A morphological and life history comparison between desert populations of a sit-and-pursue antlion, in reference to a co-occurring pit-building antlion

A morphological and life history comparison between desert populations of a sit-and-pursue antlion, in reference to a co-occurring pit-building antlion
of 10
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  ORIGINAL PAPER  A morphological and life history comparison betweendesert populations of a sit-and-pursue antlion, in referenceto a co-occurring pit-building antlion Inon Scharf   &  Ido Filin  &  Aziz Subach  &  Ofer Ovadia Received: 7 April 2009 /Revised: 24 May 2009 /Accepted: 28 May 2009 /Published online: 27 June 2009 # Springer-Verlag 2009 Abstract  Although most antlion species do not construct  pits, the vast majority of studies on antlions focused on pit- building species. We report here on a transplant experiment aiming to test for morphological and life history differences between two desert populations of a sit-and-pursue antlionspecies,  Lopezus fedtschenkoi  (Neuroptera: Myrmeleonti-dae), srcinating from habitats, which mainly differ in plant cover and productivity. We raised the antlion larvae inenvironmental chambers simulating either hyper-arid or Mediterranean climate. We found significant differences inthe morphology and life history of   L. fedtschenkoi  larvae between the two populations. For example, the larvaesrcinating from the more productive habitat pupated faster and had a higher growth rate. In agreement with thetemperature  –  size rule, antlions reached higher final mass inthe colder Mediterranean climate and exhibited a higher growth rate, but there was no difference in their develop-mental time. Observed differences in morphology between populations as well as those triggered by climate growingconditions could be explained by differences in sizeallometry. We also provide a quantitative description of the allometric growth axis, based on 12 morphologicaltraits. Comparing the responses of   L. fedtschenkoi  withthose observed in a co-occurring pit-building antlionindicated that there were neither shape differences that areindependent of size nor was there a difference in the plasticity level between the two species. Keywords  Foragingmode.Myrmeleontidae.Reactionnorm.Temperature.Transplantexperiment  Introduction Species with a broad geographical range may be subjectedto a wide spectrum of different environmental conditions.When the strength and/or direction of selection within sucha species range are not uniform, adaptations to localconditions can emerge (Blanckenhorn 1997; Reznick andTravis 2001). Local adaptations, however, are not the singlesolution for coping with environmental variability. Whenenvironmental conditions vary over time in a predictablemanner, phenotypic plasticity (i.e., the environmentallysensitive production of more than one phenotype by thesimilar genotypes) should be preferred (Nylin and Gotthard1998; David et al. 2004). Clearly, what is usually found in natural systems is a mixture of local adaptations selectingfor a local optima of different traits, combined with a population-specific range of phenotypes developed inresponse to different environmental conditions (e.g.,Reznick and Travis 2001; Volis et al. 2002). Occasionally, the strength of phenotypic plasticity also differs across populations. Some populations are extremely plastic, I. Scharf ( * ) : I. Filin : A. Subach : O. OvadiaDepartment of Life Sciences,Ben-Gurion University of the Negev,PO Box 653, Beer-Sheva 84105, Israele-mail: Filine-mail: ido.filin@helsinki.fiA. Subache-mail: Ovadiae-mail: FilinDepartment of Mathematics and Statistics, University of Helsinki,PO Box PL 68, 00014 Helsinki, Finland Naturwissenschaften (2009) 96:1147  –  1156DOI 10.1007/s00114-009-0576-z  greatly responding to a change in environmental conditions,whereas others exhibit a constant response. Such agenotype×environment interaction implies that phenotypic plasticity is a trait by itself, which can be susceptible tonatural selection (e.g., Piglliucci 2005).Climate is one of the most important environmentalfactors potentially affecting all traits characterizing anectotherm, including growth, development, activity, lon-gevity, and starvation endurance (e.g., Speight et al. 1999).In general, ectotherms originating from warmer environ-ments or experiencing warmer growing conditions developfaster but reach a smaller final body size (e.g., Atkinson1994; David et al. 2004; Kingsolver and Huey 2008). The causes for this  “ temperature-size rule ”  in ectotherms(Bergmann ’ s rule and neo-Bergmannian rule for endo-therms; James 1970) are not yet clear, and both adaptiveand non-adaptive mechanisms have been suggested for explaining this pattern (Blanckenhorn and Demont  2004;Kingsolver and Huey 2008). Supporting an adaptivemechanism requires a presentation of some trade-off amongfitness, temperature, and body size of ectotherms. Indeed,Reeve et al. (2000) have