Digestiveecology of two omnivorous Canarian lizardspecies (Gallotia,Lacertidae)

AlfredoV alido 1,ManuelNogales

Departamento deBiología Animal (Zoología), Universidadde La Laguna,E-38206, Tenerife, Canary Islands, Spain 1Authorfor reprint request; current address: Dept. of Ecologyand Genetics, AarhusUniversity, Ny Munkegade Building540 DK-8000 Aarhus C, Denmark e-mail: [email protected]

Abstract. Omnivorousendemic Canarianlacertids ( Gallotiaatlantica and G. galloti)donotpresent any speciŽ c digestiveand physiological adaptations to herbivorous diet, compared to species andpopulations with a different degree ofherbivory in the Canarian archipelago. The only characteristics thatcould be related tothe type of diet were thenumber of cusps per tooth (between species) andthe number of small stonescontained in droppings (between species andpopulations). The rest ofmeasured traits were correlated withlizard size andfor this reason G. galloti has longerintestines, heavier stomachs andlivers, more teethand cusps, and longer gut passage. These data suggestthat body size isamajor determinantof the reliance onplant food (mainly  eshyfruits) in these lizards andfacilitates mutualisticinteractions with  eshy-fruitedplant species.

Introduction Structuraland functional adaptations of the digestive system for optimalprocessing of plantmatter have been described for severalomnivorous and herbivorous vertebrates (see reviewsin Lö nnberg, 1902; Ziswiler and Farner, 1972; Skoczylas, 1978; Chivers and Hladik,1980; Guard, 1980; Sibly, 1981; Jordano, 2000). Althougha lowproportion (about 2-3%) of knownlizard species eat some quantities of plantfood (Pough, 1973; Iverson, 1982; King, 1996), several morphological, anatomical andphysiologicalmodiŽ cations have also been described for herbivorouslizards (Gekkota, Iguania,and Scincomorpha), which may increase the digestiveefŽ ciency of lizardsfeeding onthis resource (Throckmorton, 1973; Johnson and Lillywhite, 1979; Zimmerman and Tracy,1989). The mostimportant of thesetraits are: larger body size(Szarski, 1962; Sokol, 1967;Pough, 1973; Iverson, 1982; Schluter, 1984, butsee V anDamme, 1999); tendency to showa largerintestine/ SVL ratio(Skoczylas, 1978; V anDevender, 1982; Dearing, 1993);

© KoninklijkeBrill NV ,Leiden,2003 Amphibia-Reptilia24: 331-344 Alsoavailable online - www.brill.nl 332 AlfredoV alido,Manuel Nogales longertransit time that permits a suitabledegradation of plant tissues (Christian et al., 1984;Troyer, 1984; Karasov, 1986; W aldschmidtet al., 1986; Zimmerman and Tracy, 1989;V anMarken Lichtenbelt, 1992); smaller stomachs (Skoczylas, 1978); three-cusp or polycuspdentition (Hotton, 1955; Sokol, 1967; Montanucci, 1968; Greer, 1976;Iverson, 1982;Mateo and Ló pez-Jurado, 1992); intestinal caeca and compartmentalizati onofthe colon(Iverson, 1982; Zug, 1993); and the tendency to ingest small stones (Sokol, 1971; Skoczylas,1978). ModiŽcations to lizard digestive systems have been described mainly in folivorous speciesand their presence has been justiŽ ed, by the need to process Ž ber-richplant components,more difŽ cult to digest (Iverson, 1982; Cooper and V itt,2002). However, littleis known about the digestive modiŽ cation of lizards that primarily feed on other, comparativelyeasy to digest, plant components (basically fruits,  owers andnectar) although,in some cases, the proportion of thisplant component in the diet is considerable. Generally,lizards inhabiting island habitats have a tendencytowards a moreherbivorous dietthan lizards on mainlands (V anDamme, 1999; Cooper and V itt,2002); this is the caseof theendemic Canarian lacertids (g. Gallotia),whichpresent the most herbivorous specieswithin Lacertidae (V alido,1999; V anDamme, 1999). The seven extant Canarian lacertidscan be considered omnivorous, but the contribution of plantmatter in their diet variesaccording to lizard species, body size, habitat, and season of theyear (e.g. Mateo andLó pez-Jurado, 1992; V alidoand Nogales, 1994; Pé rez-Mellado et al., 1999; V alido, 1999).However, in some cases (e.g. G. galloti from xerophyticzones in Tenerife), the plantcomponent (basically  eshyfruits) is predominantthrough the year (V alido,1999). Theprimary aim of thisstudy is to characterizethe assemblageof traitsin some Canarian lizards.W eselecteda pairof speciesthat are more distant their diet along an omnivorous gradient:G.atlantica from Fuerteventura,which is omnivorousbut basically insectivorous, andthree distant populations of G. galloti from Tenerife(T enoBajo, Bajamar e Izaña), whichvary in the importance of the plant component in their diet (from omnivorous, basicallyinsectivorous to omnivorous, mainly herbivorous). Lizardsof the genus Gallotia (Lacertidae)represent an oceanic radiation of reptiles endemicto CanaryIslands and are resolved as abasaltaxon in the family Lacertidae (Fu, 1998).This genus is characterized by seven extant species and one extinct taxon. This genushas a monophyleticorigin and the radiation took place along an eastern-western geographicgradient from themainland (Gonzá lez et al.,1996). Gallotiaatlantica inhabitsFuerteventura and Lanzarote; it isthesmallest species of the genus(adult SVL mean s 60:8 9:9mm;V alido,1999) and shows an omnivorous § D § diet,although animal components are the main preys. Gallotiagalloti inhabitsthe islands ofTenerifeand La Palma; it isa medium-sizedlizard (SVL: mean 106:4 12:1; Valido, D § 1999)and is characterized by an omnivorous diet in which  eshyfruit is the principal componentin those populations inhabiting altitudinal lower zones (V alidoand Nogales, 1994;V alido,1999). Omnivorouslizards 333

