Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Biogeographical Survey Identifies Consistent Alternative Physiological Optima and a Minor Role for Environmental Drivers in Maintaining a Polymorphism

Biogeographical Survey Identifies Consistent Alternative Physiological Optima and a Minor Role for Environmental Drivers in Maintaining a Polymorphism

  • Arne Iserbyt, 
  • Hans Van Gossum, 
  • Robby Stoks
PLOS
x

Abstract

The contribution of adaptive mechanisms in maintaining genetic polymorphisms is still debated in many systems. To understand the contribution of selective factors in maintaining polymorphism, we investigated large-scale (>1000 km) geographic variation in morph frequencies and fitness-related physiological traits in the damselfly Nehalennia irene. As fitness-related physiological traits, we investigated investment in immune function (phenoloxidase activity), energy storage and fecundity (abdomen protein and lipid content), and flight muscles (thorax protein content). In the first part of the study, our aim was to identify selective agents maintaining the large-scale spatial variation in morph frequencies. Morph frequencies varied considerably among populations, but, in contrast to expectation, in a geographically unstructured way. Furthermore, frequencies co-varied only weakly with the numerous investigated ecological parameters. This suggests that spatial frequency patterns are driven by stochastic processes, or alternatively, are consequence of highly variable and currently unidentified ecological conditions. In line with this, the investigated ecological parameters did not affect the fitness-related physiological traits differently in both morphs. In the second part of the study, we aimed at identifying trade-offs between fitness-related physiological traits that may contribute to the local maintenance of both colour morphs by defining alternative phenotypic optima, and test the spatial consistency of such trade-off patterns. The female morph with higher levels of phenoloxidase activity had a lower thorax protein content, and vice versa, suggesting a trade-off between investments in immune function and in flight muscles. This physiological trade-off was consistent across the geographical scale studied and supports widespread correlational selection, possibly driven by male harassment, favouring alternative trait combinations in both female morphs.

Introduction

Polymorphisms are attractive model systems for understanding fundamental processes related to the origin and maintenance of genetic variation [1], [2]. For understanding processes maintaining polymorphisms, there is increasing awareness that, rather than focussing on one single mechanism, a combination of several adaptive and/or neutral mechanisms may determine polymorphism [3][7]. An ongoing debate in this context is the relative importance of neutral versus adaptive mechanisms [8]. One possibility to explore the relative contribution of these mechanisms is to compare genetic variability of neutral loci with geographic and temporal variation in morph frequencies [9][11]. Alternatively, one may evaluate to what extent spatiotemporal variation in morph frequencies and morph fitness correlates can be explained by variation in ecological parameters [12], [13].

Many polymorphisms show large-scale geographic variation in morph frequencies [14][18], which provides an elegant setting to explore co-varying ecological variables as indicators of spatially varying selection. For example, in the polymorphic snail Littorina obtusata, shell colour morph frequencies changed gradually in accordance with environmental temperature regimes within and between estuaries [13], suggesting that temperature acts as a major selective agent to maintain this colour polymorphism. Other examples in which morph frequencies relate to ecological variables involve niche occupancy in the barn owl, Tito alba [19], altitude related solar radiation in the polymorphic damselfly Megalagrion calliphya [20] and soil coloration to improve crypsis in Agouti mice [21]. In many cases, spatial morph frequency variation resembles a cline. However, such clines may also result from neutral processes [22][24]. To firmly point at spatial varying selection in maintaining polymorphisms, fitness-related traits should co-vary with the ecological variables in opposite ways between morphs, with a certain morph having optimal fitness-related traits at sites where it has the highest morph frequencies. For example, colour morphs of the walking stick Timema cristinae show spatial morph frequency variation related to the presence of host plants and the survival of a given morph is highest at the host plant where its frequency is highest [25].

While spatially varying selection may contribute to the maintenance of different morphs at large spatial scales, alternative fitness optima linked to trade-offs between fitness-related traits may underlie local coexistence of different morphs [26]. Fitness-related physiological traits that received special attention to explain the maintenance of polymorphisms are those related to immune function [27][30] and energy storage [31], [32]. Morphs may differentially trade off investments in fitness-related physiological traits, thereby generating alternative fitness optima in an adaptive landscape, caused by correlational selection on a multivariate suit of physiological and life-history traits [29], [30], [33], [34].

A classic and much debated example of intra-specific polymorphism can be found in many species of damselflies [35], [36]. In these systems, polymorphism is restricted to the female sex and shows simple Mendelian inheritance [37]. The observed biogeographic variation in female morph frequencies and the mechanisms underpinning these frequency dynamics remain an intriguing and controversial topic, e.g. [7], [15], [20], [38], [39]. To understand spatiotemporal patterns in female morph frequencies, partial support has been found for fluctuating selection pressure caused by male-female interactions [40], [41]. The presence of such distinct phenotypes are therefore generally explained in the context of sexual conflict theory, in which multiple female morphs co-exist as counter-adaptations to avoid costly male sexual harassment [42][44]. However in addition to this harassment-reduction hypothesis, recent studies point at a potential role for stochastic effects during recolonisation when it comes to explain morph frequency variation [45], differential dispersal capacity between female morphs [46], and especially differential preference or tolerance to local abiotic conditions [15], [20], [32], [47]. In support of the latter, a recent large scale population genetic study indicated the importance of divergent selection [7], in which certain morphs may be favoured in local populations that differ in ecological (biotic and abiotic) conditions. Specifically, the authors suggested potentially important factors like temperature and precipitation regimes that are likely to affect the different colour morphs in contrasting ways; see also [15], [32], [47]. Indeed, given the difference in body pigmentation, melanin pattern and behaviour it is not surprising that female morphs in these damselflies are expected to differ in thermal properties [32], [48]. Further supporting evidence comes from recent studies that showed significant clinal variation in female morph frequencies [15], [20], [29], [47], including co-variations with ambient temperature. Together, after several decades focusing on the harassment-reduction hypothesis, it became clear that one single hypothesis may not suffice to explain this polymorphism [5], [7].

In the current study, we used the polymorphic damselfly Nehalennia irene, a species with large geographical variation in morph frequencies [38], [49]. The stronger differentiation in morph frequencies at two sites, separated by only 8 km in Eastern Canada compared to genetic differentiation in neutral microsatellite markers indicated that divergent selection rather than neutral processes caused spatial variation in morph frequencies [50]. Yet, using this method one cannot identify which selective agents caused the geographic variation. Furthermore, morph frequencies appear to resemble a cline at continental scale from Northwest to Southeast Canada [49], which has been suggested to co-vary with both, ambient temperature and male density [38]. However, this suggested cline in based on groups of study populations with a discontinuity of sometimes more than 3000 km. Here, we elaborate on several previous studies [38], [45], [49], [50] and aim to identify selective agent(s) in maintaining this polymorphism. Therefore, we surveyed 89 populations along a linear 1100 km transect, representing a continuous ecological cline in population density, temperature and precipitation regimes. Following the St.-Laurence river in Eastern Canada provided us with an excellent opportunity to sample with consistent continuity.

In a first part of the study, our aim is to identify selective agents maintaining the large spatial variation in morph frequencies. If a female morph relative to the other, is favoured in local populations that differ in ecological parameters, we then expect (1) morph frequencies to co-vary with these ecological parameters [13], [19][21], [47], and (2) that in populations where this morph is favoured, it has more optimal values of fitness-related physiological traits, relative to the other morph [25], [32], [51]. The studied physiological traits involve key parameters related to investments in immune function, in energy storage and fecundity, and in flight muscles that are strong candidates to affect fitness in damselflies [52][55]. In a second, related part of the study, we aim at identifying trade-offs between fitness-related physiological traits that may contribute to the local maintenance of both colour morphs and test the spatial consistency of morph differences that define alternative phenotypic optima [29], [30], [33], [34].

Materials and Methods

Model species

The sedge sprite N. irene is a small non-territorial damselfly (Zygoptera; Odonata), which inhabits marshy or boggy waters and is common throughout most of Canada and the Northern parts of the United States [56]. It is not an endangered nor a protected species (see COSEWIC, federal government Canada). Nehalennia irene has one generation per year, with the winged adult life stage and reproduction typically occurring between early June and mid-August. After locating a potential mate, a male will attempt to grasp the female in the so-called tandem formation, where the male attaches his anal appendages to the female's prothorax [57]. In a next step, the male will then try to copulate with the female. A female's mate status can thus either be mated (tandem or copulating position) or either be single. The adult stage exhibits a clear dimorphism restricted to the female sex, with morphs being easily classified based on their body colouration and melanin pattern into andromorphs and gynomorphs [58]. Mature andromorph females resemble the conspecific male's blue body colouration and melanin pattern [59], [60], whereas gynomorph females have distinctive yellowish lateral thorax sides and a less conspicuous abdominal melanin pattern; for colour figures see [58], for pictures see [61]. Earlier research indicated large spatial but temporally fairly consistent variation in relative female morph frequencies, with proportions of andromorphs among females being atypically high at the Western edge of the species distribution range (>90%) relative to the central and Eastern part of the range (0–63%) [38], [49].