shown that the advantage of large over small males is much more considerable under low thanhigh temperatures. On the other hand, a common non-adaptive explanation is related to the smaller expected cellsize of organisms under warmer conditions: Cells should bereduced in size because oxygen consumption dependstightly on temperature, while oxygen diffusion depends ontemperature much less (Blanckenhorn and Demont  2004). Recently, He (2007) has shown that, by combiningallometry with  E  -infinity theory, one can arrive at a simplescaling law, explaining the observed negative relationship between temperature, cell size, and consequently body size.Alternatively, some studies have suggested that precipitation,moisture, and habitat productivity can better explaindifferences in body size than temperature (James 1970;Yom-Tov and Geffen 2006). These environmental variablesare often strongly intercorrelated, especially in desert environments.The effects of temperature or some other biotic andabiotic factors on body size are especially interesting because size is tightly correlated with fitness components,such as fecundity, survival, and mating success. Body sizeis therefore considered as one of the most important organismal traits (e.g., Honek  1993; Blanckenhorn andDemont  2004; Kingsolver and Pfennig 2004; Berger et al. 2006). Life history studies dealing with thermal adaptationsnot only focus on body size per se; rather, they also relate tohow developmental time and growth rate interact toinfluence adult size (Nylin and Gotthard 1998; Roff  2001; Berger et al. 2006).We report here on a transplant experiment whose aimwas to test for differences in morphology and life history(i.e., body size, growth rate, and developmental time) between two desert populations of the sit-and-pursueantlion species. These two populations originate fromsandy habitats, varying mainly in plant cover and produc-tivity. We also aimed at comparing the morphology and lifehistory of this sit-and-pursue antlion with a co-occurring pit-building antlion, exhibiting a more sedentary foragingmode. Finally, we discuss how these differences in foragingmode should be translated into morphological and lifehistory divergence. Generally speaking, comparative studiesof related populations or species serve as an important first step towards understanding patterns of local adaptations.This is because they link environmental conditions (e.g.,temperature) with different organismal traits (e.g., foragingmode and body size; Reznick and Travis 2001).The foraging behavior of antlion species can be dividedinto two characterizing groups and is very distinctive: Onone hand, some species construct conical pits, serving astraps for small arthropod prey. These pit-building antlionsare sedentary and rarely relocate (Scharf and Ovadia 2006). On the other hand, some species do not construct pits but simply burry themselves beneath the sand and wait for preyto enter their detection range. These non-pit-buildingantlions can be considered as sit-and-pursue predators sincethey relocate much more often and up to several meters per night of activity (Cain 1987). Previous studies have alreadydemonstrated differences in morphology and life history between populations of pit-building antlion species alonggeographical gradients (Arnett and Gotelli 1999; Scharf et al. 2008; Scharf et al. 2009) while following the “ temperature-size rule ”  and showing impressive levels of  phenotypic plasticity. The present study, however, is thefirst experimental quantification of such patterns in a non- pit-building antlion species. Another suggested difference between sessile and mobile organisms is the much moreconstrained ability of the former to buffer environmentalconditions (e.g., Huey et al. 2002). It is thus possible that  some of the phenotypic plasticity characterizing pit- building antlions can be attributed to their sedentary nature.Based on this logic, we may expect the phenotypicresponses of the pit-building antlions to environmentalconditions to be more prominent than those of the moremobile sit-and-pursue antlions.The sit-and-pursue antlion studied here,  Lopezus fedtschenkoi  (Neuroptera: Myrmeleontidae), occurs insandy Mediterranean and desert habitats in Israel. It oftenco-occurs with the pit-building species of   Myrmeleonhylainus , which belongs to the same family, and serveshere as a reference for comparisons. We predicted that   L. fedtschenkoi  should follow the temperature  –  size rule, i.e.,individuals should be smaller when grown under warmtemperatures or when srcinating from a warmer environ-ment. We also predicted that growth rate would be slower  1148 Naturwissenschaften (2009) 96:1147  –  1156  in the less productive habitat, since animals in poor habitatsoften have a slow fixed growth rate, even when conditionsare improved (Arendt  1997). Owing to the more mobile or  active nature of the sit-and-pursue antlion  L. fedtschenkoi ,we expect it to show a weaker