If thetraits described above represent adaptations for anherbivorous diet, we would expectthe more insectivorous populations ( G.atlantica from Fuerteventuraand G. galloti from Izaña inTenerife)to have similar characteristics (shorter intestines and transit period, heavierstomachs, presence of mono/bicuspiddentition) and to differ from thetwo more herbivorouspopulations ( G. galloti from TenoBajo and Bajamar). However, an alternative hypothesisis thatthese lizards do notneed special anatomical traits like true herbivorous lizards(Iverson, 1982; Cooper and V itt,2002), since the principal parts of theplant in the dietare  eshyfruits and  owers. Theuse of these species can also help us to understand the mechanisms in uencing dietaryvariations among insular lizard populations, speciŽ cally the general trend for relativelymore herbivorous diets when compared to continentalareas (V anDamme, 1999; Cooperand Vitt, 2002). This insular trend to increase herbivory (mainly fruits) has also importantecological and evolutionaryimplications for mutualisticinteractions with  eshy- fruitedplant species in islands(V alido,1999), a phenomenonwhich seems to occurmore frequentlyin islandecosystems (Olesen and V alido,2003).

Materialand methods

Thelizards understudy were capturedin July 1994 at fourdifferent localities on two islands: G.atlantica in Fuerteventura(V alle deTetir — 400m a.s.l.;Tetir hereafter) and G. galloti inthree differenthabitats of Tenerife (Barranco deLas Cuevas,Teno Bajo — 150m a.s.l.;Barranco deV argas, Bajamar —110m a.s.l.;and Corral del Niño, Izañ a —2293m a.s.l.;hereafter TenoBajo, Bajamar andIzañ a, respectively). Theclimate ofTetir is characterized byan annual rainfall of less than200 mm andmean temperature of 21±C.Characteristic plantsof these xeric zonesare Launaeaarborescens (Compositae), Salsolavermiculata (Chenopodiaceae),and Lyciumintricatum (Solanaceae). InTeno Bajo, the climate is dry,with an annual rainfall < 300mm, anda mean temperature ofabout 21 ±C.Themore abundantplants in this habitat include some species belongingto the genus Euphorbia: E.obtusifolia , E.canariensis , E.balsamifera .Inaddition, Plocama pendula, Rubiafruticosa (Rubiaceae) andthe introduced Opuntiadillenii (Cactaceae) are abundant.The other localityon Tenerife (Bajamar) presentsa mean annualrainfall < 400mm, amean temperature of21 ±C and is characterized byscrubvegetation where Asparaguspastorianus , A.umbellatus (Liliaceae), and Boseayervamora (Amaranthaceae), are abundant.The third locality on Tenerife, Izañ a, is ahighmountain site witha cool,dry climate andannual rainfall of about600 mm andmean temperature of10 ±C;inwinter, rainfall can turnto snow. Spartocytisussupranubius and Adenocarpusviscosus (Fabaceae) composeshrubs basically.

Procedures. We collectedrecent lizarddroppings ( n 150,at each site) ineach ofthe four localities, in July D 1994,in order to study the diet composition. Therefore, in order to cover those different microhabitats within each localityand to have a representativeapproximation of the lizard diet at populationlevel, linear transects were carried out.To avoid seudoreplication during the sampling collection, no more thanŽ velizard droppings were collectedat thesame site.Each droppingwas keptindividually and analyzed under 16 magniŽcation after £ thematerial hadbeen soaked in water. We visuallyestimated thepercentage volumein each faecal sample made upby animal andplant matter (tothe nearest 10%),and the number of animal items andseeds were recordedand identiŽed. Small stones inside droppings were alsonoted. Although this method is less accurate thanstomach contentsto trying to knowa taxonomicdescription of invertebratescomponents, in the approximation carried out inthis study (animal vs vegetalcomponent), feces analyses gavesimilar resultsto those from stomach (angular transformationon percentages ofplant matter; ANOVA, F 0:016, df 1, 118, P 0:89forcomparative data D D D obtainedfrom both stomach- ushings from 60 lizards and60 fresh droppingscollected in the same daysand sites inTeno Bajo; V alido,1999). 334 AlfredoV alido,Manuel Nogales

Theexperiments carried outto assess thegut passage time (hereafter GPT)were performedby using 20-21 differentadult lizards fromthree localities(not Bajamar) (12males and9 females fromTeno Bajo and Izañ a; 11 males and9 females fromTetir) during August to September 1994. They were putin adjacent 32 23 22 cm £ £ individualcages. These cages were locatedan environmental chamber where thetemperature variedbetween 28-30±Canddaylight period was 12h.Each individualwas previouslyacclimatized for20 daysbefore starting thetrials. This temperature rangewas chosenin accordance whitthe operative temperature ofthese lizards (see Díaz, 1994for G. galloti andMá rquez et al.,1997 for G.atlantica ).Allindividuals had ad libitum access toa diet consistingof insects (adultcrickets andColeoptera larvae) andfruits of ‘ Balo’( Plocama pendula )andtomatoes (Solanumlycopersicum ,Solanaceae). However,insects were always theprincipal remains inthe droppings of all individualssince theyprefer thisfood to fruits when they are present;water was alsocontinuously available. Lizards were simultaneouslyforce-fed twosmall glass beads(maximum diameter 3 mm; mass 0.02 g) D D togetherwith food. Cages were inspectedevery three orfour hours and faeces were crushedand visually examined forglass beads.Individual GPT was deŽned as theaverage time untilappearance inthe droppings of thetwo glass beads.Body mass was recordedat thebeginning and end of the trials, and remained constantfor the lizards from Izaña (Wilcoxontest, Z 1:08, P 0:27/ and Tetir (Z 1:53, P 0:12/ whilethe lizards fromTeno Bajo D D D D were heavier(about 2.8 g increase) at theend of the trials ( Z 3:17, P 0:002/. D D Totest ifthe data fromthe laboratory trials were similar tonatural conditions, we made parallel experiments inthe Ž eld(Teno Bajo). We put13 adult lizards inindividual cages (40 30 25cm) andthey had ad libitum £ £ access tofood (the same providedin the captivity trials) underconditions of sun/ shade.Lizards were alsoforce- fedsimultaneously two glass beadswith the food. Taking into account that GPT is inuenced by the SVL of each individual,lizards fromthe Ž eld(average S.D.; 104:9 10:8 mm, n 13/ showedno signiŽ cant size § § D differences fromthose used in the laboratory experiments (111 :2 10:4 mm, n 21)(Mann-Whitney test, § D Z 1:76, P 0:079/. D D Morphologicaland anatomical traits (SVL,intestine length, stomach andliver mass) were takenfrom some lizards collectedat thesame localitiesand dates. Furthermore, the number of teethand the cusp number per tooth inthe jaws andmaxillary bones were counted.