Study sites and sampling procedure

Frequencies and densities of males and female morphs were determined at 58 populations during the reproductive seasons of 2009 and 2010. This was done along a linear and continuous 1100 kilometre transect in the central to Eastern part of the species' distribution range, in Ontario and Quebec, Canada (Figure 1). Along this transect, annual mean temperatures range from 6.5°C at the south-west up to 0.8°C at the north-east. In addition, frequency and density data of 31 additional populations in Ontario, Quebec and New Brunswick that were sampled in 2004 and 2007 were used from Iserbyt et al. [38]. We aimed to sample populations minimally five kilometre separated from one another (mean ± SE: 13.0±1.1 km) and not being part of the same lake or river system. No specific permits were required for the described field studies. Neither were any of the locations privately owned or protected in any way.

thumbnail
Figure 1. Geographical distribution of the study populations.

The 89 populations are presented in three groups according to andromorph frequencies. Average frequencies are given when populations were sampled in multiple years. Numbers one to eight indicate populations where specimens were collected for quantification of physiological parameters (further details of these populations are provided in Table 1).

https://doi.org/10.1371/journal.pone.0032648.g001

Population parameters

Frequencies and densities of males and female morphs were determined at each site as described by Van Gossum et al. [49]. Shortly, an insect net was swept transcribing ‘eight-shaped’ figures, while walking slowly through the shoreline vegetation and recording the time elapsed. Sex, morph and age class (immature or mature) of each individual N. irene netted was noted. We aimed to catch at least 30 females at each site (mean ± SE = 62±3). This allowed calculating six population parameters either based on matures or matures plus immature [38]: i.e. the proportion of females being andromorph, the ratio of andromorphs to males (i.e. mimics to models), the ratio of males to females (operational sex ratio, OSR), the population and male density (respectively number of individuals and number of males caught per time unit) and finally the proportion of mature individuals. Density and sex ratio are considered relevant proxies of male harassment, because when males are more numerous or when the number of males per female increases, females are expected to be approached more frequently by mate-searching males [62][64]. The proportion of mature individuals can be used as surrogate for moment in the reproductive season, with a proportion equal to zero at the onset and a proportion of one at the end of the season.

Weather parameters

Given that female morph frequencies and physiological parameters (see further) can be influenced by long-term as well as short-term weather conditions [32], [20], [47], we obtained detailed weather data from the Canadian National Climate Data and Information Archive (http://www.climate.weatheroffice.ec.gc.ca). The average distance between our 89 study populations and the closest weather station was 16.8±1.0 km. Long-term annual and seasonal (winter, spring, summer and autumn) climate normals were derived from the weather stations closest to each site. These climate normals are the arithmetic averages of weather elements over the 30-year interval (1971–2000). Our extracted dataset includes annual and seasonal temperature (mean, minimum and maximum) and precipitation averaged over this thirty-year period. To obtain short-term weather data we derived seasonal mean precipitation and mean, minimum and maximum temperature, during the preceding year (starting with the summer of 2009). Additionally, on a very short-term time scale we derived daily precipitation, and mean, minimum and maximum temperatures for four periods: the day of capture, the day of capture plus the preceding day, the preceding two days and the preceding three days.

Physiological parameters

To quantify physiological parameters, individuals were collected at a subset of eight sites along our transect (Table 1; Figure 1). We selected these populations based on the minimal distance between them (min: 65 km; mean ±1SE = 93±14 km), in such a way that large variation was present in both, weather and population parameters (Table 1). This large variation provides us with a good opportunity to explore morph-specific variation in the physiological estimates with respect to these ecological variables. The eight selected sites may thus be viewed as a representative subset of the total population dataset (N = 89) along our transect. Minimally 20 individuals (24.1±0.5) of each female morph were collected in 2010. These were all mature individuals judged by the brightness of their body colours and stiffness of the wings [57]. Mate status at the moment of capture, i.e. being single or mated, was noted for every female. Each female was stored separately and preserved immediately in liquid nitrogen in the field. Afterwards, all individuals were further stored at −80°C in the laboratory for further use.

thumbnail
Table 1. Ecological parameters of the eight populations where individuals were collected during the reproductive season of 2010 for quantification of physiological parameters.

https://doi.org/10.1371/journal.pone.0032648.t001

We studied three key physiological parameters related to investments in immune function (phenoloxidase (PO) activity), in energy storage and fecundity (abdomen lipid and protein content) and in flight muscles (thorax protein content) that are strong candidates to affect fitness in damselflies [32], [54], [55], [65]. PO activity is one of the most important components of insect immune function [66][68]. Lipids are the most important form of energy storage in both the adult and the egg stage [69], [70]. Proteins are the major component of flight muscles [71] and are important for the development of the eggs [72]. Because flight muscles make up to 95% of the thorax mass [73], thorax protein content can be seen as a proxy for investment in flight muscles. Abdomen protein and lipid content reflect the investment in fecundity. All three physiological parameters have been shown to be influenced by male harassment [32], [74] and weather conditions [75], [76], and may potentially differ between morphs [29], [32,].

To quantify PO activity, we closely followed the protocol by Stoks et al. [77], here optimised for N. irene. Specifically, the thorax was homogenised using a hand-held pistil and 0.3 ml cacodylate buffer was added (0.01 mol/l Na-Coc, 0.005 mol/l CaCl2). The cell walls were removed via centrifugation (4°C, 13000 rpm, 10 min). Each well of a 96-well microtiterplate was filled with 100 µl sample supernatant, 35 µl PBS buffer, 5 µl chymotripsine (5 mg/ml) and after five minutes 60 µl L-Dopa (dihydrophenyl-l-alanine; 10 mM in cacodylate buffer). The reaction proceeded for 30 minutes at 30°C. Readings were taken every 10 seconds on a temperature-controlled microplate reader at 490 nm. Enzyme activity was measured as the slope during the linear phase of the reaction when the enzyme catalyses the transition from L-DOPA to dopachrome.

Protein content was quantified separately in thorax (mainly flight muscles) and abdomen (mainly eggs) using the Bradford method [78]. Therefore, 5 µl of the homogenised sample was added to 155 µl milli-Q-water and 40 µl Bradford reagent (Sigma®, San Louis – USA). The absorbance was read at 595 nm after 10 minutes on a microplate reader. Concentrations were calculated from standard curves of bovine serum albumine (United States Biochemical Corp, Bath – UK).

Abdomen lipid content was assayed using the protocol described in Bligh & Dyer [79]. 200 µl of the homogenised sample was mixed with 400 µl chloroform, 400 µl methanol and 200 µl milli-Q-water to dissolve the lipids. Lipids were precipitated via centrifugation (4°C, 13000 rpm, 5 min). 200 µl of the lower chloroform fraction was mixed with 500 µl concentrated H2SO4 and was incubated for 15 minutes at 200°C. Then 2 ml milli-Q-water was added and subsequently 200 µl of each sample was read at 405 nm on a microplate reader. Concentrations were calculated from standard curves of tripalmitine (Acros Organics, Geel– Belgium). Thorax lipids could not be quantified because all homogenised thorax sample was used to measure PO activity and protein content.

All physiological parameters were assayed twice per individual and the mean of both readings was used in all further analyses (all repeatabilities >0.83 [80]). A digital picture of the right hind wing was taken of each individual (Nikon D70/Tamron macro lens 90 mm 1∶2.8). Using ImageJ 1.38× [81], wing length was determined from the second antenodal cross vein to the stigma; for more detail see [61].

Statistical analyses

We first tested for spatial autocorrelation to examine geographical dependency in the studied population and long-term weather parameters [82]. Therefore, we used the Morans' I index, with I-values significantly different from zero indicating that spatial heterogeneity in the ecological parameters increases with distance among populations. Nearby populations are thus expected to be more similar than populations further apart. Population parameters were averaged per site when sampled in multiple years and tested for spatial autocorrelation with a lag distance of 5 km. This value equals our minimal distance between nearby population and is much higher than the average dispersal distance of comparable zygopteran damselflies in a network of ponds [83]. In addition, we tested for clinal variation, i.e. latitudinal and longitudinal effects, in these ecological parameters. We accounted for sampling over multiple years by treating site as a random factor in the mixed models.

In a next step, we explored to what extent spatial variation in morph frequencies could be explained by linear and quadratic effects of latitude and longitude, harassment proxies (OSR and densities) and weather parameters (temperature and precipitation normals). As expected, several of these parameters were strongly correlated and may therefore not be included together in one regression model, because of multi-collinearity problems [84]. To make an a priori selection among the explanatory variables and meanwhile avoiding problems with multi-collinearity, we used two independent methods which have been proven to be successful in the past: classification and regression trees (CART) [85][87] and partial least squares regressions (PLS) [88], [89]. CART explains variation of a single response variable by repeatedly splitting the data with the best predictive variables into more homogeneous groups. PLS is a regression technique developed to deal with many explanatory variables and one or several response variables. Predictors selected by the PLS that are highly correlated (R2>0.9) are considered to have equal explanatory value. In such cases, the variable that best corresponded with the CART analyses was selected as best predictor. CART and PLS both selected the same best predictors in our dataset (see results). Therefore, these analyses can be seen as replicated and independent methods, which increases the robustness of our results. Subsequently, we tested whether the two best predictive parameters had a significant effect on the spatial variation in morph frequencies using separate general linear mixed models (GLM). These are basically multiple regression analyses, in which we controlled for possible within-season variation in morph frequencies [49], by including the proportion of mature individuals as linear and quadratic effects into the model. Also, given that some populations were sampled in multiple years, source population was treated as a random variable.

The ecological variables that could best explain variation in morph frequencies were also used to analyse variation in the physiological parameters. Therefore, we performed ANCOVA models, including morph, mate status and the selected ecological variables, plus interactions with morph. Source population was treated as a random variable. To correct for individual and morph-specific differences in size, protein and lipid contents were divided by wing length [61], [90]. To correct for individual and morph-specific differences in protein content when analyzing the enzyme PO, we calculated PO residuals obtained by regressing PO activity against thorax protein content. Correcting PO for wing length as the other physiological parameters did not change the outcome of the analyses. Similarly, using protein or lipid residuals obtained by the regression against wing length, did not alter the conclusions. All analyses were performed in SAS 9.2 (SAS Institute Inc, Carry, NC, USA), except for the CART analyses which were completed in SPSS 18.0 (SPSS Inc, Chicago, IL, USA).

Results

Variation in morph frequencies

Considerable spatial variation in the frequency of andromorph females was identified ranging from 0.0 to 62.7% (mean ± SE: 8.8±1.2%; see Figure 1). No sign of spatial autocorrelation in morph frequencies was detected, nor were there any effects of latitude or longitude (Table 2). Similar patterns were observed with the ratio of andromorphs to males and the OSR (Morans' I, latitude and longitude, all p>0.05). In contrast, male density and annual temperature normals both decreased towards Eastern and Northern directions, while annual precipitation normals increased towards the East. Strong signals of spatial autocorrelation were detected in these three ecological parameters (Table 2). Similar patterns were observed for population density and all seasonal mean, minimum and maximum temperature and precipitation normals (Morans' I, latitude and longitude, all p<0.0001).

thumbnail
Table 2. Outcome of the spatial autocorrelation analyses for andromorph frequency, male density, annual temperature and precipitation normals, using Morans' I index.

https://doi.org/10.1371/journal.pone.0032648.t002

The CART analysis put forward adult male density (split at 11.2 individuals/minute) as most important variable for explaining the spatial variation in morph frequencies, followed by maximum spring temperature normals (11.0°C). These results exactly correspond with the selected predictors of the PLS analysis (see Figure 2). Overall, precipitation regimes had very limited explanatory value (Figure 2). From the two selected ecological parameters, maximum spring temperature did explain only a minor, non-significant, part of the variation in andromorph frequencies (GLM: F1,40 = 2.38, p = 0.13; Pearson correlation: R2 = 0.02; Figure 3A). Andromorph frequencies increased with mature male densities (F1,40 = 5.55, p = 0.023; R2 = 0.04; Figure 3B). The proportion of mature individuals, as a controlling factor for seasonal variation, had no effect on morph frequencies (linear: F1,39 = 2.72, p = 0.11 and quadratic: F1,38 = 0.84, p = 0.36).

thumbnail
Figure 2. Variable importance plot of the partial least squares (PLS) regression explaining spatial variation in morph frequencies.

Conform the CART analysis, mature male density and maximum spring temperature were selected as most important explanatory variables. Note that all density estimates are highly correlated (all R2>0.9). Hence, only adult male density is used in further analyses (see methods).

https://doi.org/10.1371/journal.pone.0032648.g002

thumbnail
Figure 3. Spatial variation in andromorph frequencies plotted against maximum temperature normals in the spring season (A) and mature male density (B).

These variables explained most of the spatial variation in morph frequencies, based on the CART and PLS analyses. Average population morph frequencies are presented when sampled over multiple years. Significant regression fit and 95% confidence interval is presented in panel (B).

https://doi.org/10.1371/journal.pone.0032648.g003

Variation in physiological parameters

Andromorphs had significantly, ca. 10% higher PO activity levels compared to gynomorphs (Table 3, Figure 4). The reverse pattern was found for thorax protein content, which was over 5% lower for andromorphs relative to gynomorphs. No morph-specific differences were found in abdominal physiological traits (Table 3). Although mature male density was the only significant ecological parameter that could explain part of the spatial variation in morph frequencies, this density effect did not differentially affect the physiological parameters of both female morphs (see interactions, Table 3). Neither did any of the various other studied population parameters, long-term and short-term weather parameters affect physiological parameters of both female morphs in contrasting ways (results not shown). Finally, mated females had significantly higher thorax protein content, compared to single females (Table 3, Figure 4B).

thumbnail
Figure 4. Differences between single (black) and mated (white) female morphs in PO activity (A) and thorax protein content (B).

Values are based on the least squares means (±1SE) of the final general linear mixed models as presented in Table 2. For graphical clearness, mean values of PO activity (residuals obtained by regressing PO activity against thorax protein content) are summed by one.

https://doi.org/10.1371/journal.pone.0032648.g004

thumbnail
Table 3. Results of the general linear mixed models explaining variation in the four physiological parameters.

https://doi.org/10.1371/journal.pone.0032648.t003

Discussion

Our results indicate large spatial variation in morph frequencies in a geographical rather unstructured way, which could be explained only weakly by numerous ecological parameters. These ecological parameters did not affect physiological fitness-related traits of both female morphs in contrasting ways. However, we showed a geographically consistent morph-specific trade-off between investment in immune function and in flight muscles. In what follows, we will discuss the above findings in a broader evolutionary context.

While most of the investigated ecological parameters showed spatial autocorrelation, this pattern was not present in female morph frequencies. Moreover, and contrary to our expectations, we found generally little proof for weather and population parameters being important in explaining biogeographic patterns in female morph frequencies. Our results may therefore indicate the importance of stochastic effects causing spatial frequency fluctuations. Alternatively, female morphs may be adapted to highly variable and currently unidentified small-scale environmental conditions. The latter alternative explanation may seem less likely as we included most of the known weather and population parameters likely to differentially affect the colour morphs, although we acknowledge that some factors may be overlooked. Considerable variation in morph frequencies in a spatially structured manner has been reported in taxonomically diverse systems, such as birds [19], mammals [21], insects [18], [39], molluscs [13] and plants [92]. Such studies provide preliminary support for selection underlying these polymorphisms. The clinal variation in morph frequencies in female damselflies has been related with variation in ambient temperature in Ischnura elegans [47] and altitude related solar radiation in Megalagrion calliphya [20]. In another approach to understand these biogeographical patterns, population genetic studies compared genetic variability of neutral loci with morph frequencies at local spatial and temporal scales [93][96]. While direction, form and magnitude of selection differed among studies, they jointly argued against drift to maintain multiple female morphs in natural populations. Specifically for N. irene, Wong et al. [50] suggested that on a local scale spatially variable selection operates on different morphs, perhaps mediated by adaptation to variable local environmental conditions, frequency- and density-dependent selection regimes or a combination of those. A similar conclusion could be drawn for I. senegalensis [39] and I. elegans at a large spatial scale [7], in which the strength of divergent selection differed among regions. Similar to our observations, one of the female morphs is extremely rare or even absent in some parts of I. elegans' distribution area [7], [15]. This could either be caused by local selection acting against this rare morph, or by stochastic effects during colonization of certain areas; for lizards see also [97]. In accordance, a recent population genetic study indicated that the extreme continental frequency variation (from 0% to ∼100%) in N. irene could in part be explained by such stochastic effects [45]. The lack of spatial autocorrelation in morph frequencies and the weak explanatory value of ecological parameters in the current study strengthen this idea.

Nonetheless, and in accordance with a number of studies, we found, admittedly weak (R2 = 0.04; see figure 3B), but significant co-variation between spatial variation in andromorph frequencies and male density [38], [49], [59], [93], [98]. Evidently, this relationship does not prove causality. However, theoretical and empirical studies support that the intensity of sexual conflict rises with male density, thereby reducing female fitness components [32], [62][64]. Harassment selection may thus be too low or even absent in populations with low male density [99]. Drift may get the upper hand in such conditions, generating fixation of a given morph [2], [14]. This may also explain why 13 out of 89 populations are monomorphic in the current study.

To point at selection maintaining spatial variation in the polymorphism, relative fitness components of the morphs should also differ in relation to the supposed selective agent [25], [32], [51]. However, in the present study no morph-specific influences of male density (or other population parameters, nor short- and long-term weather parameters) on physiological traits were detected. Contrarily, body condition measures in the damselfly Enallagma cyathigerum were differentially affected by short-term ambient temperature in both morphs [32]. One reason for this difference between studies may be the smaller variation in weather parameters in the current study (e.g. 11–22°C in [32], vs 17–23°C in the current study). However, in line with the current results, neither did other recent studies find such morph-specific influences of environmental conditions on several behavioural and life-history traits [48], [100], [101].

With regard to trade-offs between fitness-related physiological traits contributing to maintaining the polymorphism at a local scale, our physiological data suggest a morph-specific and geographically consistent trade-off where andromorphs, relative to gynomorphs, tend to invest more in immune function and less in flight muscles. Because traits related to immune function and flight muscles are costly to produce and maintain [102][104], our results likely reflect an energy-allocation trade-off [105], [106]. Interestingly, similar trade-offs have been suggested in polymorphic butterflies [107], but see [108] and are particularly well described in several species of polymorphic lizards. In these latter model species, colour morphs may differentially allocate resources towards traits related to immune function versus life-history, morphology or performance-related traits (Uta stansburiana: [27]; Podarcis muralis: [30]; Anolis sagrei: [29]). These lizard studies suggest that correlational selection driven by density of neighbouring individuals favours successful and alternative trait combinations among morphs [27], [29]. In our study, both morphs may differentially allocate resources to traits related to immune function and flight muscles because of differences in degree of male sexual harassment and/or differences in behavioural strategies to avoid this cost. In this scenario, gynomorphs would invest more in flight muscles that enhance aerial competitive ability [71], [73], [109] to fend of harassing males. This effect was even stronger for mated females, perhaps because males particularly search and mate females that are most active and in best condition [110]. In line with the suggested enhanced flight ability, N. irene gynomorphs display more refusal behaviour when they are the majority morph [111], see also [100]. Also in I. elegans, female morphs differ in behavioural tactics in order to escape from excessive male harassment [110], [112]. Specifically andromorphs occupy less open habitat, fly within shorter ranges and directly fend off approaching males [112]. These observed behavioural differences may relate to differential investment in flight muscles and differential exposure to parasites [113] and thus investment in immune function.

Whatever the underlying mechanism, our data and other studies suggest correlational selection favouring successful and alternative trait combinations among female morphs [114]. This multivariate suit of traits involve physiological (this study) as well as behavioural [111], morphological [60], [61] and life-history [115] traits. Identifying trade-offs among fitness-related traits may increase our knowledge how polymorphisms are maintained, because they may generate morph-specific fitness optima in phenotypic space [27], [29], [30].

Conclusions

Despite many studies, the mechanisms underlying female polymorphism in damselflies are still under debate and this study area urgently needs to open its perspective and accept multiple non-exclusive explanations [7], [20], [35], [36]. In the present study, we provided several new insights. We first showed that morph frequencies are spatially unstructured and are only weakly related to numerous investigated ecological parameters. This may indicate that morph frequencies either vary randomly caused by stochastic processes or alternatively, that female morphs are adapted to local ecological conditions that largely vary at a small spatial scale. Furthermore we showed that ecological parameters did not differentially affect fitness-related physiological traits of both female morphs. Instead, we documented geographically consistent morph differences in physiology reflecting a trade-off between investments in immune function and in flight muscles. Our results share much resemblance with results on polymorphic lizards and highlight an overlooked candidate mechanism contributing to the maintenance of multiple female morphs within damselfly species.

Acknowledgments

We wish to thank B. Hansson and two anonymous referees for valuable suggestions on previous drafts of this manuscript, S. Van Dongen, L. De Bruyn and V. Sluydts for statistical advice, T. Sherratt for local support, J. Ting and S. Puls for field assistance, N. Van Houtte, R. Van Houdt and S. Puls for laboratory assistance and EBT research group (Antwerp University) for the use of their micro plate reader.

Author Contributions

Conceived and designed the experiments: AI. Performed the experiments: AI. Analyzed the data: AI. Contributed reagents/materials/analysis tools: AI RS. Wrote the paper: AI HVG RS.

References

  1. 1. Stearns SC, Hoekstra RF (2000) Evolution, an introduction. New York: Oxford University Press. 381 p.SC StearnsRF Hoekstra2000Evolution, an introductionNew YorkOxford University Press381
  2. 2. Gray SM, McKinnon JS (2007) Linking color polymorphism maintenance and speciation. Trends Ecol Evol 22: 71–79.SM GrayJS McKinnon2007Linking color polymorphism maintenance and speciation.Trends Ecol Evol227179
  3. 3. Jones JS, Leith BH, Rawlings P (1977) Polymorphism in Cepaea - problem with too many solutions. Annu Rev Ecol Syst 8: 109–143.JS JonesBH LeithP. Rawlings1977Polymorphism in Cepaea - problem with too many solutions.Annu Rev Ecol Syst8109143
  4. 4. Magurran AE (2005) Evolutionary ecology: the Trinidadian guppy. New York: Oxford University Press. 224 p.AE Magurran2005Evolutionary ecology: the Trinidadian guppyNew YorkOxford University Press224
  5. 5. Punzalan D, Hosken DJ (2010) Sexual dimorphism: why the sexes are (and are not) different. Curr Biol 20: R972–R973.D. PunzalanDJ Hosken2010Sexual dimorphism: why the sexes are (and are not) different.Curr Biol20R972R973
  6. 6. Cornwallis CK, Uller T (2010) Towards an evolutionary ecology of sexual traits. Trends Ecol Evol 25: 145–152.CK CornwallisT. Uller2010Towards an evolutionary ecology of sexual traits.Trends Ecol Evol25145152
  7. 7. Sánchez-Guillén RA, Hansson B, Wellenreuther M, Svensson EI, Cordero-Rivera A (2011) The influence of stochastic and selective forces in the population divergence of female colour polymorphism in damselflies of the genus Ischnura. Heredity 107: 513–522.RA Sánchez-GuillénB. HanssonM. WellenreutherEI SvenssonA. Cordero-Rivera2011The influence of stochastic and selective forces in the population divergence of female colour polymorphism in damselflies of the genus Ischnura.Heredity107513522
  8. 8. Millstein RL (2002) Are random drift and natural selection conceptually distinct? Biol Philos 17: 33–53.RL Millstein2002Are random drift and natural selection conceptually distinct?Biol Philos173353
  9. 9. Leinonen T, O'hara RB, Cano JM, Merila J (2008) Comparative studies of quantitative trait and neutral marker divergence: a meta-analysis. J Evol Biol 21: 1–17.T. LeinonenRB O'haraJM CanoJ. Merila2008Comparative studies of quantitative trait and neutral marker divergence: a meta-analysis.J Evol Biol21117
  10. 10. Pujol B, Wilson AJ, Ross RIC, Pannell JR (2008) Are Q(ST)-F(ST) comparisons for natural populations meaningful? Mol Ecol 17: 4782–4785.B. PujolAJ WilsonRIC RossJR Pannell2008Are Q(ST)-F(ST) comparisons for natural populations meaningful?Mol Ecol1747824785
  11. 11. Whitlock MC (2008) Evolutionary inference from Q(ST). Mol Ecol 17: 1885–1896.MC Whitlock2008Evolutionary inference from Q(ST).Mol Ecol1718851896
  12. 12. Hoffman EA, Blouin MS (2000) A review of colour and pattern polymorphisms in anurans. Biol J Linn Soc 70: 633–665.EA HoffmanMS Blouin2000A review of colour and pattern polymorphisms in anurans.Biol J Linn Soc70633665
  13. 13. Phifer-Rixey M, Heckman M, Trussell GC, Schmidt PS (2008) Maintenance of clinal variation for shell colour phenotype in the flat periwinkle Littorina obtusata. J Evol Biol 21: 966–978.M. Phifer-RixeyM. HeckmanGC TrussellPS Schmidt2008Maintenance of clinal variation for shell colour phenotype in the flat periwinkle Littorina obtusata.J Evol Biol21966978
  14. 14. Corl A, Davis AR, Kuchta SR, Sinervo B (2010) Selective loss of polymorphic mating types is associated with rapid phenotypic evolution during morphic speciation. Proc Natl Acad Sci U S A 107: 4254–4259.A. CorlAR DavisSR KuchtaB. Sinervo2010Selective loss of polymorphic mating types is associated with rapid phenotypic evolution during morphic speciation.Proc Natl Acad Sci U S A10742544259
  15. 15. Gosden TP, Stoks R, Svensson EI (2011) Range limits, large-scale biogeographic variation, and localized evolutionary dynamics in a polymorphic damselfly. Biol J Linn Soc 102: 775–785.TP GosdenR. StoksEI Svensson2011Range limits, large-scale biogeographic variation, and localized evolutionary dynamics in a polymorphic damselfly.Biol J Linn Soc102775785
  16. 16. Silvertown J, Cook L, Cameron R, Dodd M, McConway K, et al. (2011) Citizen science reveals unexpected continental-scale evolutionary change in a model organism. PLoS ONE 6: e18927.J. SilvertownL. CookR. CameronM. DoddK. McConway2011Citizen science reveals unexpected continental-scale evolutionary change in a model organism.PLoS ONE6e18927
  17. 17. Stapley J, Wordley C, Slate J (2011) No evidence of genetic differentiation between anoles with different dewlap color patterns. J Hered 102: 118–124.J. StapleyC. WordleyJ. Slate2011No evidence of genetic differentiation between anoles with different dewlap color patterns.J Hered102118124
  18. 18. Saccheri IJ, Rousset F, Watts PC, Brakefield PM, Cook LM (2008) Selection and gene flow on a diminishing cline of melanic peppered moths. Proc Natl Acad Sci U S A 105: 16212–16217.IJ SaccheriF. RoussetPC WattsPM BrakefieldLM Cook2008Selection and gene flow on a diminishing cline of melanic peppered moths.Proc Natl Acad Sci U S A1051621216217
  19. 19. Antoniazza S, Burri R, Fumagalli L, Goudet J, Roulin A (2010) Local adaptation maintains clinal variation in melanin-based coloration of european barn owls (Tyto alba). Evolution 64: 1944–1954.S. AntoniazzaR. BurriL. FumagalliJ. GoudetA. Roulin2010Local adaptation maintains clinal variation in melanin-based coloration of european barn owls (Tyto alba).Evolution6419441954
  20. 20. Cooper IA (2010) Ecology of sexual dimorphism and clinal variation of coloration in a damselfly. Am Nat 176: 566–572.IA Cooper2010Ecology of sexual dimorphism and clinal variation of coloration in a damselfly.Am Nat176566572
  21. 21. Mullen LM, Hoekstra HE (2008) Natural selection along an environmental gradient: A classic cline in mouse pigmentation. Evolution 62: 1555–1569.LM MullenHE Hoekstra2008Natural selection along an environmental gradient: A classic cline in mouse pigmentation.Evolution6215551569
  22. 22. Slatkin M (1973) Gene flow and selection in a cline. Genetics 75: 733–756.M. Slatkin1973Gene flow and selection in a cline.Genetics75733756
  23. 23. Endler JA (1977) Geographic Variation, Speciation and Clines. Princeton, Princeton University Press. 262 p.JA Endler1977Geographic Variation, Speciation and ClinesPrinceton, Princeton University Press262
  24. 24. Excoffier L, Ray N (2008) Surfing during population expansions promotes genetic revolutions and structuration. Trends Ecol Evol 23: 347–351.L. ExcoffierN. Ray2008Surfing during population expansions promotes genetic revolutions and structuration.Trends Ecol Evol23347351
  25. 25. Nosil P, Crespi BJ, Sandoval CP, Kirkpatrick M (2006) Migration and the genetic covariance between habitat preference and performance. Am Nat 167: E66–E78.P. NosilBJ CrespiCP SandovalM. Kirkpatrick2006Migration and the genetic covariance between habitat preference and performance.Am Nat167E66E78
  26. 26. Sinervo B, Svensson E (1998) Mechanistic and selective causes of life history trade-offs and plasticity. Oikos 83: 432–442.B. SinervoE. Svensson1998Mechanistic and selective causes of life history trade-offs and plasticity.Oikos83432442
  27. 27. Svensson E, Sinervo B, Comendant T (2001) Density-dependent competition and selection on immune function in genetic lizard morphs. Proc Natl Acad Sci U S A 98: 12561–12565.E. SvenssonB. SinervoT. Comendant2001Density-dependent competition and selection on immune function in genetic lizard morphs.Proc Natl Acad Sci U S A981256112565
  28. 28. Joop G, Mitschke A, Rolff J, Siva-Jothy MT (2006) Immune function and parasite resistance in male and polymorphic female Coenagrion puella. BMC Evol Biol 6: 19.G. JoopA. MitschkeJ. RolffMT Siva-Jothy2006Immune function and parasite resistance in male and polymorphic female Coenagrion puella.BMC Evol Biol619Biomedcentral website. Available: http://www.biomedcentral.com/1471-2148/6/19. Accessed 2012 Feb 3. Biomedcentral website. Available: http://www.biomedcentral.com/1471-2148/6/19. Accessed 2012 Feb 3.
  29. 29. Calsbeek R, Bonneaud C, Smith TB (2008) Differential fitness effects of immunocompetence and neighbourhood density in alternative female lizard morphs. J Anim Ecol 77: 103–109.R. CalsbeekC. BonneaudTB Smith2008Differential fitness effects of immunocompetence and neighbourhood density in alternative female lizard morphs.J Anim Ecol77103109
  30. 30. Calsbeek B, Hasselquist D, Clobert J (2010) Multivariate phenotypes and the potential for alternative phenotypic optima in wall lizard (Podarcis muralis) ventral colour morphs. J Evol Biol 23: 1138–1147.B. CalsbeekD. HasselquistJ. Clobert2010Multivariate phenotypes and the potential for alternative phenotypic optima in wall lizard (Podarcis muralis) ventral colour morphs.J Evol Biol2311381147
  31. 31. Zhao Z, Zera AJ (2006) Biochemical basis of specialization for dispersal vs. reproduction in a wing-polymorphic cricket: Morph-specific metabolism of amino acids. J Insect Physiol 52: 646–658.Z. ZhaoAJ Zera2006Biochemical basis of specialization for dispersal vs. reproduction in a wing-polymorphic cricket: Morph-specific metabolism of amino acids.J Insect Physiol52646658
  32. 32. Bots J, De Bruyn L, Van Dongen S, Smolders R, Van Gossum H (2009) Female polymorphism, condition differences, and variation in male harassment and ambient temperature. Biol J Linn Soc 97: 545–554.J. BotsL. De BruynS. Van DongenR. SmoldersH. Van Gossum2009Female polymorphism, condition differences, and variation in male harassment and ambient temperature.Biol J Linn Soc97545554
  33. 33. Sinervo B, Bleay C, Adamopoulou C (2001) Social causes of correlational selection and the resolution of a heritable throat color polymorphism in a lizard. Evolution 55: 2040–2052.B. SinervoC. BleayC. Adamopoulou2001Social causes of correlational selection and the resolution of a heritable throat color polymorphism in a lizard.Evolution5520402052
  34. 34. Sinervo B, Svensson E (2002) Correlational selection and the evolution of genomic architecture. Heredity 89: 329–338.B. SinervoE. Svensson2002Correlational selection and the evolution of genomic architecture.Heredity89329338
  35. 35. Van Gossum H, Sherratt TN, Cordero-Rivera A (2008) The evolution of sex-limited colour polymorphisms. Dragonflies: model organisms for ecological and evolutionary research. New York: Oxford University Press. pp. 219–229.H. Van GossumTN SherrattA. Cordero-Rivera2008The evolution of sex-limited colour polymorphisms.Dragonflies: model organisms for ecological and evolutionary researchNew YorkOxford University Press219229
  36. 36. Svensson EI, Abbott JK, Gosden TP, Coreau A (2009) Female polymorphisms, sexual conflict and limits to speciation processes in animals. Evol Ecol 23: 93–108.EI SvenssonJK AbbottTP GosdenA. Coreau2009Female polymorphisms, sexual conflict and limits to speciation processes in animals.Evol Ecol2393108
  37. 37. Sánchez-Guillén RA, Van Gossum H, Cordero-Rivera A (2005) Hybridization and the inheritance of female colour polymorphism in two ischnurid damselflies (Odonata : Coenagrionidae). Biol J Linn Soc 85: 471–481.RA Sánchez-GuillénH. Van GossumA. Cordero-Rivera2005Hybridization and the inheritance of female colour polymorphism in two ischnurid damselflies (Odonata : Coenagrionidae).Biol J Linn Soc85471481
  38. 38. Iserbyt A, Bots J, Ting JJ, Jvostov FP, Forbes MR, et al. (2009) Multi-annual variation in female morph frequencies of the polymorphic damselfly, Nehalennia irene, at continental and regional scales. Anim Biol 59: 313–326.A. IserbytJ. BotsJJ TingFP JvostovMR Forbes2009Multi-annual variation in female morph frequencies of the polymorphic damselfly, Nehalennia irene, at continental and regional scales.Anim Biol59313326
  39. 39. Takahashi Y, Morita S, Yoshimura J, Watanabe M (2011) A geographic cline induced by negative frequency-dependent selection. BMC Evol Biol 11: 256.Y. TakahashiS. MoritaJ. YoshimuraM. Watanabe2011A geographic cline induced by negative frequency-dependent selection.BMC Evol Biol11256Biomedcentral website. Available: http://www.biomedcentral.com/1471-2148/11/256. Accessed 2012 Feb 3. Biomedcentral website. Available: http://www.biomedcentral.com/1471-2148/11/256. Accessed 2012 Feb 3.
  40. 40. Svensson EI, Abbott J, Härdling R (2005) Female polymorphism, frequency dependence and rapid evolutionary dynamics in natural populations. Am Nat 165: 567–576.EI SvenssonJ. AbbottR. Härdling2005Female polymorphism, frequency dependence and rapid evolutionary dynamics in natural populations.Am Nat165567576
  41. 41. Takahashi Y, Yoshimura J, Morita S, Watanabe M (2010) Negative frequency-dependent selection in female color polymorphism of a damselfly. Evolution 64: 3620–3628.Y. TakahashiJ. YoshimuraS. MoritaM. Watanabe2010Negative frequency-dependent selection in female color polymorphism of a damselfly.Evolution6436203628
  42. 42. Hinnekint BON (1987) Population-dynamics of Ischnura elegans (Vander Linden) (Insecta, Odonata) with special reference to morphological color changes, female polymorphism, multiannual cycles and their influence on behavior. Hydrobiologia 146: 3–31.BON Hinnekint1987Population-dynamics of Ischnura elegans (Vander Linden) (Insecta, Odonata) with special reference to morphological color changes, female polymorphism, multiannual cycles and their influence on behavior.Hydrobiologia146331
  43. 43. Robertson HM (1985) Female dimorphism and mating-behavior in a damselfly, Ischnura ramburi: females mimicking males. Anim Behav 33: 805–809.HM Robertson1985Female dimorphism and mating-behavior in a damselfly, Ischnura ramburi: females mimicking males.Anim Behav33805809
  44. 44. Fincke OM (2004) Polymorphic signals of harassed female Odonates and the males that learn them support a novel frequency-dependent model. Anim Behav 67: 833–845.OM Fincke2004Polymorphic signals of harassed female Odonates and the males that learn them support a novel frequency-dependent model.Anim Behav67833845
  45. 45. Iserbyt A, Bots J, Van Gossum H, Jordaens K (2010) Did historical events shape current geographic variation in morph frequencies of a polymorphic damselfly? J Zool 282: 256–265.A. IserbytJ. BotsH. Van GossumK. Jordaens2010Did historical events shape current geographic variation in morph frequencies of a polymorphic damselfly?J Zool282256265
  46. 46. Svensson EI, Abbott J (2005) Evolutionary dynamics and population biology of a polymorphic insect. J Evol Biol 18: 1503–1514.EI SvenssonJ. Abbott2005Evolutionary dynamics and population biology of a polymorphic insect.J Evol Biol1815031514
  47. 47. Hammers M, Van Gossum H (2008) Variation in female morph frequencies and mating frequencies: random, frequency-dependent harassment or male mimicry? Anim Behav 76: 1403–1410.M. HammersH. Van Gossum2008Variation in female morph frequencies and mating frequencies: random, frequency-dependent harassment or male mimicry?Anim Behav7614031410
  48. 48. Bots J, De Bruyn L, Van Damme R, Van Gossum H (2008) Effects of phenotypic variation onto body temperature and flight activity in a polymorphic insect. Physiol Entomol 33: 138–144.J. BotsL. De BruynR. Van DammeH. Van Gossum2008Effects of phenotypic variation onto body temperature and flight activity in a polymorphic insect.Physiol Entomol33138144
  49. 49. Van Gossum H, Beirinckx K, Forbes MR, Sherratt TN (2007) Do current hypotheses explain continental and seasonal variation in female morph frequencies of the damselfly, Nehalennia irene? Biol J Linn Soc 90: 501–508.H. Van GossumK. BeirinckxMR ForbesTN Sherratt2007Do current hypotheses explain continental and seasonal variation in female morph frequencies of the damselfly, Nehalennia irene?Biol J Linn Soc90501508
  50. 50. Wong A, Smith ML, Forbes MRL (2003) Differentiation between subpopulations of a polychromatic damselfly with respect to morph frequencies, but not neutral genetic markers. Mol Ecol 12: 3505–3513.A. WongML SmithMRL Forbes2003Differentiation between subpopulations of a polychromatic damselfly with respect to morph frequencies, but not neutral genetic markers.Mol Ecol1235053513
  51. 51. Parkash R, Singh S, Ramniwas S (2009) Seasonal changes in humidity level in the tropics impact body color polymorphism and desiccation resistance in Drosophila jambulina - Evidence for melanism-desiccation hypothesis. J Insect Physiol 55: 358–368.R. ParkashS. SinghS. Ramniwas2009Seasonal changes in humidity level in the tropics impact body color polymorphism and desiccation resistance in Drosophila jambulina - Evidence for melanism-desiccation hypothesis.J Insect Physiol55358368
  52. 52. Córdoba-Aguilar A (2009) A female evolutionary response when survival is at risk: male harassment mediates early reallocation of resources to increase egg number and size. Behav Ecol Sociobiol 63: 751–763.A. Córdoba-Aguilar2009A female evolutionary response when survival is at risk: male harassment mediates early reallocation of resources to increase egg number and size.Behav Ecol Sociobiol63751763
  53. 53. González-Tokman D, Córdoba-Aguilar A, González-Santoyo I, Lanz-Mendoza H (2011) Infection effects on feeding and territorial behaviour in a predatory insect in the wild. Anim Behav 81: 1185–1194.D. González-TokmanA. Córdoba-AguilarI. González-SantoyoH. Lanz-Mendoza2011Infection effects on feeding and territorial behaviour in a predatory insect in the wild.Anim Behav8111851194
  54. 54. Rolff J, Siva-Jothy MT (2004) Selection on insect immunity in the wild. Proc Roy Soc B 271: 2157–2160.J. RolffMT Siva-Jothy2004Selection on insect immunity in the wild.Proc Roy Soc B27121572160
  55. 55. Stoks R, Córdoba-Aguilar A (2012) Evolutionary ecology of Odonata: a complex life cycle perspective. Annu Rev Entomol 57: 249–265.R. StoksA. Córdoba-Aguilar2012Evolutionary ecology of Odonata: a complex life cycle perspective.Annu Rev Entomol57249265
  56. 56. Westfall MJ, May ML (1996) Damselflies of North America. Gainesville: Scientific Publishers. 649 p.MJ WestfallML May1996Damselflies of North AmericaGainesvilleScientific Publishers649
  57. 57. Corbet PS (1999) Dragonflies: Behaviour and ecology of Odonata. Essex: Harley Books. 829 p.PS Corbet1999Dragonflies: Behaviour and ecology of OdonataEssexHarley Books829