phenotypic flexibility thanthe more sedentary pit-building antlion  M. hyalinus  becausethe former can compensate for environmental variation bymovement. Finally, morphological comparisons of widelyforaging predators with their sit-and-wait relatives haveshown that the former are more  “ streamlined ”  or   “ elongated, ” while the latter tend to be more  “ stocky ”  or   “ heavy bodied ” (Vitt and Congdon 1978; Huey and Pianka 1981). This suggests that the elongated shape improves movement. Wehypothesized that   L. fedtschenkoi , regardless of the habitat of srcin, should have a more elongated body shape than  M.hyalinus . Materials and methods  Natural history and study sites  Lopezus fedtschenkoi  is a nocturnal sit-and-pursue antlionspecies, which does not construct a pit. It occurs in sandyhabitats in desert and Mediterranean regions in Israel and isstrongly restricted to shifting sand dunes, avoiding theshaded areas under trees or bushes (Simon 1988).  L. fedtschenkoi  develops through three instar stages (the third being the longest). Adults emerge between early April andearly August, and compared to the long lifespan of thelarvae, the adults are much short lived. Adults mate at night, and the females oviposit eggs directly in sand. Antscompose about 70% of its diet. When ambushing prey, itsentire body, except for the mandibles and eyes, is coveredwith sand.  L. fedtschenkoi  moves several times each night while being active. However, its effective detection range at each ambush location is considerably smaller than that of  pit-building antlions (Cain 1987). The larvae of   L. fedtschenkoi , similarly to other antlion species, prey onother small arthropods, mainly ants. They penetrate the prey with their mandibles and inject toxins into the prey body. After the prey is dead, enzymes are injected, and the body contents are digested and extracted (Griffiths 1980).The maximal size of potential prey increases with the instar stage (Griffiths 1980; Scharf and Ovadia 2006). Similarly to pit-building antlions,  L. fedtschenkoi  discards the preyafter consumption, but thereafter, the antlion would almost always change its ambush site (Simon 1988).  Lopezus fedtschenkoi  larvae were collected in twodistinct sandy sites in the Negev desert of Israel: NahalSecher and Holot Agur. Both sites are part of the extensionof the sand belt of northern Sinai. The sandy landscape ineach site can be divided into three major habitat types:shifting (where  L. fedtschenkoi  mainly occurs), semi-stabilized, and stabilized sand dunes. In spite of the short geographical distance between the two sites (~38 km), theydiffer in their annual rainfall, temperature, and vegetationcover (Abramsky et al. 1985; Scharf et al. 2009; Table 1).  Nahal Secher receives more precipitation and has a higher annual plant cover than Holot Agur. Temperature differ-ences between these two sites are less clear: Nahal Secher has colder winters but slightly warmer summers than Holot Agur and, therefore, has a broader seasonal temperaturerange (Table 1). Additionally, there is a large concentrationof planted tamarisk trees in Nahal Secher, which useunexploited underground water. These trees and the denser vegetation cover characterizing Nahal Secher probablyaffect the antlions in two different ways. First, they increasethe productivity of this sandy landscape; more seeds andgreen material in general are present there, and consequent-ly, more insect prey, such as seed-eating ants, is availablefor antlions. Second, trees and denser vegetation may also provide more stable microclimatic conditions owing to the broader shaded areas they create. Therefore, Nahal Secher  probably provides better conditions for antlions than Holot Agur. Antlions in the former habitat should be larger anddevelop faster owing to the increased prey abundance andimproved microclimatological conditions.Experimental designDuring June and July 2006 we collected a total number of 75  L. fedtschenkoi  larvae from the two sandy sitesdescribed above and brought them to the laboratory.Individual larvae were assigned randomly to either aMediterranean or a desert environmental chamber main-tained under a 24-h cycle of temperature, light, andhumidity. The two environmental chambers simulatedthe summer ambient conditions of Tel-Aviv or Eilat,respectively (Tel-Aviv, Mediterranean climate, day/night  Table 1  Climatic conditions and annual and perennial plant cover characterizing the two study sites, Nahal Secher and Holot Agur  Nahal Secher Holot Agur Latitude 31.1033 31.0023Longitude 34.8111 34.4241Altitude (m) 326.9 216.8Mean Jan. temp. (ºC) 10.8 11.2Mean Aug. temp. (ºC) 25.9 24.7Seasonal temp. range (ºC) 15.1 13.5Annual rainfall (mm) 140.2 103.7Annual cover (%) 30.3 10.5Perennial cover (%) 9.2 2.8 Naturwissenschaften (2009) 96:1147  –  1156 1149  temperatures of 29.1°C/20.7°C and humidity of 60%/82%;Eilat, desert climate with day/night temperature of 39.0°C/ 25.4°C and humidity of 