DataAnalysis. Data analyses includedANOV Atostudy differences inlizard body size between thetwo differentislands (and among populations), and ANCOV A(usingdiameter offecal pelletsas covariates) to compare amountof plant matter inthe diet. This approximation was usedto control lizard-size effects ondiet typeand to avoid bias in the diet comparisons among Tenerife lizardlocalities. The relationship between the proportionof plantin the diet and body size ofCanarianlizards iswell documented(Mateo and Ló pez-Jurado, 1992;V alido,1999). Also, there is asigniŽcant relationshipbetween diameter offecal pelletsand lizard size (Pearsoncorrelation, rp 0:79, P 0:036on log-transformed data; n 24mean diameter ofdroppings D D D obtainedfrom 24 individuals of G. galloti keptin captivity and with an omnivorous diet). We thereforeused droppingsdiameter as acovariate tocontrol the effects oflizard’ s size. Analysisof simple regressionwith the pooleddata (withSVL as anindependent variable) was usedto control for allometric effects inthe comparisons of anatomical andphysiological characters (Packardand Boardman, 1988). The use ofpooled data inthe regression analysisis justiŽed because theinteraction term (locality SVL)in the ANCOV A(usingSVL as acovariate and £ localityas themain effect) isnot signiŽ cant ( P > 0:11inall thetraits), indicating similar regressionslopes in thestudied populations. Residuals (deviations) of observationsfor the allometric regressionswere retainedfor ANOVAandmultivariate analysis of variance (MANOVA)usinga GLMprocedureand Type III sums ofsquares. Throughoutthe paper, means are accompaniedby standard deviations unless otherwise indicated.

Results

Variationin DietComposition. Table1 showsthe compositionof plantmatter and animal preyitems in the diet of both species of lizards from thefour studied localities. The percentageof total plant matter showed signiŽ cant differences among the four studied localities(ANOV A, F 243:45, df 3, 596, P < 0:001on arc-sintransformed data), D D Omnivorouslizards 335 withsigniŽ cantly more plant material in the lizard’ s dietin TenoBajo(79 19%, n 150/ § D andBajamar (81 18%, n 150/ thanIzañ a (21 25%, n 150/ andT etir(28 30%, § D § D § n 150)(Scheffé test, P < 0:001).The frequency distributions of plant matter show D similartrends in Teno Bajo and Bajamar (Tenerife), while the opposite tendency was observedin both Izañ a (Tenerife)and T etir(Fuerteventura) populations (Ž g. 1). All the droppingsanalyzed in TenoBajo and Bajamar had some remains of plantmatter, while in theother two localities plant material was presentin 61%(Izañ a) and69% (Tetir) of the droppingsanalyzed (table 1). Oneof theclearest differences is thehigh values of eshyfruit remains in the droppings collectedfrom TenoBajo and Bajamar(66% and 59%, respectively),and their near absence inthe other two zones: Izañ a andT etir(3% and 8%), where other plant parts (basically owers butleaves too) were moreimportant than  eshyfruits. Therefore, lizards from the highshrub mountain of Izaña (Tenerife)and from thexeric habitats of Fuerteventura(T etir) were moreomnivorous, practically insectivorous, than those from thexerophytic shrub of TenoBajo and Bajamar (Tenerife). Inthe case of thethreelocalities of Tenerife,the differencethat appeared in thedietcould berelated to bias toward collecting droppings from lizardsof different body size (e.g. a morerepresentative sample of droppingsbelonging to smalllizards in Izañ a, characterized byan insectivorous diet). However, the pattern of intra-insular differences is consistent whencontrolling the diameter (mm) ofdroppings analyzed (6 :0 1:4 mm, n 142 § D inTeno Bajo, 6 :37 1:4 mm, n 149in Bajamar and 4 :5 1:4 mm, n 147 for § D § D Izaña) (arc-sintransformation; F 241:15, df 2, 435, P < 0:001for theANCOV Aof D D percentageof plantmatter as themain effect and diameter of droppingsas thecovariate), andno signiŽcant differences were observedwhen comparing the percentage of vegetable matterin thedroppings of Bajamarand T enoBajo when controlling the dropping diameter (ANCOVA, F 0:016, df 2, 290, P 0:89/. D D D Thefrequency of smallstones (grid) contained in the droppings was remarkelydifferent inthe each populations. Teno Bajo ( n 23/ andBajamar ( n 20/ presentedthe higher D D numbersto compare with Izañ a ( n 5/ and Tetir (n 2/,for theentire sample of D D droppingsanalyzed at eachlocality.