  58. 58. Lam E (2004) Damselflies of the Northeast. A guide to the species of Eastern Canada and the Northeastern United Stades. New York: Biodiversity Books. 96 p.E. Lam2004Damselflies of the Northeast. A guide to the species of Eastern Canada and the Northeastern United StadesNew YorkBiodiversity Books96
  59. 59. Forbes MRL, Richarson JML, Baker RL (1995) Frequency of female morphs is related to an index of male density in the damselfly, Nehalennia irene (Hagen). Ecoscience 2: 28–33.MRL ForbesJML RicharsonRL Baker1995Frequency of female morphs is related to an index of male density in the damselfly, Nehalennia irene (Hagen).Ecoscience22833
  60. 60. Van Gossum H, Robb T, Forbes MR, Rasmussen L (2008) Female-limited polymorphism in a widespread damselfly: morph frequencies, male density, and phenotypic similarity of andromorphs to males. Can J Zool 86: 1131–1138.H. Van GossumT. RobbMR ForbesL. Rasmussen2008Female-limited polymorphism in a widespread damselfly: morph frequencies, male density, and phenotypic similarity of andromorphs to males.Can J Zool8611311138
  61. 61. Iserbyt A, Bots J, Van Dongen S, Ting JJ, Van Gossum H, et al. (2011) Frequency-dependent variation in mimetic fidelity in an intra-specific mimicry system. Proc Roy Soc B 278: 3116–3122.A. IserbytJ. BotsS. Van DongenJJ TingH. Van Gossum2011Frequency-dependent variation in mimetic fidelity in an intra-specific mimicry system.Proc Roy Soc B27831163122
  62. 62. Le Galliard JF, Fitze PS, Ferriere R, Clobert J (2005) Sex ratio bias, male aggression, and population collapse in lizards. Proc Natl Acad Sci U S A 102: 18231–18236.JF Le GalliardPS FitzeR. FerriereJ. Clobert2005Sex ratio bias, male aggression, and population collapse in lizards.Proc Natl Acad Sci U S A1021823118236
  63. 63. Kokko H, Rankin DJ (2006) Lonely hearts or sex in the city? Density-dependent effects in mating systems. Philos T Roy Soc B 361: 319–334.H. KokkoDJ Rankin2006Lonely hearts or sex in the city? Density-dependent effects in mating systems.Philos T Roy Soc B361319334
  64. 64. Xu MZ, Fincke OM (2011) Tests of the harassment-reduction function and frequency-dependent maintenance of a female-specific color polymorphism in a damselfly. Behav Ecol Sociobiol 65: 1215–1227.MZ XuOM Fincke2011Tests of the harassment-reduction function and frequency-dependent maintenance of a female-specific color polymorphism in a damselfly.Behav Ecol Sociobiol6512151227
  65. 65. Rolff J, Van de Meutter F, Stoks R (2004) Time constraints decouple age and size at maturity and physiological traits. Am Nat 164: 559–565.J. RolffF. Van de MeutterR. Stoks2004Time constraints decouple age and size at maturity and physiological traits.Am Nat164559565
  66. 66. Sugumaran H (2002) Comparative biochemistry of eumelanogenesis and the protective roles of phenoloxidase and melanin in insects. Pigm Cell Res 15: 2–9.H. Sugumaran2002Comparative biochemistry of eumelanogenesis and the protective roles of phenoloxidase and melanin in insects.Pigm Cell Res1529
  67. 67. Boughton RK, Joop G, Armitage SAO (2011) Outdoor immunology: methodological considerations for ecologists. Funct Ecol 25: 81–100.RK BoughtonG. JoopSAO Armitage2011Outdoor immunology: methodological considerations for ecologists.Funct Ecol2581100
  68. 68. González-Santoyo I, Córdoba-Aguilar A (2012) Phenoloxidase: a key component of the insect immune system. Entomol Exp Appl 142: 1–16.I. González-SantoyoA. Córdoba-Aguilar2012Phenoloxidase: a key component of the insect immune system.Entomol Exp Appl142116
  69. 69. Ziegler R, Van Antwerpen R (2006) Lipid uptake by insect oocytes. Insect Biochem Mol Biol 36: 264–272.R. ZieglerR. Van Antwerpen2006Lipid uptake by insect oocytes.Insect Biochem Mol Biol36264272
  70. 70. Lease HM, Wolf BO (2011) Lipid content of terrestrial arthropods in relation to body size, phylogeny, ontogeny and sex. Physiol Entomol 36: 29–38.HM LeaseBO Wolf2011Lipid content of terrestrial arthropods in relation to body size, phylogeny, ontogeny and sex.Physiol Entomol362938
  71. 71. Marden JH (2000) Variability in the size, composition, and function of insect flight muscles. Annu Rev Physiol 62: 157–178.JH Marden2000Variability in the size, composition, and function of insect flight muscles.Annu Rev Physiol62157178
  72. 72. Wheeler D (1996) The role of nourishment in oogenesis. Annu Rev Entomol 41: 407–431.D. Wheeler1996The role of nourishment in oogenesis.Annu Rev Entomol41407431
  73. 73. Marden JH (1989) Bodybuilding dragonflies - costs and benefits of maximizing flight-muscle. Physiol Zool 62: 505–521.JH Marden1989Bodybuilding dragonflies - costs and benefits of maximizing flight-muscle.Physiol Zool62505521
  74. 74. Córdoba-Aguilar A (2009) A female evolutionary response when survival is at risk: male harassment mediates early reallocation of resources to increase egg number and size. Behav Ecol Sociobiol 63: 751–763.A. Córdoba-Aguilar2009A female evolutionary response when survival is at risk: male harassment mediates early reallocation of resources to increase egg number and size.Behav Ecol Sociobiol63751763
  75. 75. Fielding DJ, Defoliart LS (2008) Discriminating tastes: self-selection of macronutrients in two populations of grasshoppers. Physiol Entomol 33: 264–273.DJ FieldingLS Defoliart2008Discriminating tastes: self-selection of macronutrients in two populations of grasshoppers.Physiol Entomol33264273
  76. 76. Karl I, Stoks R, De Block M, Janowitz SA, Fischer K (2011) Temperature extremes and butterfly fitness: conflicting evidence from life history and immune function. Glob Change Biol 17: 676–687.I. KarlR. StoksM. De BlockSA JanowitzK. Fischer2011Temperature extremes and butterfly fitness: conflicting evidence from life history and immune function.Glob Change Biol17676687
  77. 77. Stoks R, De Block M, Slos S, Van Doorslaer W, Rolff J (2006) Time constraints mediate predator-induced plasticity in immune function, condition, and life history. Ecology 87: 809–815.R. StoksM. De BlockS. SlosW. Van DoorslaerJ. Rolff2006Time constraints mediate predator-induced plasticity in immune function, condition, and life history.Ecology87809815
  78. 78. Bradford MM (1976) Rapid and sensitive method for quantitation of microgram quantities of protein utilizing principle of protein-dye binding. Anal Biochem 72: 248–254.MM Bradford1976Rapid and sensitive method for quantitation of microgram quantities of protein utilizing principle of protein-dye binding.Anal Biochem72248254
  79. 79. Bligh EG, Dyer WJ (1959) A rapid method of total lipid extraction and purification. Can J Biochem Physiol 37: 911–917.EG BlighWJ Dyer1959A rapid method of total lipid extraction and purification.Can J Biochem Physiol37911917
  80. 80. Lessells CM, Boag PT (1987) Unrepeatable repeatabilities: a common mistake. Auk 104: 116–121.CM LessellsPT Boag1987Unrepeatable repeatabilities: a common mistake.Auk104116121
  81. 81. Abramoff MD, Magelhaes PJ, Ram SJ (2004) Image processing with ImageJ. Biophoton Int 11: 36–42.MD AbramoffPJ MagelhaesSJ Ram2004Image processing with ImageJ.Biophoton Int113642
  82. 82. Koenig WD (1999) Spatial autocorrelation of ecological phenomena. Trends Ecol Evol 14: 22–26.WD Koenig1999Spatial autocorrelation of ecological phenomena.Trends Ecol Evol142226
  83. 83. Conrad KF, Willson KH, Whitfield K, Harvey IF, Thomas CJ, et al. (2002) Characteristics of dispersing Ischnura elegans and Coenagrion puella (Odonata): age, sex, size, morph and ectoparasitism. Ecography 25: 439–445.KF ConradKH WillsonK. WhitfieldIF HarveyCJ Thomas2002Characteristics of dispersing Ischnura elegans and Coenagrion puella (Odonata): age, sex, size, morph and ectoparasitism.Ecography25439445
  84. 84. Farrar DE, Glauber RR (1967) Multicollinearity in regression analysis - problem revisited. Rev Econ Stat 49: 92–107.DE FarrarRR Glauber1967Multicollinearity in regression analysis - problem revisited.Rev Econ Stat4992107
  85. 85. De'ath G, Fabricius KE (2000) Classification and regression trees: A powerful yet simple technique for ecological data analysis. Ecology 81: 3178–3192.G. De'athKE Fabricius2000Classification and regression trees: A powerful yet simple technique for ecological data analysis.Ecology8131783192
  86. 86. Zheng H, Chen L, Han X, Zhao X, Ma Y (2009) Classification and regression tree (CART) for analysis of soybean yield variability among fields in Northeast China: The importance of phosphorus application rates under drought conditions. Agric Ecosyst Environ 132: 98–105.H. ZhengL. ChenX. HanX. ZhaoY. Ma2009Classification and regression tree (CART) for analysis of soybean yield variability among fields in Northeast China: The importance of phosphorus application rates under drought conditions.Agric Ecosyst Environ13298105
  87. 87. Nakamoto A, Itabe S, Sato A, Kinjo K, Izawa M (2011) Geographical distribution pattern and interisland movements of Orii's flying fox in Okinawa Islands, the Ryukyu Archipelago, Japan. Popul Ecol 53: 241–252.A. NakamotoS. ItabeA. SatoK. KinjoM. Izawa2011Geographical distribution pattern and interisland movements of Orii's flying fox in Okinawa Islands, the Ryukyu Archipelago, Japan.Popul Ecol53241252
  88. 88. Li BB, Morris J, Martin EB (2002) Model selection for partial least squares regression. Chemometr Intell Lab 64: 79–89.BB LiJ. MorrisEB Martin2002Model selection for partial least squares regression.Chemometr Intell Lab647989
  89. 89. Hubert M, Vanden Branden K (2003) Robust methods for partial least squares regression. J Chemometr 17: 537–549.M. HubertK. Vanden Branden2003Robust methods for partial least squares regression.J Chemometr17537549
  90. 90. Mucklow PT, Vizoso DB, Jensen KH, Refardt D, Ebert D (2004) Variation in phenoloxidase activity and its relation to parasite resistance within and between populations of Daphnia magna. Proc Roy Soc B 271: 1175–1183.PT MucklowDB VizosoKH JensenD. RefardtD. Ebert2004Variation in phenoloxidase activity and its relation to parasite resistance within and between populations of Daphnia magna.Proc Roy Soc B27111751183
  91. 91. Holm S (1979) A simple sequentially rejective multiple test procedure. Scand J Stat 6: 65–70.S. Holm1979A simple sequentially rejective multiple test procedure.Scand J Stat66570
  92. 92. Olsson K, Agren J (2002) Latitudinal population differentiation in phenology, life history and flower morphology in the perennial herb Lythrum salicaria. J Evol Biol 15: 983–996.K. OlssonJ. Agren2002Latitudinal population differentiation in phenology, life history and flower morphology in the perennial herb Lythrum salicaria.J Evol Biol15983996
  93. 93. Andrés JA, Sánchez-Guillén RA, Cordero Rivera A (2002) Evolution of female colour polymorphism in damselflies: testing the hypotheses. Anim Behav 63: 677–685.JA AndrésRA Sánchez-GuillénA. Cordero Rivera2002Evolution of female colour polymorphism in damselflies: testing the hypotheses.Anim Behav63677685
  94. 94. Andrés JA, Sánchez-Guillén RA, Cordero Rivera A (2000) Molecular evidence for selection on female color polymorphism in the damselfly Ischnura graellsii. Evolution 54: 2156–2161.JA AndrésRA Sánchez-GuillénA. Cordero Rivera2000Molecular evidence for selection on female color polymorphism in the damselfly Ischnura graellsii.Evolution5421562161
  95. 95. Wong A, Smith ML, Forbes MRL (2003) Differentiation between subpopulations of a polychromatic damselfly with respect to morph frequencies, but not neutral genetic markers. Mol Ecol 12: 3505–3513.A. WongML SmithMRL Forbes2003Differentiation between subpopulations of a polychromatic damselfly with respect to morph frequencies, but not neutral genetic markers.Mol Ecol1235053513
  96. 96. Abbott JK, Bensch S, Gosden TP, Svensson EI (2008) Patterns of differentiation in a colour polymorphism and in neutral markers reveal rapid genetic changes in natural damselfly populations. Mol Ecol 17: 1597–1604.JK AbbottS. BenschTP GosdenEI Svensson2008Patterns of differentiation in a colour polymorphism and in neutral markers reveal rapid genetic changes in natural damselfly populations.Mol Ecol1715971604
  97. 97. Calsbeek R, Bonvini L, Cox RM (2010) Geographic variation, frequency-dependent selection, and the maintenance of a female-limited polymorphism. Evolution 64: 116–125.R. CalsbeekL. BonviniRM Cox2010Geographic variation, frequency-dependent selection, and the maintenance of a female-limited polymorphism.Evolution64116125
  98. 98. Cordero Rivera A, Egido JFP (1998) Mating frequency, population density and female polychromatism in the damselfly Ischnura graellsii: an analysis of four natural populations. Etología 6: 71–78.A. Cordero RiveraJFP Egido1998Mating frequency, population density and female polychromatism in the damselfly Ischnura graellsii: an analysis of four natural populations.Etología67178
  99. 99. Van Gossum H, Sherratt TN (2008) A dynamical model of sexual harassment in damselflies and its implications for female-limited polymorphism. Ecol Model 210: 212–220.H. Van GossumTN Sherratt2008A dynamical model of sexual harassment in damselflies and its implications for female-limited polymorphism.Ecol Model210212220
  100. 100. Iserbyt A, Van Gossum H (2009) Unexpected absence of behavioural differences between female damselfly colour morphs. Anim Behav 78: 1463–1469.A. IserbytH. Van Gossum2009Unexpected absence of behavioural differences between female damselfly colour morphs.Anim Behav7814631469
  101. 101. Bouton N, Iserbyt A, Van Gossum H (2011) Thermal plasticity in life-history traits in the polymorphic blue-tailed damselfly, Ischnura elegans: No differences between female morphs. J Insect Sci 11: 112.N. BoutonA. IserbytH. Van Gossum2011Thermal plasticity in life-history traits in the polymorphic blue-tailed damselfly, Ischnura elegans: No differences between female morphs.J Insect Sci11112Insectscience website. Available: http://www.insectscience.org/11.112/. Accessed 2012 Feb 3. Insectscience website. Available: http://www.insectscience.org/11.112/. Accessed 2012 Feb 3.
  102. 102. Siva-Jothy MT, Thompson JJW (2002) Short-term nutrient deprivation affects immune function. Physiol Entomol 27: 206–212.MT Siva-JothyJJW Thompson2002Short-term nutrient deprivation affects immune function.Physiol Entomol27206212
  103. 103. Schmid-Hempel P (2005) Natural insect host-parasite systems show immune priming and specificity: puzzles to be solved. Bioessays 27: 1026–1034.P. Schmid-Hempel2005Natural insect host-parasite systems show immune priming and specificity: puzzles to be solved.Bioessays2710261034
  104. 104. De Block M, Stoks R (2008) Short-term larval food stress and associated compensatory growth reduce adult immune function in a damselfly. Ecol Entomol 33: 796–801.M. De BlockR. Stoks2008Short-term larval food stress and associated compensatory growth reduce adult immune function in a damselfly.Ecol Entomol33796801
  105. 105. Lochmiller RL, Deerenberg C (2000) Trade-offs in evolutionary immunology: just what is the cost of immunity? Oikos 88: 87–98.RL LochmillerC. Deerenberg2000Trade-offs in evolutionary immunology: just what is the cost of immunity?Oikos888798
  106. 106. King EG, Roff DA, Fairbairn DJ (2011) Trade-off acquisition and allocation in Gryllus firmus: a test of the Y model. J Evol Biol 24: 256–264.EG KingDA RoffDJ Fairbairn2011Trade-off acquisition and allocation in Gryllus firmus: a test of the Y model.J Evol Biol24256264
  107. 107. Ohsaki N (2005) A common mechanism explaining the evolution of female-limited and both-sex Batesian mimicry in butterflies. J Anim Ecol 74: 728–734.N. Ohsaki2005A common mechanism explaining the evolution of female-limited and both-sex Batesian mimicry in butterflies.J Anim Ecol74728734
  108. 108. Kunte K (2009) Female-limited mimetic polymorphism: a review of theories and a critique of sexual selection as balancing selection. Anim Behav 78: 1029–1036.K. Kunte2009Female-limited mimetic polymorphism: a review of theories and a critique of sexual selection as balancing selection.Anim Behav7810291036
  109. 109. Berwaerts K, Van Dyck H (2004) Take-off performance under optimal and suboptimal thermal conditions in the butterfly Pararge aegeria. Oecologia 141: 536–545.K. BerwaertsH. Van Dyck2004Take-off performance under optimal and suboptimal thermal conditions in the butterfly Pararge aegeria.Oecologia141536545
  110. 110. Gosden TP, Svensson EI (2009) Density-dependent male mating harassment, female resistance, and male mimicry. Am Nat 173: 709–721.TP GosdenEI Svensson2009Density-dependent male mating harassment, female resistance, and male mimicry.Am Nat173709721
  111. 111. Forbes MRL, Schalk G, Miller JG, Richardson JML (1997) Male-female morph interactions in the damselfly Nehallenia irene (Hagen). Can J Zool 75: 253–260.MRL ForbesG. SchalkJG MillerJML Richardson1997Male-female morph interactions in the damselfly Nehallenia irene (Hagen).Can J Zool75253260
  112. 112. Van Gossum H, Stoks R, De Bruyn L (2001) Frequency-dependent male mate harassment and intra-specific variation in its avoidance by females of the damselfly Ischnura elegans. Behav Ecol Sociobiol 51: 69–75.H. Van GossumR. StoksL. De Bruyn2001Frequency-dependent male mate harassment and intra-specific variation in its avoidance by females of the damselfly Ischnura elegans.Behav Ecol Sociobiol516975
  113. 113. Locklin JL, Vodopich DS (2010) Patterns of gregarine parasitism in dragonflies: host, habitat, and seasonality. Parasitol Res 107: 75–87.JL LocklinDS Vodopich2010Patterns of gregarine parasitism in dragonflies: host, habitat, and seasonality.Parasitol Res1077587
  114. 114. McKinnon JS, Pierotti MER (2010) Colour polymorphism and correlated characters: genetic mechanisms and evolution. Mol Ecol 19: 5101–5125.JS McKinnonMER Pierotti2010Colour polymorphism and correlated characters: genetic mechanisms and evolution.Mol Ecol1951015125
  115. 115. Iserbyt A (2012) Maintenance of intra-sexual polymorphism in female damselflies. PhD thesis in press. Antwerp: Antwerp University. A. Iserbyt2012Maintenance of intra-sexual polymorphism in female damselflies. PhD thesis in pressAntwerpAntwerp University