15%/48%; day/night length 12:12 hin both chambers). Both populations srcinate from a desert region; however, larvae were raised also under Mediterraneanconditions. The species range includes also the Mediterraneansandy habitats, so it should be insightful to study its responsetoward that climate. Moreover, the Mediterranean climateshould be considered here also as a milder climate incomparison to the desert climate, and it enables us to quantifythe effect of a harsh (desert) climate in comparison to a milder (Mediterranean) one. Antlion larvae were kept separately inroundplastic cups(5-cmdiameter) filledwithsand(2cm)andwere fed with flour beetle larvae twice a week. We used arelatively intensive feeding in order to neutralize possibleeffectsofhunger.Weaimedatsimulatingahabitatrichofprey but which differs in the climatic conditions. Antlions wereweighed once every 2  –  3 weeks. The experiment lasted for ayear, after which ~17% of the larvae pupated and ~59%reached an asymptotic mass without pupating (their mass wasstable for at least a month). We interpreted this to mean that some cue required for pupation is missing but documentedtheir asymptotic-final mass.Larval morphologyLarvae were photographed after spending about 6 weeks inthe environmental chambers. For each of the 61 antlionlarvae photographed, we took 12 different larval bodymeasurements:distancebetweenmandibles(DBM),mandiblelength (ML), curved mandible length (CML), mandible width(MW), head length (HL), head width (HW), distance betweenmandible ’ s first and third tooth (T13), distance betweenmandible ’ s second and third tooth (T23), thorax length (TL),thorax width (TW), abdomen length (AL), and abdomenwidth (AW). All trait values were log-transformed prior tostatistical analyses, as described next.We used multivariate analysis of variance (MANOVA;Zelditch et al. 2004, Chapter 9) to test for differences in morphology between the two populations and the twoclimatic treatments. We used principal component analysis(PCA; Zelditch et al. 2004, Chapter 7) to generate a new composite variable, describing the overall larval size (thefirst principal component; PC1). We then used a three-wayANOVA to test for the effects of instar, population, andclimatic treatment on larval size (the effect of instar wasobvious, but we included it in order to test for possibleinteractions). We additionally used discriminant functionanalysis (Zelditch et al. 2004, Chapter 7) in order to find the axis (i.e., a linear combination of the morphological traits),which best separates instar stages.We compared the morphology of the non-pit-buildingantlion  L. fedtschenkoi  with that of the pit-building antlion  M. hyalinus , originating from the same sites. The  M.hyalinus  data used for this comparison was taken from asimilar study by Scharf et al. (2008). We used PCA once more, on a shared dataset of   L. fedtschenkoi  and  M.hyalinus  second and third instar larvae. PC1 representedsize, as described above. The purpose here, in addition tothe obvious differences in size, was to understand whether the sit-and-wait   M. hyalinus  is more  “ heavy bodied ”  or  “ stocky ”  than the sit-and-pursue  L. fedtschenkoi .Life history analysisFollowing Nylin and Gotthard (1998), we refer to finalmass, growth rate, and development time as the most important life history traits. We consider the last measured body mass prior to pupation or the asymptotic larval massas the  “ final mass. ”  Growth rate was calculated using thecommon formula (e.g., Gotthard et al. 1994):ln mass t  ð Þ ln mass 0 ð Þ t   ð 1 Þ In the analysis of growth rate and change in mass, weremoved two individuals that had only a single bodymass measurement. Pupation rate served to measuredevelopmental time (i.e., they are inversely related). Weused two-way ANOVAs to test for the effect of populationsand climatic conditions on final mass, growth rate, andchange in mass. The comparison of pupation rate between populations and climatic conditions was performed using aCox regression model (i.e., proportional hazards model;Kalbfleisch and Prentice 2002). We used the counting process notation of the proportional hazards model toaccount for body mass of individuals (log-transformed) asa time-varying covariate (Kalbfleisch and Prentice 2002;S-Plus 2000 guide to statistics, vol. 2, Mathsoft, Seattle,WA, USA).Finally, we were interested in differences in the level of  plasticity between  L. fedtschenkoi  and  M. hyalinus . Data for this comparison was taken once more from Scharf et al.