for ,urvivors and non\urvivor\ ( o = 5~1rviv01.s. upper line: x = nonsurvivor\. lower line).

Significantly more lizards n,ere observed on the unnioclified plot half than the moclifiecl half (40 vs. 22, respectively: X' = 4.7, p < .05). Although the olcler juveniles distinguished the 2 sides. the hatchlings apparentlq clicl not. Evidence of rest[-ictecl movement of hatchlings a h a ) from theit- hatch sites ('1 inhle I967h). probitbl) explains the fitilut-e of the hatchlings to clistinguish the sides. I t is conceivable that the hatchlings were distributed ~tnequall)between dense and sparse vegetation ~\,ithinthe plot hitlves. Nonsut-vivot-s, either by choice 01- aggressive exclusion from the control sicle, utilizecl the modified plot half significantlq more than did the surviving lizards. Of the older juveniles. tn.0 thirds of the observations of eventual nonsurvivors were on the modified sicle. ~ h i l eonly one thit-el of the observations of survivors were macle on that side ( l a b l e 8). Among these same older juveniles. the survivors (those pt-edominant on the unmodified half) were signific:tntly larger than those that perished (Mann-Whitney U-test: Z = 4.54 standarcl deviations). Preponderant observation of nonsut-vivot-s on the cle2u-ed plot half I-uns parallel to the findings that nonsurvivot-s n.ere more often seen :tssoci:~ted with sparse vegetation and maintained home I-anpeswith greater clensities of such vegetation.

7 he results of this stud) appear to reaffirm the concept of natural selection within a beh;tvioral context: among the total phenotbpic variation a limitecl range of phenotqpes enjoys enhanced sut-vivorship. 1wo important considerations weaken this intet-p~-etation. however. Fit-st. 1 have no information on the he[-itabilit) of the behavior 1 observed. Natural selection requires such a genetic basis. I wo lines of evidence imperfectly suggest, though, that the behavioral difh differences. ferences might be associatecl ~ i t genetic Some of the same lizards of 1972. which w r e fc)llowed behaviorall), were also characterized for several niorphological and biochemical traits. Stabilizing selection Mas observed on 3 uncorrelatecl scale ch;tracters over the same summer intet-vltl (Fox 1975). Lizards that survived the summer were of medial scale character states and exhibited the strategy of behavior outlined in this paper. Because scltle characters are generally thought to be geneticall) determined. the "sut-vivortype" suite of behavioral characters appears associateel ~ i t hparticular genotqpes. ltlthough not necessarily clirectlq. The biochemical investigation showed a significant heterozqgote advantage at I locus (M-6-

Summer 1978

NATURAL SELtCTlON ON U T A

PI) out of 6 sho~\,ingvariation ( F o x 1973). Survivors h e r e heterozbgotes at this locus and s h o ~ \ , e patterns d of activit) different ft-on1 homozygotes (nonsurvivors). -1 his enz) matic eviclence implies also that the behavioral differences h e r e associatecl hith different genotqpes. 1' he second line of evidence that indirectl) suggests a genetic basis to the behavioral cliffel-ences lies in a studq on the same species b) l inkle (1965). His observation of the unequal survival of offspring from different parents cluring increased juvenile densities argues fur selective survivorship within a genetic context. '1 he association of a particular pattern of activitq to survival (this stud)) and associittion of particular genotqpes to surviv;tl (Tinkle's [I9651 study) imples a corresponding association of genotq pes to pattern of activity. Unfot-tunittely . offspring from the same parents share :I common niicl-oenvironniental influence as ell as similar genomes, so that the relative importitnce of each to survival is unknown. I his, of course, clilures the implied association of ecological behavior and genot) pe. '1 he second heitkness I-elating to the interpretation of natural selection on the observed behavioral patterns is that on14 survival n,as nieasurecl and assumecl to be a measure of titness. Overall fitness inclucles aspects of reproduction as ell. I here I-emtins the possibility that those juveniles that lived but shohed the nialitditptive pattern of behavior as regards to survival greh to become verq fecuncl aclults indeed. Fretell (1972) hits s h o ~ \ , an n inverse relation between survival and successful pt-oduction of offspring in field sparrows and implied similar I-elations in other birds. Because 1 Iitck data on reproductive success of the lizards I studied. 1 hesitate to extend Fretwell's (1972) finding to U . . s t c o ~ . ~ h ~ t r i teven r ~ ~ t r though . i t remains a possibilit) . '1 he 2 studq seasons appear to illustt-ate the effect of intense and relitxed selection pressure. The 1st ) r was chat-acterized b) fewer liz;tt-cls (less competition) and fewer introduced predators (less predation). The 2nd qr wits charactel-ized both by more lizards ancl more introduced predators. For the most part, survival in the 1st yr h a s inclependent of the behavioral vat-iables. but for the 2nd wits relatecl to the strategq of behavior each lizard sho~\,ecl.I t is my contention that in the 2nd qr. aspects peculiar to the individuals or to the home ranges they maintained itllowed survivors to forage quickl) ancl efticientl) under optimal ecological conclitions before retiring to safer seclusion underground or cleep hithin i t shrub or other refuge. Those lizards that hould not live out the summet-. h o ~ e v e r . foraged less efticientl) and so requiring more time each day to gather fc)od. were active under a greater diversity of ecologic:tl conditions including suboptim:tl ones. The comp:~risonof observation frequencq (dabs between successive observations) of individual lizards

843

T A B L8.~ Obser~edand expected (in parentheses) numbers of obser~ation\ ol' \ L I ~ \ i \ ors and non\ur\ ilors older than 30 day\ on t h e ~~nclearcd control \idc and the cleared s ~ d eo f plot 5 . X' = 14.01: p < .001 -.

Plot hall'

Sun I \ or\

-

Nonsur\ i\ors

Control \ ~ d e Cleared \ ~ d e Totals

suggests that i t subgroup of the population (the survivors) h e r e inactive h h e n others were active. Such voluntat-q inactivity has I-arelq been reportecl for lizards (Regal 1967. Soul6 1968. Pearson and Braclfc~-d1976). Part of the explanation may lie in the relative costs of thermoregulation for survivol.s ancl nonsurvivot-s. Sniitller liz:trds. ~ \ , h i c ha r e mostlq nonsurvivors. woulcl be predicted to assume i t c t ~ v ~ t ) over a bt-o21clet-temperature (and thus ditil) temporal) range than larger lizarcls. assuming careful thermoregulator) behavior (Huey ancl Slatkin 1976). Even hen size is contt-ollecl, however. the nonsut-vivors shoh activitq over a greater diversitq of ambient temperatures than sut-vivors ('1 able 5). and survivors arid nonsurvivors respond differentiallq to ambient teriiperatures as theq uniforml) grow olelet- (Table 3). he more open home ranges of nonsurvivot-s (Iable 6) could recluce t h e r m o r e g u l a t o ~ costs as per Huey (1974). although suitable basking sites on home ranges of survivors seemed abundantl) available. I t seems likelq that the reduced tenipot-:tl activit) schedule of the survivors reflects a risk, not an energetic cost, of thermoregulatot-q beh;tvior. Lizarcls that minimize their aboveground itctivit) shuttling between sun and shade also minimize their risk of preclittion. The specialized conditions under ~ \ , h i c hsut-vivors h e r e seen may have contributed to individuitl fitness in 2 waqs. First. the specialization ma) relate to microh:tbititt differences in qualitq of foocl (insects). Similarity of individual fecal production between survivors and nonsut-vivot-s suggests that inclividual fc)od consumption was equivalent over the survival classes. But the activity patterns suggest that the nonsut-vivors took more time to find their food. Because survivors and nonsurvivors were associatecl hith different vegetation, it is conceivable that each group selected food from a different arraq of preq items 01- from different densities of the same preq items. O r second, the specialization may relate to relative predation risks under different ecologic;tl conclitions. It is re;tsonable that lizarcls active under conditions clifferent from those of theit- preclators would s h o ~increased survivorship. Because most of the predators introcluced into the plots h e r e nocturnal or crepuscular. activity earl) in the morning and late in the evening n.as avoided bq survivors. Associated ~ \ , i t these h times of preclatot- activit) are lowet- temperatures. and in the evening. cloudier conditions- eathe^ he^- avoicled bq survivors.

S T A N L E Y F. FOX

-

Xo PHENOTYPE

Flc,. 4. Modcl illu\trating '~~nidirectionals t a g i r i n g sel r c t i o n ' . X , , = population mean before selection: X , ulation mean after selection.