Morphologicaland Anatomical Traits. Adultlizards from Tetir( G.atlantica ) (SVL: 57:0 3:6 mm, n 19/ were smallerin size than G. galloti from thelocalities of § D Tenerife(Teno Bajo: 105 :9 13:7 mm, n 36 Bajamar: 95:7 7:1 mm, n 15 § D I § D I Izaña: 105:1 14:2, N 15) (ANOVA, F 132:9, df 3, 81, P < 0:001 on log- § D D D transformeddata) (table 2). Additionally, no statisticaldifferences were evidentamong the three G. galloti populationsfrom Tenerife(Scheffé test, P > 0:05/. Despitethis clear difference in size between species, no signiŽ cant differences in anatomicalmeasures were obtainedamong the lizards of thesefour localities if variation inSVL istaken into account (MANOV A,Wilks’ ¸; F 0:62, P 0:85, N 60 D D D I withresiduals from thesigniŽ cant models obtained in the regression analysis of eachtrait; 336 AlfredoV alido,Manuel Nogales

Figure 1. Numberof droppingsand percentage ofplant matter (pooledfruits,  owers andleaves) in G. galloti fromTeno Bajo, Bajamar andIzañ a (Tenerife); and G.atlantica fromTetir (Fuerteventura) ( n 150droppings, D respectively). Omnivorouslizards 337

Table 1. Diet compositionin the droppings of Gallotiagalloti inTeno Bajo, Bajamar andIzañ a (Tenerife) and G.atlantica inTetir (Fuerteventura) ( n 150,respectively). % FO:Frequency of occurrence (percentage of D droppingswith the taxon present). % F:percentage ofthetotal seeds andinvertebrates found in the droppings.

Gallotiagalloti Gallotiaatlantica Teno Bajo Bajamar Izaña Tetir %FO%F%FO%F%FO%F%FO%F

Fruitremains Rubiafruticosa 11:3 3:9 2:7 2:7 – – – – Withaniaaristata 5:3 0:8 2:7 3:8 – – – – Neochamaeleapulverulenta 2 0:2 – – – – – – Lyciumintricatum 1:3 0:3 – – – – – – Plocamapendula 46 86:1 – – – – – – Opuntiadillenii 26:7 6:0 – – – – – – Opuntiamaxima – – 3:3 2:2 – – – – Atriplexsemibaccata 0:7 0:1 6:7 6:3 – – 2 17:2 Spartocytisussupranubius – – – – 9:3 88:9 – – Ficuscarica – – 19:3 66:3 – – 9:3 81 Asparaguspastorianus – – 20 9:3 – – – – Asparagusumbellatus – – 1:3 0:4 – – – – Boseayervamora – – 26 9:1 – – – – Undetermined 2 2:6 2 1 1:3 11:1 0:7 1:7 Invertebrates Coleoptera 6:7 18:2 2 8:8 40 29:5 21:3 12:9 Hemiptera 2:7 7:3 1:3 5:9 22 18:4 3:3 1:9 Diptera 6 16:4 12 58:8 12:7 12:3 2:7 2:7 Formicidae 10:7 50:9 6 26:5 19:3 25:4 34:7 72:6 OtherHymenoptera 2 5:5 – – 15:3 13:1 4 3 Orthoptera 0:7 1:8 – – 2 1:2 2:7 1:5 Gasteropoda – – – – – – 9:3 5:3

n % n % n % n %

Droppingswith plant matter 150100 150 100 92 61 :3 104 69:3 Droppingswith only plant matter 2315 :3 22 14:7 0 0 1 0:7 Droppingswith invertebrates 127 84:7 128 85:3150100 149 99 :3 Droppingswith only invertebrates 0 0005838 :7 46 30:7

%volumeof fruit remains 66.3 59.3 3.3 8.4 %volumeof otherplant parts 12.5 21.6 17.3 19.9 seetable 2). These results conŽ rm thatapparent differences between species, or among populations,are clearly associated with differences in lizard size. An example of this patternis the positive relationship observed between intestine length and SVL ineach population(T enoBajo: r 0:42, P 0:015,Bajamar: r 0:32, P 0:24, Izaña: s D D s D D r 0:79, P < 0:001,and T etir: r 0:41, P 0:078/. s D s D D Bothspecies have heterodont dentition and cylindrical teeth attached to the jaw in the pleurodontfashion. Gallotiaatlantica showstwo cusps per tooth (back cusp bigger in size) occupyingpractically all the dentary except 4-6 unicusp teeth located in theanterior part. 338 AlfredoV alido,Manuel Nogales

Table 2. Descriptivestatistics ofmorphologicaland anatomical traits of G. galloti inTeno Bajo, Bajamar and Izaña (Tenerife),and G.atlantica inTetir (Fuerteventura). Number of teeth and cusps have been counted in one jaw perindividual. Data are mean 1 s (sample size inbrackets). § Gallotiagalloti Gallotiaatlantica Teno Bajo Bajamar Izaña Tetir

SVL (mm) 119:0 13:4 (119) 95:7 7:1 (15) 103:0 13:1 (33) 60:8 9:9 (65) § § § § Intestinelength (mm) 190 :6 39:2 (32) 130:9 24:4 (16) 164:1 34:2 (15) 85:3 14:7 (19) § § § § Stomachmass (g) 0:3 0:2 (33) 0:3 0:1 (16) 0:3 0:1 (15) 0:1 0:1 (19) § § § § Livermass (g) 0:9 0:3 (33) 0:6 0:2 (16) 1:2 0:5 (15) 0:2 0:2 (19) § § § § Numberof teeth 18:4 2:3 (18) 18:2 1:7 (16) 18:8 0:9 (13) 17:4 0:9 (18) § § § § Numberof cusps 44:3 3:7 (18) 44:1 2:1 (16) 44:5 1:8 (13) 29:8 1:8 (18) § § § §

Gallotiagalloti alsopresents three cusps per tooth distributed in nearly all the dentary except4-5 unicusp teeth in theanterior part. The central cusp of eachtooth is clearly more developedthan the lateral cusps, and this pattern is verysimilar among the three studied lizardpopulations from Tenerife.

PhysiologicalT raits. GPT was shorterin the T etirindividuals (2 :4 1:5days;Range § 0.6-5.9days) than both those from Izaña (6 :1 5:3days;Range 0.7-25.7days) and D § D Teno Bajo (6:9 3:8days;Range 2.3-19.0days) under a similardiet in captivity. This § D isanexpectedresult because G.atlantica hasa muchshorter digestive tract or SVL than G. galloti (see table2). However, this apparent difference between species is not conŽ rmed whenthe difference in size of the lizards is controlled by residual analysis (ANOV A, F 0:124, df 2, 59, P 0:88/ inthis sense, obviously larger sizes generate higher D D D I GPT values.Furthermore, no signiŽcant differences between the two populations studied inTenerife (Teno Bajo and Izañ a) were observed(Scheffé test, P > 0:05/, although a signiŽcant difference in thediet was detectedin the wild. Wecalculatedthe food passage rate (hereafter FPR) bytheratio: SVL/ GPT. Differences inFPR were onlymarginally signiŽ cant, being slower in the lizards from TenoBajo (20:93 10:33mm SVL/day)relative to Tetir (36 :7 25:9mm SVL/day)(ANOV A, § § F 3:24, df 2, 58, P 0:046)(Ž g. 2). Lizards from Izaña showedintermediate D D D values (23:4 13:3mm/day)and did not differ from theother two localities (Scheffé test, § P > 0:05/.Furthermore,a highindividual variation was observedin thethree populations ranges:5.2-44.3 mm/ day(TenoBajo); 3.8-56.1 mm/ day(Izañ a); 10.1-87.3 mm/ day(T etir). Inthe case of Teno Bajo lizards, GPT obtainedunder laboratory conditions (6 :9 § 3:8 days, n 21/ was similarto Ž eldtrials (7 :2 2:6 days, n 13)(Mann-Whitney D § D test, Z 0:87, P 0:38/ thispattern is also observed in the case of theFPR ( Z 1:33, D D I D P 0:18/.Therefore,these results conŽ rm thatdata from thelaboratory experiments D agreewith those obtained under natural conditions. Omnivorouslizards 339

Figure 2. Foodpassage rate (FPR)(mm SVL/day)found in the lizards fromthe three localitiesstudied: Teno Bajoand Izañ a (Tenerife) andTetir (Fuerteventura). The box indicates the mean ( 1SE).The bars extending § aboveand below each boxshow the maximum andminimum values. Sample sizes are alsoindicated.

Discussion

DietComposition and V ariation. Analysesof faecal pellets from thefour populations studiedindicate that these lizards showed a remarkableomnivorous diet. However, an importantvariation in diet composition at both intra and inter-speciŽ c levelswas observed. Lizardsfrom TenoBajo and Bajamar (Tenerife) are omnivorous, basically frugivorous, whileindividuals from Izaña (Tenerife)and T etir(Fuerteventura) are omnivorous, basically insectivorous.These results are also in agreement with previous studies, which show an omnivorousdiet in these species (Lö nnberg, 1902; Klemmer, 1976; Dí az, 1980; V alido andNogales, 1994; Ló pez-Jurado and Mateo, 1995; Nogales et al.,1998). Althoughwe lackestimates of resource abundance in these localities, the observed differencesin the diet could be attributable to abundance variations in food item, both betweenislands and among habitats. In this respect, some authors such as Pé rez-Mellado andCorti (1993), V anDamme (1999),and Cooper and Vitt (2002) have mentioned that thelimitation of animalprey availability in insularenvironments, which together with the densitycompensation suffered by theisland lizards (Olesen and V alido,2003), has been oneof themain factors that might explain the higher occurrence of plantmatter in the diet ofinsularlizards with respect to their mainland relatives. Althoughthis pattern has been used to compare mainland and island habitats, this tendencycould also explain differences among islands within an archipelago (Schluter, 1984),and also between habitats within islands, but this aspect is still rather unknown. Additionally,lizard size (lesser in G.atlantica from Fuerteventura)can also contribute to increaseddifferences in thedegree of herbivorybetween the two species.

Morphologicaland Anatomical T raits. Thedigestive system in these lizards shows dif- ferencesthat are correlated with the amountof theenergeticallydiluted food components in 340 AlfredoV alido,Manuel Nogales thediet(Lö nnberg, 1902; Skoczylas, 1978; King, 1996). This fact conŽ rms thepredictions raisedin this study based on optimal digestion theory of energetically poor diets (Sibly, 1981).The lizard species with a moreenergy-poor diet ( G. galloti)islarger in size and thispermits them to have(secondarily) more effective anatomical and physiological traits tomaintain a higherdigestive efŽ ciency when feeding on vegetablematter .Extensiveuse ofplant matter, as a mainfood resource, is constrainedby bodysize (allometric constraint hypothesis).Smaller lizards have higher energy requirements relative to bodysize (Pough, 1973;Troyer, 1984; Mautz and Nagy, 1987; Zimmerman and Tracy, 1989; Zari, 1991) and for thisreason, the majority of ‘small’species show a markedpreference for adietbased oninvertebrates (Ostrom, 1963;Sokol, 1967; Pough, 1973; Greene, 1982; Demment and VanSoest,1985; Bozinovic, 1993; King, 1996; V anDamme, 1999).Furthermore, Ostrom (1963)and Sokol (1967) indicated that smaller lizards do nothave sufŽ cient jaw power to crushplant matter efŽ ciently; therefore, only large lizards can successfully exploit it. How- ever,some authors such as Johnson and Lillywhite (1979) and King (1996) have criticized thisinterpretation. Accordingto Pough’ s modelon the relationship between lizard size and diet, it isinteresting to note that G. galloti from TenoBajo and Bajamar (the species does notnormally exceed 55 g) shows a markedtendency forward herbivoryeven with a smallersize than other lizards belonging to thefamilies Agamidae, Iguanidae, Scincidae, Gerrhosauridaeand Cordylidae (Pough, 1973). However more extreme deviations from thismodel are known, and expected, under particular ecological situations, such as low predationrisk, behavioral and/ orphysiological mechanisms, resource availability and thermoregulatorybehavior (Pough, 1973; Throckmorton, 1973; Johnson and Lillywhite, 1979;Greene, 1982; V anDevender,1982; Schluter, 1984; W aldschmidtet al., 1986; Mautz andNagy, 1987; Zimmerman and Tracy, 1989; V anMarken Lichtenbelt, 1992; Bozinovic, 1993;Dearing, 1993; V anDamme, 1999). Furthermore, it is interestingto note that those lizardspecies in whichpoor-Ž ber plantparts (fruits and owers) constitutethe main part of theplant component do notshow speciŽ c anatomicaltraits to digestcellulose as occursin folivorouslizard species (Iverson, 1982; Cooper and Vitt, 2002). This could explain why herbivoryis relatively more common among small lizards (see reviewsin Greene, 1982; VanDamme, 1999 and Cooper and V itt,2002) than expected under Pough’ s model.In thisrespect, the size of organswill be largeraccording to ageneralincrease in body size (allometricrelated; see e.g. Carretero 1993 for somelacertid species). For example,with respectto the intestine length, an increase of the absorption surface would be important for lizardswith qualitatively poor diets due to the increase in digestion rate and nutrient absorptionefŽ ciency. Thenumber of cusps per tooth has been suggested to be adaptively variable among lizardsin relation to herbivory (Hotton, 1955; Montanucci, 1968; Johnson and Lillywhite, 1979;Zimmerman and Tracy, 1989; Mateo and Ló pez-Jurado, 1992) because the cusps havepotentially a shearingfunction and could be usedfor reducingplant materials (Sokol, 1967).Nevertheless, the presence of teeth without cusps and the absence of a clear Omnivorouslizards 341 masticationhave been pointed out ascentraltraits of typicallyherbivorous species (Ostrom, 1963;Throckmorton, 1973 and 1976). Besides,this degree of shearingof plantmatter would also be favoredby theingestion ofsmall stones (lithophagy) (Sokol, 1971; Skoczylas, 1978). This tendency has also been observedin our studied organisms where a highernumber of stones (grit) was foundin TenoBajo and Bajamar than Izañ a andT etir.This comparative feeding behavior would bea goodsolution that does not require any remarkable change in digestiveanatomy, and suitablefor allin habitats where the resource availability and abundance is unpredictable. PhysiologicalTraits. Physiologicalvariables measured in the Canarian lizard populations followthe general patterns previously reported by several authors for differentlizard speciesalong an insectivory-herbivory gradient. Long transit times (about three to seven days)have been mentioned for differentherbivorous lizard species (Throckmorton, 1973; Johnsonand Lillywhite, 1979; Christian et al., 1984; W aldschmidtet al., 1986; Mautz and Nagy,1987; Zimmerman and Tracy, 1989), while values shorter values than two days have beenobserved among insectivorous ones (T addei,1951; Avery,1973; Windeland Sarokon, 1976;Zimmerman and Tracy, 1989). Theselong retention times of theplantfood in the gut arenecessary to increasedigestive efŽciency (Throckmorton, 1973; Johnson and Lillywhite, 1979; Iverson, 1982; Christian etal., 1984; Troyer, 1984; Zimmerman and Tracy, 1989). The absence of differences betweenthe lizards from TenoBajo and Izañ a, duringthe GPT experiments,could be a consequenceof their maintenance under a similardiet. Therefore, it might be possible thatunder natural conditions the Izañ a lizards(with a basicallyinsectivorous diet) will presenta shorterGPT, inaccordance with the predictions of Sibly (1981). In this respect, adifferentthermoregulatory behavior can produce a shorterGPT (Throckmorton,1973; Waldschmidtet al., 1986; Mautz and Nagy, 1987; Zimmerman and Tracy, 1989; V an MarkenLichtenbelt, 1992), and this functional change could be more advantageous ecologicallyand evolutionarily, than anatomical modiŽ cations. Conclusions. Nospecial anatomical or physiological adaptations were observedin the omnivorousCanarian lizards studied. Body size was themajor determinant explaining the levelof herbivory(or frugivory)in these omnivorous lizards; in the latter case, large body sizefacilitates the exploitation of fruitsand the lizards function importantly as mutualistic agentsin theseed-dispersal of plants(V alido,1999). Onthe other hand, herbivory in island lizards could be related to a lowerdensity of arthropodsin oceanicislands (Janzen, 1973; V anDamme, 1999;Cooper and Vitt, 2002). Thiscould explain the tendency of Izaña lizardsto show an moreinsectivorous diet with respectto theBajamar and Tenopopulations,probably due toa higherarthropod abundance inthis locality when this study was performed(V alido,1999). Besides, differences in theamount of arthropods in the lizard diet, and their relationship with the abundance ofinsects, has been observed in some localities in the high mountain shrub of Tenerife wherea systematicsampling of arthropodswas realized(Oromí et al.,1996; V alido,1999). 342 AlfredoV alido,Manuel Nogales

Inadition, the Canarian lizards show preferences for arthropodswhen both food items (insectsand plants) are available (A. Valido,personal observation with captive lizads; see alsoSchuter, 1984 in the case of theGalá pagos lava lizard). Anotherargument that could explain the absence of anatomical or physiological adaptationsto the herbivorous diet is that these lizards show an important temporal and spatialvariation of plant material in the diet during the year .Inthis respect, it was observedthat the vegetal component ranged between 23.1% (January) - 79.3%(April) in the case of G. galloti inTenoBajo and it isanormalresponse in differentCanarian lizards inhabitingseveral habitats through the year (V alido,1999). This seasonal diet change could belimiting the evolution of permanent anatomical adaptations in these lizards (as has beenobserved in frugivorous birds; Herrera, 1984;Jordano, 1987), and other behavioral changeslike thermoregulatory behavior, ingesting small stones which are less costly and moreecological and evolutionarily advantageous for generalistlizards.

Acknowledgements. We thankLuisa Herná ndez and Manuel V alido‘ Pololo’for assisting us duringthe follow upofthe GPT experiments. Fé lix Medina ‘ Feluco’and Juan Carlos Sá nchez capturedsome lizards andprovided uswithpreserved specimens, respectively.Jaime Uriostegave us insects duringthe GPT experimental period, and theServicio Electró nico de la Universidadde La Lagunaassembled thecontrol system oftheclimatic chamber. PedroJordano helped us to plan the experiments and encouraged in several ways at allstages ofdevelopment of thisstudy. Mario Dí az andspecially Pedro Jordano, Francisco Bozinovic, Donald B. Miles,Raoul V anDamme andtwo anonymous referees improvedan initial manuscript with useful comments. Thecontrol system usedin theGPT experiments was supportedby Direcció n General deUniversidades e Investigación delGobierno de Canarias (grant93/ 150).