(2008). We compared plasticity in the final mass of individuals using a two-way ANOVA. We tested for asignificant species×environmental chamber interaction.The experimental procedure in the present study and Scharf et al. (2008) were almost identical (antlion populationswere sampled and raised under two distinct laboratoryconditions). The same variables were measured (e.g., timeto pupation, mass prior to pupation, and morphologicalcharacters), the climatic growth conditions were identical(two environmental chambers simulating desert and Med-iterranean treatments) and so was the feeding regime (two prey items per week). The only difference is the larger  1150 Naturwissenschaften (2009) 96:1147  –  1156  number of populations in Scharf et al. (2008)  —  five insteadof two in the current experiment. Results Larval morphologyWe performed PCA on the 12 morphological traitsmeasured. PC1 and PC2 accounted for 86.53% and 5.15%of the total variance, respectively. Morphological traits inPC1 all had a positive sign, and therefore, this axis isinterpreted as representing overall body size (Table 2).Since PC1 explained the vast majority of the variance, werefer only to that axis in further analyses. Second and thirdinstar stages are clearly separated along the PC1 axis(Fig. 1a; first instar-stage larvae were not collected).Finally, canonical discriminant function analysis on mor- phology provided an expression for the maximal separationof instars. Based on this expression, the trait best suited todiscriminate between the second and third instar stages ishead width (F-to-remove=10.56), and the second most indicative trait is curved mandible length (F-to-remove=1.95). The canonical variate (standardized by within-groupvariances), best separating between instars was:CV ¼ 0 : 0587  DBM þ 0 : 1131  ML þ 0 : 3527  CML þ 0 : 1407  MW þ 0 : 0386  HL þ 0 : 7319  HW þ 0 : 0351  T13 þ 0 : 0626  T23 þ 0 : 1434  TL þ 0 : 0335  TW þ 0 : 1113  AL þ 0 : 1393  AW :  ð 2 Þ By standardizing the PC1 loadings of all traits, using thehead width as a reference (i.e., dividing the loadings of alltraits by that of head width), we found that all traits, except the thorax length, grow with negative allometry in relationto the head width (i.e., have a loading value of less than 1;Table 2). Abdomen dimensions in particular increased in aslower rate compared to head length and thorax length.Therefore, as the larva progresses from second to thirdinstar, there is an allometric change in shape associatedwith the overall increase in body size. Its thorax becomesrelatively longer, the head becomes proportionally wider  but shorter and the abdomen proportionally smaller (Fig. 1b). In other words, not only size but also shapechanges through development, i.e., growth is allometric.There was a significant difference in body size (i.e.,morphological traits) of second instar   L. fedtschekoi  between the Nahal Secher and Holot Agur populations(MANOVA; Wilk  ’ s  l =0.3334,  P  =0.0347,  df   =12, 16).However, the effect of climatic treatment (i.e., desert vs.Mediterranean growth conditions) was not significant (Wilk  ’ s  l =0.4559,  P  =0.1903,  df   =12, 16). A similar separation between populations, though less significant,was evident among third instar larvae (Wilk  ’ s  l =0.3945,  P  =0.0698,  df   =12, 17), and once more, the effect of climatic treatment was not significant (Wilk  ’ s  l =0.5794,  P  =0.4671,  df   =12, 17). Next, we used another statisticalapproach to test for the effect of population of srcin andclimatic treatment on body size. Using a three-wayANOVA, we tested for those same effects on PC1 scores,which represent body size, while also taking the instar stageinto account (we tested the main effects and two-wayinteraction terms, instar×population and instar×climatechamber). The only significant main effect impacting bodysize (PC1) was the instar stage (note that the instar stagewas determined using PC1). Interestingly, the interactionterm instar×population was significant (  F  1,55 =4.8259,  P  =0.0323). Indeed, second instar larvae srcinating from Nahal Secher (i.e., the more productive habitat) were larger than those originating from Holot Agur, but the reverse pattern was evident among third instar larvae (Fig. 2a). Theinstar×climate interaction was marginally non-significant  Table 2  PC loadings for the first PC in the intra-specific morpho-logical analysis of   L. fedtschenkoi  (LfPC1) and for the first PC in theinter-specific morphological analysis, including  L. fedtschenkoi  and  M. hyalinus  (LfMhPC1)LfPC1 LfMhPC1Variance explained (%) 86.53% 86.43%DBM 0.7280 0.8889ML 0.8609 0.9264CML 0.8843 0.9639MW 0.9323 1.0117HL 0.8974 1.0513HW 1.0000 1.0000T13 0.7799 0.6834T23 0.7369 0.4031TL 1.1765 0.9150TW 0.9489 1.0433AL 0.4196 0.6052AW 0.8843 0.7072The PC loadings presented are normalized according to the head width(HW), which is the factor that provides the best separation amonginstars and species and which was set to a value of one  DBM   distance between mandibles,  ML  mandible length,  CML  curvedmandible length,  MW   mandible width,  HL  head length,  HW   headwidth,  T13  distance between mandible first and third tooth,  T23 distance between mandible second and third tooth,  TL  thorax length, TW   thorax width,  AL  abdomen length,  AW   abdomen width Naturwissenschaften (2009) 96:1147  –  1156 1151
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks