Body size is the most important feature of all organisms and in this study is represented by the variables “body length” and “body width”. Bergmann’s Rule is a hypothesis concerning body size increase with latitude for homeotherms, whereas geographic version of Rensch’s Rule predicts greater body size variation with latitude for males. The trend predicted by the first rule is confirmed here for a group of ectotherms, terrestrial isopods, while that predicted by the second was not found to apply in these organisms. Several hypotheses have been proposed to explain body size variation in terms of environmental factors and interspecific interactions underlying latitudinal gradients. Thus, we used our results to test certain hypotheses proposed to account for Bergmann’s Rule. The Island Rule examines body size as it is affected by spatial parameters and reflects a complex pattern from gigantism in taxa with small sized mainland relatives, to dwarfism in taxa with large mainland relatives. Our results are not conclusive regarding this rule, but are indicative of trends that need to be examined in the future.
The most important feature of all organisms through which physiological
, ecological and life-history traits are reflected, is body size. Although size is treated as a fixed variable, it is continuously altered by a variety of factors, and final adult size (maximum size) is depended on interactions among time constraints, resource allocation to growth and/or reproduction, mortality, physiological costs, ageing and food quality De Block et al. (2008). There are several variables that can represent ‘body size’. Mass is regarded the best
, but it varies on a seasonal or even daily basis. Body length, along with body width, is also an attractive index because it measures the whole animal Meiri (2007). Although maximum size may be sensitive to sampling and it is expected that mainland areas are more likely to include the largest individuals due to the existence of a greater number of populations there, it is used in this study because it follows maturation of terrestrial isopods and it is indicative of developmental rate, as well as to the responses of individuals to environmental thermal variation Folguera et al. (2009).
Latitudinal patterns with body size and species richness
The observation of a latitudinal pattern in body size predates Darwin’s evolutionary theory and is known as Bergmann’s Rule. It relates temperature with body size -although latitude is often used as a proxy to temperature Ashton (2002). According to the rule, there is an increase in body size towards cold climates Bergmann (1847). Bergmann’s rule was initially referring to warm-blooded vertebrate species, thus, to endotherms, at an interspecific level Watt (2010). Bergmann’s rule does not make predictions for ectotherms Entling et al. (2010). Even though evidence for the dominance of Bergmann’s clines in ectotherms is controversial Adams et al. (2007), similar gradients, but also their converse, have been observed in ectotherms at the interspecific level Olalla-Tarraga et al. (2007). An 80% of the ectotherms tested grow larger at lower temperatures - a trend also known as temperature-size rule Blanckenhorn et al. (2004); Ho et al. (2010)
The mechanism behind Bergmann’s rule is a thermoregulatory one, on the basis that an increase in size involves a more rapid increase of body volume compared to the increase of its surface area. Due to the fact that heat loss is related to the surface of a homeotherm, while heat production to its volume, larger animals will tend to produce more and loose less heat, an advantageous fact in cold climates. In addition, increase of body size with resource availability, rather than with decreasing temperature Rosenzweig (1968), may be another mechanism that explains Bergmann’s rule. Increasing seasonality and lower predictability of environments from the tropics to the poles were thought to select for large body size, because large animals are more resistant to cold and starvation stress Olson et al. (2009).
The mechanism of heat conservation is not applicable to ectotherms, especially to those that are small, aquatic or whose body temperature largely fluctuates with ambient temperature Ho et al. (2010) or to those that do not thermoregulate behaviorally Adams et al. (2007). Many ectotherms, though, can regulate their internal temperature by modifying their behavior or by creating heat through muscular activity Watt et al. (2010). So, Heat Conservation Hypothesis, proposed by Bergmann for endotherms, could be extended to terrestrial ectotherms that can control body temperature through physiological and/or behavioral adjustments Olalla-Tarraga et al. (2007). Along with this hypothesis, there are two other interspecific mechanisms that predict increase of body size with latitude. Accelerated Maturation Hypothesis predicts a smaller adult size in warm environments Walters et al. (2006); Entling et al. (2010). Starvation Resistance Hypothesis states that an increase in body size encourages starvation resistance and this trend is regarded more important in cold, seasonal environments Entling et al. (2010). In addition, Heat Dissipation Hypothesis is based on the conviction that average body size is better described by variables reflecting both environmental parameters, temperature and moisture. Moreover, covariation of size with these variables should be stronger for large-bodied than small-bodied species Blackburn et al. (2004). There is also another unifying physiological mechanism beyond Bergmann’s rule that extends to ectotherms and tries to explain how low temperatures (the factor constraining growth) result in larger body sizes. Since the rate of growth is mainly affected by protein synthesis and secondarily by temperature, whereas development is highly dependent on temperature, organisms are expected to reach maturity more rapidly at higher temperatures while time of growth increases less rapidly, leading to smaller final size Van der Have et al. (1996). These physiological constraints apply to all parts of the body, such as eggs, sperm or single cells. Another mechanism explaining the smaller size of eggs and cells at higher temperatures is that, while oxygen diffusion depends weakly on temperature, consumption of oxygen depends strongly on it, so that large cells may suffer from hypoxia at high temperatures Woods (1999). This hypothesis provides a simple mechanism for body size clines in ectotherms, since the total size of an organism is largely the sum of its cells Van Voorhies (1996).