References

Avery,R.A. (1973): Morphometric and functional studies on the stomach ofthe lizard Lacertavivipara . J. Zool. 169: 157-167. Bozinovic,F. (1993): Fisiologí a ecológica dela alimentación ydigestión envertebrados: modelos y teorías. Rev. Chil.Hist. Nat. 66: 375-382. Carretero, M.A.(1997): Digestive size anddiet in Lacertidae: apreliminaryanalysis. In: Herpetologia Bonnensis, p.43-49. Bö hme, W .,Bischoff,W .,Ziegler,T., Eds, Bonn, S.E.H. Chivers,D.J., Hladik, C.M. (1980): Morphology of thegastrointestinal tract inprimates: comparisonswith other mammals inrelation to diet. J. Morphol. 166: 337-386. Christian,K.A., Tracy, A.C., Porter, W .P.(1984):Diet, digestion, and food preferences ofGalá pagos land iguanas. Herpetologica 40: 205-212. Cooper,W .E.,Vitt,L.V .(2002):Distribution, extent, and evolution of plant consumption by lizards. J. Zool., Lond. 257: 487-517. Dearing,M.D. (1993): An alimentary specialization for herbivory in the Tropical Whiptail lizard Cnemidophorus murinus.J.Herpetol. 27: 111-114. Demment, M.W.,VanSoest, P .J.(1985):A nutritionalexplanation for body-size patterns of ruminant and nonruminantherbivores. Am. Nat. 125: 641-672. Díaz, E.(1980):Notas sobreel lagartotizó n ( Lacertagalloti ).Doñana, Acta Vert. 7: 95-96. Díaz, J.A.(1994): Field thermoregulatory behavior in the western Canarianlizard Gallotiagalloti .J.Herpetol. 28: 325-333. Fu,J. (1998):Toward the phylogeny of the family Lacertidae: implicationsfrom mitochondrial DNA 12sand 16s genesequences (Reptilia:Squamata). Mol. Phyl. Evol. 9: 118-130. Omnivorouslizards 343

González, P.,Pinto,F., Nogales, M., Jiménez-Asensio, J., Herná ndez, M., Cabrera, V.(1996):Phylogenetic relationshipsof the Canary Islands endemic lizardgenus Gallotia (Sauria:Lacertidae), inferredfrom mitochondrialDNA sequences.Mol. Phyl. Evol. 5: 63-71. Greene, H.W.(1982): Dietary andphenotypic diversity in lizards: why are some organismsspecialized? In: Environmentaladaptation and evolution, p. 107-128. Mossakowski, D., Roth, G., Eds, Stuttgart, Gustav Fischer. Greer, A.E.(1976): On the evolution of the giant Cape Verde scincidlizard Macroscincuscoctei .J.Nat. Hist. 10: 691-712. Guard,C.L. (1980): The reptilian digestive system: generalcharacteristics. In:Comparative Physiology: Primitive Mammals, p.43-51. Schmidt-Nielsen, K., Bolis, L., Taylor,C.R., Eds, Cambridge, Cambridge University Press. Herrera, C.M.(1984): Adaptation to frugivory of Mediterranean avian seed dispersers.Ecology 65: 609-617. Hotton,N. (1955):A surveyof adaptiverelationships of dentitionto diet in North America Iguanidae.Am. Midl. Nat. 55: 88-114. Iverson,J.B. (1982): Adaptations to herbivory in iguanine lizards. In: Iguanas of the World: their Behavior, Ecologyand Conservation, p. 60-76.Burghardt, G.M., Rand, A.S., Eds, New Jersey, Noyes,Park Ridge. Janzen,D.H. (1973): Sweep samples oftropicalfoliage insects: effects ofseasons, vegetation types, elevation, time ofday, and insularity. Ecology 54: 687-701. Johnson,R.N., Lillywhite, H.B. (1979): Digestive efŽ ciency of the omnivorous lizard Klauberinareversiana . Copeia 1979: 431-437. Jordano,P .(1987):Frugivory, external morphology and digestive system inmediterranean sylviidwarblers Sylvia spp. Ibis 129: 175-189. Jordano,P .(2000):Fruits and frugivory. In: Seeds, the Ecology of Regenerationin Plant Communities, p. 125- 165.Fenner, M., Ed.,Wallingford, Commonwealth Agricultural Bureau International. Karasov,W .H.(1986):Energetics, physiology and vertebrate ecology.Trends Ecol. Evol. 1: 101-104. King,G. (1996):Reptiles and Herbivory. London, Chapman and Hall. Klemmer, K.(1976):The amphibia and reptilia of the Canary Islands. In: Biogeography and Ecology in the CanaryIslands. Monographia Biologica 30, p. 433-456. Kunkel, G., Ed.,The Netherlands, The Hague, W . JunkPublishers. Lönnberg, E. (1902):On some pointsof relation between themorphological structure of the intestine and the diet ofreptiles. Bihang Till K. SvenskaV et.-Akad.Handlingar 28: 1-51. López-Jurado, L.F., Mateo, J.A. (1995): Origin, colonization, adaptive radiation, intrainsular evolution and species substitutionprocesses inthe fossil and living lizards oftheCanary Islands. In: Scienta Herpetologica, p.81-91. Llorente, G.A., Montori, A., Santos, X., Carretero, M.A.,Eds, Barcelona, S.E.H. Márquez, R., Cejudo,D., Pérez-Mellado, V .(1997):Selected bodytemperatures offour lacertid lizards fromthe CanaryIslands. Herpetol. J. 7: 122-124. Mateo,J.A., Ló pez-Jurado, L.F. (1992): Study of dentitionin lizards fromGran Canaria Island(Canary Islands) andits ecological and evolutionary signiŽ cance. Biol.J. Linn.Soc. 46: 39-48. Mautz,W.J., Nagy, K.G. (1987): Ontogenetic change in diet, Ž eldmetabolic rate, andwater uxinthe herbivorous lizard Dipsosaurusdorsalis .Physiol.Zool. 60: 640-658. Montanucci,R.R. (1968): Comparative dentition in four iguanid lizards. Herpetologica 24: 305-315. Nogales,M., Delgado,J.D., Medina, F.M. (1998): Shrikes, lizards and Lyciumintricatum (Solanaceae) fruits:a case ofindirect seed dispersalon anoceanic island(Alegranza, CanaryIslands). J. Ecol. 86: 866-871. Olesen, J.M.,V alido,A. (2003):Lizards as pollinatorsand seed dispersers:A islandphenomenon. Trends Ecol. Evol. 18 (4):177-181. Oromí, P.,Zurita,N., Arechavaleta, M.,Camacho, A.(1996):Inventario de la faunade invertebradosdel Parque Nacionaldel Teide. II Estudiode las comunidadesde invertebrados. Depto. de Biología Animal.Universidad deLa Laguna(unpublished report), pp. 370. Ostrom,J.H. (1963): Further comments onherbivorouslizards. Evolution 17: 368-369. Packard,G.C., Boardman, T.J. (1988): The misuse ofratios, indices, and percentages inecophysiological research. Physiol.Zool. 61: 1-9. Pérez-Mellado, V .,Corti,C. (1993):Dietary adaptationsand herbivory in lacertid lizards ofthe genus Podarcis fromwestern Mediterraneanislands (Reptilia: Sauria). Bonn. zool. Beitr. 44: 193-220. 344 AlfredoV alido,Manuel Nogales