=

pop-

Even during the dab. in cool. overcast. and usuallq wind) i:onclitions. normallq nocturnal snakes h e r e ob\ervecl in and near the stud) area. It is not certainlq known if survivors curtaileel activit) under these specific conclitions. but the results indicate that the) probabl? did. The observed differences in activity between survivors and nonsurvivors suggest that predation pre:,sure %;IS significant. Survivors h e r e apparentlq galhering food as quicklq as possible so ;is to minimize predation risk; the) were time minimizers (Schoener 1971). Regardless of the proximate agent of mortality. it is evident that differenti:~l survival of juvenile phenot) pe4 with respect to 4 different ecological variables ultimatelq produceel nieasur;tble behavioral changes in resource utilization over at least I summer in this population. If the behavioral phenotbpes are controlled to some extent geneticall) (either directly or indirectl)) then natural selection on behavioral strategies has been observed. Is this selection directional or st;tbilizing'.' Survivors differed from nonsurvivors along 4 behavioral variables in 1971. This is not, however. necessaril) a demonstration of true clirectional selection. Rather. this may be a special case of stabilizing selection. When extreme phenotypes are equivalently reniovecl from both sides of the adaptive mode. no mean difference in the distributions of wrvivors and nonsurvivors will be observeel. If, ho\+,ever. the elimination of sides is elisproportionate. then the population mean may well qhift cluring selection. I'his effect occurs in the situation when the aclaptive mode is up against an ecological or phqsiological barrier. e.g.. the optimum sky conditions for activity are no clouds. or the optimum c l ~ ~ t csize h is 1 egg. Because there exists no clearer skq condition than no cloucls nor clutch size less than 1 egg. 2111 the phenotypic variation accumulates in only 1 direction. Stabilizing selection (the elimination of


extreme phenotypes) pushes the drooping population the barrier and back to~\,arclits mean back to~\,:~rcl aclaptive mean (Fig. 4 ) . Although comparison of a generation's means before and after selection maq show a shift. it is quite likelq that the population mean ma) remain unchanged over several generations of uniclirectional stabilizing selection. Breedlove and Ehrlich (1968) have described :I similiu- example of this tqpe of selection for early f l o ~ e r i n gtimes in a lupine. What maintains the behavioral v;~riation'! Under relaxed selection pressures ( I97 1 ) variouh phenotq pes survivecl equivalently. But this is insufficient to retain the extreme tqpes selected against cluring conditions similar to those of 1972. To persist, those phenotqpe4 must s h o ~;in o c c a s i o ~ ~ arelative l fitness greater than the fitness of the phenotbpes which survived in my studq and subject to specific limits (Haldane and . l a akar 1963) or specific conditions of environmental tolerance (Levins 1968). Pianka (1967) has s h o ~ nthat precipitation in the North American deserts is highl) variable. Plant and insect densities are tied to precipitation ancl thus create temporall) variable environments for the lizards of these deserts (Blair 1960: 1inkle I967h: Hocldenbach and Turner 1968; '1 urner et al. I969tr. h. 1970: Pianka 1970: Nagy 1973: Parker and Pianka 1975). I t is possible that in niq studq area different strategies of lizard activity are fr~vorecl as the environment and conconiiti~nt selection pressure change. In years of scarce food and few preclators, generalized behavioral phenotypes might be vorecl. ~ h i l eunder conditions of abundant foocl ancl severe predation. speci:~lized phenotqpes might gain advantage. '1 he latter conclitions probablq prevailed cluring the second gear of mq study. Indeed, i t n,as seen that the more specialized phenotqpe M:IS the \urvivor. 1he extent to M hich this and other conventioni~l mechanisms that maintain genetic variation are applicable to the population and traits I studied is i~nknown. An alternative hypothesis explains the behavioral variation as a bq-prociuct of the social organization of this lizard. It is outlineci be lo^.

I t is like14 that the fitness of a lizard is a propertq of its home range. Home ranges of s ~ ~ r v i v o rwere s qualitatively different from those of nonsurvivors. Procurement of a superior home range may depencl somewhat on genotype. Even if the behavioral variables measurecl in this stud) prove to be solel) a property of a lizard's home range. genotqpe probably influences home-range acquisition (either b) the juvenile or through ovipository site selection by its mother) and exclusion of others (territorial defense). Other studies. primarily of birds. have indicated a relationship between fitness and territory size or quali t ) (Watson and Miller 1971. Holm 1973. revie\+,eclby Brown and Orians 1970). '1 hose indivicluals ~ i t larger h o r otherwise superior territories survive better o r pro-