Another pattern that can be associated with clinal variation in body size is Rensch’s rule that predicts, under the condition of the existence of geographic variation in body size for a species, geographic variation in sexual size dimorphism Pyron et al. (2007) where latitudinal clines in males are steeper than those in females Blanckenhorn et al. (2006). According to another hypothesis deriving from the positive relationship between mean and variance, whichever sex is larger should be more variable and display a steeper latitudinal slope (Larger = Steeper Hypothesis). This rule is based on the observation that male body size varies more than that of females among related species, in a way that male-biased sexual size dimorphism increases and female decreases with body size Fairbairn (1997). A possible mechanism behind this pattern may be sexual selection for large male size along with a high genetic correlation in body size between sexes, and results in greater among-population variation in male body size Blanckenhorn et al. (2006).
Another strong pattern is the decrease in body size with increasing richness Olson et al. (2009), a hypothesis consistent with an increase of body size with latitude Roy et al. (2001), i.e., with ‘Bergmann’s rule’-like trends. Species richness is related to latitude in what probably is the oldest ecological pattern observed, namely that the tropics (low latitudes) hold more species than higher latitudes Turner (2004). Although latitude cannot be a determinant of species richness per se, its effect is mediated through the systematic spatial variation of a variety of other factors, such as climate etc. Many explanations of this effect have been proposed, based inter alia on chance, historical perturbation, environmental stability, habitat heterogeneity, productivity and interspecific interactions Gaston (2000).
Body size and insularity
Foster (1964) found differences in body size between different groups of terrestrial mammals inhabiting islands and their mainland congeners. He showed that rodents evolve large size on islands whereas carnivores and artiodactyls usually grow smaller. It is a pattern that describes a clear tendency for different evolutionary trends and selective pressures on islands Lomolino et al. (2006). Lomolino (1985) extended this trend by proposing that many insular forms become dwarfs or giants in comparison with their mainland relatives and predicted a trend that promotes gigantism in the smaller mainland species and dwarfism in the larger ones. The driving forces that promote gigantism are character release in response to reduced interspecific competition or predation and a force promoting dwarfism is resource limitation Meiri et al. (2005). Dwarfism can improve fitness in the case that the resources available from the reduction of size are channeled into reproduction Palmer (2002). High population densities may promote gigantism because large individuals can more effectively deal with intraspecific competition. Different diet preferences among animals may lead to different responses to the insular environment Meiri (2007). A general factor determining the evolution of body size is the limited area of islands. Small area weakens competition and predation pressure through impoverished communities, and diminishes the abundance of resources Schillaci et al. (2009). Isolation too, can affect size evolution, because large animals have a greater chance of colonizing very isolated islands. Although the effect of phylogeny has not yet been explicitly examined, the Island Rule may hold within families or orders Meiri et al. (2009). In order to avoid effects of phylogeny as much as possible, we restricted analyses to species belonging to the same genus and/or family.
The overall aims of this study were to collate available data on morphological characteristics of terrestrial isopods in order to test: (1) whether there is a consistent trend between body size and latitude, as well as between those two parameters and species richness; (2) whether the ‘island rule’ applies to ectotherms at an interspecific level; (3) which are the main ecological factors driving body size variation; and (4) which hypotheses can explain body size variation in our study.