Pérez-Mellado, V .,Romero-Beviá, M.,Ortega,F., Martí n-Garcí a, S.,Perera, A.,Ló pez-Vicente, M.,Galache, C. (1999):El usode losrecursos tróŽ cos en Gallotiasimonyi (SauriaLacertidae) dela isla deElHierro(Islas Canarias). In:El Lagarto Gigante de ElHierro. Bases para suConservació n, p. 63-84.Ló pez-Jurado, L.F., Mateo,J.A., Eds, Asociació n Herpetológica Española. Pough,H. (1973):Lizard energetic anddiet. Ecology 54: 837-844. Schluter,D. (1984):Body size, preysize andherbivory in the Galá pagos lava lizard Tropidurus. Oikos 43: 291- 300. Sibly,R.M. (1981): Strategies of digestionand defecation. In: Physiological Ecology, an Evolutionary Approach toResource Use, p.109-139. Townsend, C.R., Calow, P .,Eds,London, Blackwell ScientiŽc Publications. Skoczylas,R. (1978):Physiology of the digestive tract. In:Biology of the Reptilia, vol. 8, p. 589-717. Gans, C., Tinkle,D.W., Eds, London, Academic Press. Sokol,O.M. (1967): Herbivory in lizards. Evolution 21: 192-194. Sokol,O.M. (1971): Lithophagy and geophagy in reptiles. J. Herpetol. 5: 69-70. Szarski,H. (1962):Some remarks onherbivorouslizards. Evolution 16: 529. Taddei,A. (1951):Contibuto allo studio del canale alimentaire in Lacerta.Boll.Zool. 18: 291-294. Throckmorton,G.S. (1973): Digestive efŽ ciency in the herbivorous lizard Ctenosaurapectinata . Copeia 1973: 431-435. Throckmorton,G.S. (1976): Oral foodprocessingin two herbiv orouslizards, Iguanaiguana (Iguanidae)and Uromastixaegyptius (Agamidae). J.Morphol. 148: 363-390. Troyer,K. (1984):Diet selectionand digestion in Iguanaiguana :theimportance of age andnutrient requirements. Oecologia 61: 201-207. Valido,A. (1999):Ecologí a dela dispersión desemillas porlos lagartos endé micos canarios (g. Gallotia, Lacertidae). PhDThesis, University of La Laguna,Tenerife. Valido,A., Nogales, M. (1994):Frugivory and seed dispersalby the lizard Gallotiagalloti (Lacertidae) ina xeric habitatof the Canary Islands. Oikos 70: 403-411. VanDamme, R.(1999):Evolution of herbivory in lacertid lizards:effects ofinsularity and body size. J.Herpetol. 33: 663-674. VanDevender, R.W. (1982): Growth and Ecology of Spiny-Tailed and Green Iguanasin Costa Rica, with comments onthe evolution of herbivoryand large bodysize. In:Iguanas of the World, pp. 162-183. Burgharat, G.M.,Rand, A.S., Eds, New Jersey, Noyespubl. Park Ridge. VanMarken Lichtenbelt, W. (1992): Digestion in an ectothermic herbivore,the green Iguana ( Iguanaiguana ): effect offoodcomposition and body temperature. Physiol.Zool. 65: 649-673. Waldschmidt,S.R., Jones, S.M., Porter, W.P .(1986):The effect ofbody temperature andfeeding regime on activity,passage time anddigestive coefŽ cient in the lizard Utastansburiana .Physiol.Zool. 59: 376-383. Windell,J.T., Sarokon, J.A. (1976): Rate ofgastric evacuationin the lizard Anoliscarolinensis (Sauria: Iguanidae).Herpetologica 32: 18-23. Zari,T.A. (1991): The in uence ofbodymass andtemperature onthe standard metabolic rate oftheherbivorous desert lizard, Uromastyx microlepis .J.Therm.Biol. 16: 129-133. Zimmerman, L.C.,Tracy, C.R. (1989): Interactions between environmentsand ectothermic andherbivory in reptiles.Physiol. Zool. 62: 374-409. Ziswiler,V .,Farner,D.S. (1972): Digestion and the digestive system. In:Avian Biology V ol.II., p. 343-430. Farner,D.S., King, J.R., Parkes, K.X., Eds, New York,Academic Press. Zug,G.R. (1993): Herpetology, an Introductory Biology of Amphibians and Reptiles. Academic Press, SanDiego.

Received: August31, 2002. Accepted: January20, 2003.