Summer I978

XATURAL SELECTIOX ON UTA

duce more offspring than those with inferior territories. 7 inkle (1969) shoned that ad~lltfemales of Urrr .\ttr11.th~iritr11rr from large territories produced more boung than females from small territories. The degrec of individual aggressiveness o r relative dominance may be important in cleterrnining the size or qualit) of a territory an animal can secure ancl retain. 'I his appears to be the case in some territorial birds (Coulson 1968. Watson and bliller 1971). lekforming birds (Hogan-Warburg 1966). lek-forming ungulates (lxutholcl 1966). ancl territorial lizards (Brattstrom 1974. Ferner 1974. Licht 1974). bly stucly showed qualitative differences between the home range4 of surviving ancl nonsurviving juvenile lizards and irlclicates the possible importance of wcial interactions influencing these differences. When one half of plot 5 of my stucly was cleared of dense vegetation. the older juveniles favored the uncleared side. .I hose who spent some time in the cleared side suffered a high rate of mortalit). It seems possible that these individual4 \+,ereaggressively excluded from the ~lnrnodifieclside and forced to remain on the high-risk plot half. Ferguson and Bohlen (1978) have shown that the later-hatched. larger, more aggressive juveniles of .5(~(~10~1orii.\ ri~~~lriltrtri.\ survive better than smaller. less aggressive juveniles of the same late season. I'he late season is ;I t i ~ n eof decreasing food suppl) and increasing lizard densit). resulting in increased competition. As competition increases, di5persal from crowded areas is also more marked. presurnablq a result of exclusion bq the more aggressive larger juveniles. The juveniles ubserved in my stucly on the cleared habitat \+,eresmaller in size than those remaining on the uncleared sicle. Presumably. these smaller lizards \+,ere found on the cleared side as a result of aggressive exclusion b) the larger lizards precipitated by increased competition when on14 one half of the plot reniainecl suitable for the same number of juveniles after habitat modification. In the other ~lnmodifiecl plots. the s~lrvivorsand nonsurvivors \+,ere5till differentially distributed across habitat of heterogeneous qualit), even though the) weren't clumped into large areas of poor arlcl rich quali t ) . 1 feel t h w e lizard4 inhabiting superior home ranges (promoting s~lrvival)suciallq exclude others to suboptimal home ranges. I'his condition is broadl) analogous (u Brown's Level 2 in his model of territorialit) (Brown 1969). Level 2 is the density at M hich some inclividuals are forced to occup) territories of lesser quality. ancl as s ~ l c hproduce fewer offspring. In my study. the superior habitat is securecl b) lizards which in turn relegate others to poorer habitat. which i 4 . however. interspersed among 4uperior habitat. I'hose with home ranges uf superior q~lalit)are marked for better s ~ ~ r v i v a l . I t appears that a superior home range extends a selective advantage tu the inclividual bq increasing the chances of Rand's (1967) item one, i.e.. its securing

845

a necessar) share of environmental resources. in this instance foocl and predation refugia. I'hose forced to establish home ranges in poor habitat consequentl) alter their behavior in the fashion of the nonsurvivors of this stud). It is possible had they obtained home ranges in wperior habitat. the) ~vouldhave shown the behavior of survivors. I'he behavior ma) folio\+, from the quality of home range inhabited. but i t seems unlikely that the acquisition of a s~litablehome range is ranclom.1n this sense. the behavior I observed Mas tnerel) a marker for those genetic qualitie4 which promote the formation of a home range in superior habitat. As long as there always exist more hatchlings than segments of superior habitat, some lizards \!,ill alnays fail to obtain home ranges of superior quality and exhibit the behavior of eventual nonsurvivors. The proairnate factor is the activit) a lizard sho\+,sunder various ecological conditions. ancl n h a t may be directl) influential to his survival. The ~lltimatefactor is the combination of characteristics M hich affects the quali t y of home range a lizard obtains. The home range is the link bet\+,eenthe 2 . The home ranges of the juveniles of this stucly were obtained by consec~ltiveresightings of inclividuals ancl not points of defense. I have. therefore. refel.recl to these areas as home ranges. I have observed social interactions comprising bobs. displabs and e v e n t ~ ~ a l supplants in juvenile Uttr stir11.\hrrriio1r1.As the lizards g r o ~ s.o d o their home ranges. and the) begirl to exclude others: aclults are territorial. Territorialit) is developing in these juveniles over their first summer ancl social relations ma) \+'ell play a major role in the qu:~li t ) of territorq an inclividual event~lall)ubtains. 1hose qualities promoting the establishment of a superior home range may include clominance. A gro\+,ingbod) of literature emphasize5 the relationship of dominance to fitness. not only in obtaining optimal territories, but also in direct competition for foocl (Fretusell 1969. 1972, revieused by CI-ook 1970). It \+,auld be enlightening to explore more filll) the possibilit) uf such a relationship in Utii .\ttr1r.,hrrriir11rr.

'l'his research is part of a dissertation wbmitted to the Graduate School of Yale Universit) in partial fulfillment of the requirements for the Ph.D. degree. I thank my thesi\ advisor. Dr. R. S. bliller. for his assistance and advice. ?'he other members of my committee. Drs. G . E . Hutchin\on. K . S. Thonison. and J . A. W. Kirsch also contributed aid. The conception and early development of thi\ research were largel) a result of discussions and exchange of ideas with Drs. X. P. Ashmole and T. hl. Uzzell. I particularl) thank Drs. M. C . Baker. E. L. Goldstein. and F. H. C . Hotchkiss for \timulating conversations. I thank Dr. 0. Tinkle for his basic study of L/. .s~trt~sh~rritrtitr from uhich m! study greu. and for his research suggestion\. Drs. P. hlarler and R . Taniarin commented on the manuscript. Texas Tech University extended the use of their field station at Wink. Texas. and Dr. R. Packard facilitated arrangements. I thank the late hlr. E . Ve\t for the use of a part of his ranchland, upon which the stud) uas conducted. hlr. J .

846

STAKLEY F. FOX


Haldane. J . B. S . . and S. L). Jaqaker. 1963. Polqmorphism due to selection of varying direction. Journal of Genetic\ 58: 237-242. Hecht. M. K. 1952. Katural selection in the lizard genus Ariatclligc,r. Evolution 6: 111-124. Hoddenbach. G. A , . and F. B. Turner. 1968. Clutch \ire of the lizard Uttr rtrt7shuritrt1tr in southern Kevada. American Midland Katuralist 80:262-265. Hogan-Warburg. A. J . 1966. Social behavior of the r ~ ~ f f . Philorntrc.hus pugtltr.~( L . ) Ardea 54: 109-229. Holm. C . H . 1973. Breeding sex ratios. territoriality. and reproductive success in the red-hinged blackbird (Agc~ltriirs Averq. R. A. 1971. E\tirnates of food con\umption bq the phr,rt~ic,cju.t).Ecologq 54:356-365. lizard Lcic.c~rttr~,i\.ipcrrtrJacquin. Journal of Animal Ecologq Hueq. R. B. 1974. Behavioral thermoregulation in lizards: 40:35 1-365. importance of associated costs. Science 184: 1001-1003. Blair. W. F. 1960. The r ~ ~ s tlizard. q A population s t ~ l d q . . and M. Slatkin. 1976. Cost and benefits of lizard Universitq of Texas Press. Austin. 'l'exas. USA. thermoregulation. Quurterlq Review of Biology 51:363Brattstrorn. B. H. 1974. 'l'he evolution of reptilian social 384. organization, American Zoologist 1 4 : 3 5 4 9 . Inger. R. F. 1943. Further notes on differential selection of Breedlove. L). E . . and P. R. Ehrlich. 1968. Plant-herbivore variant juvenile snakes. American Katuralist 77:87-90. coevolution: lupines and lycaenids. Science 162:671-672. Iruin. L . K . 1965. Diel activitq and social interaction of the Broun. J. L . 1969. Territorial behavior and population regulation in birds. Wilson Bulletin 81:293-329. lizard U r u stutishuriutici sttgtirgc,ri. Copeia 1965:99-101. . and G . H. Orians. 1970. Spacing patterns in mobile L e ~ ~ t h o l dW. . 1966. Variations in territorial behavior of ( N e ~ ~ m a n1896). n Beanimals. Annual Revieu of Ecologq and Systematics Uganda Kob Atlct~oftr K o h tho~~rcisi havior 27:2 15-258. 1:239-257. Charno\. E . L . . G . H. Orians. and K. Hyatt. 1976. Ecolog- Levins. R. 1968. Evolution in changing environment\. Princeton University Press. Princeton. N e u Jerseq. USA. ical implication\ of resource depression. American NdtuLevinton. J. S . 1971. Control of tellinacean (Molluscs: Birali\t 110:247-1-59, valvia) feeding behavior by predation. Limnologq and Clarke. R. L). 1972. The effect of toe clipping on survival in Oceanography 16:660-662. Copeia 1972: 182Fowler's toad (Byfi, ~t~r,r,tll~oir.sri,fi,~~~leri). Licht. P. 1974. Response of At~oli.\ lizards to food supple185. Colwell. R. K.. and 0. J . Futuqma. 1971. On the measurementation in nature. Copeia 1974:2 15-22 1. McKinneq. C . 0..and E. R. E . Ballinger. 1966. Snake predment of niche breadth and overlap. Ecology 52:567-576. ators of lizards in western '1 exas. Southwestern Naturalist Coul\on. J . C. 1968. Differences in the qualit) of birds nesting in centre and on the edges of a colon). Nature 217:47811:410412. Medica. P. A,. and F. B. 'l'urner. 1976. Reproduction by 479. L/tci srut~~huriciritr(Reptilia. Lacertilia. Iguanidae) in southCrook. J . H . 1970. Social organization and the environment: ern Nevada. Journal of Herpetology 10:123-128. aspects of contemporary social ethologq. Animal Behavior Mertens. R. 1947. Studien zur Eidonomie ~ ~ 'l'axonomie n d 18: 197-209. der Ringelnatter (,\trrrir t ~ c i r r i . ~Abhandlungen ). der SenckCulver. D. C. 1972. A niche analysi\ of Colorado ants. enbergischen Naturforschenden Gesellschaft 476: 1-38. Ecologq 53: 11-6-1 3 1. Dunn. t:. R. 1 9 1 . The variations of a brood of water snakes. Nagq. K. A. 1973. Behavior. diet. and reproduction in a Proceeding\ of the Biological Societq of Washington 28:6 1desert lizard. Sa~rrot)iuluaohcaus. Copeia 1973:93-102. Parker. W. S . . and E. R. Pianka. 1975. Comparative ecology 68. Ehrman. L.. and P. A. Parsons. 1976. The genetics of beof populations of the lizard Uttr artrtr.thirritrt~tr. Copeia 1975:615-63?, havior. Sinauer Associates. Sunderland. Massachusetts. Pearson. 0. P.. and D. F. Bradford. 1976. 'l'hermoregulation USA. of lizards and toads at high 'iltit~ldes in Peru. Copeia Ferguson. G . W.. and C. H . Bohlen. 1978. it7 press. De1976: 155- 170. mographic analy\is: a tool for the stud) of natural selection of behavioral traits. Pages 227-243 it1 K . Greenburg. and Pianka. E . R. 1967. On lizard species diversity: North American flatland deserts. Ecology 48:333-35 1 . P. McLean, editors. 'I he behavior and neurologq of lizards. Kational Institutes of Health. Rockville. Marqland. USA. -- . 1970. Comparative autecology of the lizard C t ~ r t l i iclophorlrs tigris in different parts of its geographic range. Ferner. J . 1974. Home-range size and overlap in Scc~loporlrs Ecologq 51:703-720. utrclu1uru.t c,rythroc.hcillrs (Reptilia: Iguanidae). Copeia Pielou. E. C . 1966. The measurement of diver\ity in differ1974:332-337. ent types of biological collections. J o ~ ~ r n of Fisher. R. A. 1970. Statistical methods for research worka l 'l'heoretical e r \ . Hafiier Publishing Company. Darien. C o n n e c t i c ~ ~ t . Biology 13: 13 1-144. Rand. A. S . 1967. The adaptive significance of territoriality USA. 14th Edition. Pages 99-101. in iguanid lizards. Pages 106-1 I5 it7 W. Milstead. editor. Fox. S . F. 1973. Katural selection in the lizard Uttr sftrt7.sSymposium of lizard ecology. University of blis\ouri hiiritirrtr. Doctoral thesis. Yale University. New Haven. Press. Columbia. M i s s o ~ ~ rUSA. i. Connecticut. USA. . 1975. N a t ~ ~ r aselection l on morphological pheno- Regal. P. J . 1967. Voluntarq hqpotherrnia in reptiles. Science 155: 1551-1553. type\ of the lizard Urtr .trcit~.shuricitiu.Evolution 29:95-107. Fretwell. S . L). 1969. Dominance behavior and winter hab- Schoener. T . W. 1971. Theorq of feeding strategies. Annual Revieu of Ecologq and Systematics 2:369404. itat distribution in Juncos. Bird Banding 40: 1-25. 1972. P o p ~ ~ l a t i o nins a seasonal environment. Shannon. C. E.. and W. Weaver. 1949. The mathematical theory of communication. Universitq of Illinoi\ Press. UrPrinceton University Press. Princeton. K e u Jerseq. USA. bana. Illinois. USA. Go\s-Custard. J . D. 1970. Feeding dispersion in some overuintering wading birds. Pages 3-35 it1 J . H . Crook. editor. Soule. bl. 1 9 6 7 ~ .Phenetics of natural populations I . Phenetic relationships of insular populations of the sideSocial behavior in bird\ and mammals. Academic Press. blotched lizard. Evolution 21:584-591. London. England and Kew York. New York. USA.

Williamson lent moral \upport and the use of a fine calculator. F i n ~ ~ l l qI .thank my wife. Karen. for her help at all stages of this work. Fund\ were provided by NSF Predoctoral Grant GB 31577. the Department of Biology. Yale Universitq. and a grant from the Societq of Sigma Xi. Major pieces of field equipment were provided by the Peabodq Museum of Natural History. Yale Univer\ity. Additional aid u a s obtained from K S F Grant GB 33102 to P. blarler.

Suninicr 1978

NATURAL SELEC'I ION ON L/TA

847

196717. Phenetics of natural populations 11. Ashni1970. Comments on laboratorq survivorship in t u o metr! and evolution in a lizard. American N a t ~ ~ r a l i s t species of the lizard genus Uttr. Copeia 1970:381-383. 101: 141-160. ----. O. LlcGregor. and S . Dana. 1962. Home range ecol. 1968. Bodq teniperaturez of quiescent Strtor ,qrtrtrogq of Utcr .srutrsl~uriutiusfc:jtit~yeri.Ecologq 43:223-229. Turner. F. B.. G . A. Hoddenbach. P. A. bledica. and J . R. c/trcj\,rr.\ in nature. Copeia 1968:622-623. Stein. R . A , . and J. J . Llagnuson. 1976. Behavioral response Lannoni. 1970. The deniographq of the lizard. Urtr .sttitrshlcricctici Baird and Girard in southern Nevada. Journal of of craqfish to a fish predator. Ecologb 57:75 1-76 I . I'inkle. L). W . 1965. Population structure and effective \ire Animal Ecologq 39:505-5 19. of a lizard population. Evolution 19:569-573. Turner. F. B.. J. R. Lannom. Jr.. P. A. bledica. and G . A. Hoddenbach. 19690. Densit! and composition of fenced 19670. Home range. den\itq. dq namics. and structure of a 'I eka5 population of the lizard Uttr .trtrrrahlrritrtrtr. populationz of leopard lizard\ (Ct.ortrph>rrr.\ \i,isli;c~trii)in Page5 5-29 it1 W. blil\tead. editor. Lizard ecology: a sqmsouthern Xevada. Herpetologica 25:247-257. pozium. U n i er\itq ~ of Llissoul-i Press. Columbia. Mi\souri. Turner. F. B.. P. A. Medica. J . R. Lannom. Jr.. and G. A. USA. Hoddenbach. 19696. A denlographic analhsis of fenced -. 196717. 1he life and deniographq of the \ide-blotched tigris . in populations of the whiptail lizard. C'trc~t~riclop/~or~r.\ wuthern Xevada. S o ~ ~ t h w e s t e rXaturalist n 14: 189-202. lizard. Uttr .\r(r~r\hro.itrtrtr. bliscellaneous Publication\ of the Lluseum of Zoology of the University of Michigan 132. M'atson. A. 1967. Population control by territorial hehavior in Red Grouse. Xature 215: 127441275, . 1969. Evolutionary implications of comparative and G . R. Miller. 1971. 'l'erritory size and aggression population \ t ~ ~ d i eins the lizard L//cr .srtrt~ahirritrtrcr.Pages -, 133-1i4 i t i International Conference on Systematic Biology Red G r o ~ ~ population. se Journal of Animal in a fluct~~ating Ecology 40:367-383. at the U n i \ e ~ - 4 t hof Michigan 1967. Systematic biologq. National Academy of Sciences Washington. District of Columbia. USA. Number 1692.

-.