2. MATERIALS AND METHODS
Based on the taxonomy of the world catalog of terrestrial isopods published by Schmalfuss (2003, 2004), we collected species records from papers, books and reviews that contained information on distribution and maximum body size of males and females. Most often, authors give maximal body lengths, but in many cases maximal body widths are also reported. In cases where more than one measurement was provided in different papers, we recorded only the largest. Maximal length and width values are believed to be a reasonable indicator of size, and according to Bertalanffy’s growth function, it is the maximum attainable body size that is mostly influenced by thermal constraints Angilletta et al. (2004). Moreover, data on mass of isopod species are not available in the literature. Male length ranged from 1.5 to 32 mm, female length from 1.5 to 30 mm, male width from 0.4 to 21mm, and female width from 0.4 to 14 mm. In addition, we recorded separately species whose distribution is confined on islands and species with strictly mainland ranges. In total, we collected data on 231 island species, 265 mainland and 109 species with mixed distribution. Overall, 30 families were represented in our data set. All analyses were performed at the interspecific level.
We collected data on three ecological factors that are possibly linked to the size of terrestrial isopods: latitude, climate and distribution. Latitude was recorded as the mean latitude of each species overall distribution (midpoint approach), see Blackburn et al. (2004). Latitudes south of the equator were transformed to positive values. Climate records were based on the Köppen – Geiger climate classification Peel et al. (2007) that incorporates, along with temperature, data on global long-term monthly precipitation. According to this system, there are 30 climatic types, divided into 3 tropical (Af, Am and Aw), 4 arid (BWh, BWk, BSh and BSk), 9 temperate (Csa, Csb, Csc, Cfa, Cfb, Cfc, Cwa, Cwb and Cwc), 12 cold (Dsa, Dsb, Dsc, Dsd, Dfa, Dfb, Dfc, Dfd, Dwa, Dwb, Dwc and Dwd) and 2 polar (ET and EF) subtypes. The set of locations defined as having a ‘B’-type climate is based on a combination of mean annual precipitation and mean annual temperature, while all other sets are mutually exclusive and are based only on temperature criteria. We created, thus, two variables describing the climate. The first, climate1, had five levels (A=tropical, B=arid, C=temperate, D=cold and E=polar), the second, climate2, included the second characteristic of the 30 climate types given by the Köppen – Geiger classification. The variable concerning the distribution was divided into three categories: ‘mainland’, ‘island’, and ‘both’. The island surface area of all islands with ‘island’ species was recorded as found in Wikipedia (http://wikipedia.org).
We analyzed the applicability of Bergmann’s rule by using bootstrap least-squares regressions and Pearson product-moment correlations. In the first place, we examined the variation of body size with the increase of the absolute values of latitude and, next, we performed the same analysis for species belonging to the same climate, as defined by climate1. Finally, we tested body size variation in relation to latitude for isopods belonging to different climates, as defined by climate2, searching for indirect relationships of body size with precipitation and temperature. For this reason, we performed Student tests for the comparison of mean body size values for species in climates with dry summer and those living in climates having no dry season, as well as Fischer and Mann-Whitney tests to compare the variances and the medians of these samples. We compared species in deserts, rainforests and savannahs, using ANOVA and Multiple Range tests, in order to compare the means or any pair of means for any size parameter. The creation of two different groups was based on the fact that species residing in climates with dry summer and a dry season were much more numerous in relation to those belonging to the climates of the second group, therefore the elimination of the species below and above the mean value for each size parameter would not leave the mean invariable.
We calculated the mean body length and width for species belonging to each region and plotted frequency histograms to define species numbers below and above the mean value, in order to see whether small-bodied isopods are over-represented in species-rich regions (tropics, temperate climates).
We applied regressions of body size parameters on latitude for both sexes, in order to conclude if the slope of the body size on latitude would be steeper for males or females, so as to test for Rensch’s. This analysis was performed separately for different phylogenetic groups, namely, Ligiidae, Synocheta and Crinocheta, for which we had abundant species numbers.
To test whether the Island Rule was valid for either body length or body width, we extracted data on body size of ‘island’ and ‘mainland’ species. In order to avoid using phylogenetically non-independent data Meiri et al. (2009), we used 230 pairs of mainland-island species, ordered from smaller to bigger, belonging to either the same genus or family, and applied regressions of ‘island’ on ‘mainland’ species for body length and width of both sexes. We also applied regressions for 15 pairs belonging to the same family and 26 belonging to the same genus residing in the same climatic zone. In cases where we had more than one mainland species belonging to the same genus or family with the respective insular one, or vice versa, we calculated the mean value of the species’ sizes and excluded the values above and below the mean that differed approximately the same, in order to avoid changing the initial mean value before the extraction. After that, we ranged the values for both mainland and island species from the smallest to the largest to form the pairs used in the regressions. We then tested the variation in body size of island species with island surface. Due to the fact that large islands are ‘mainland-like’ in terms of predators and competitors, we tested size variation in a consecutive order of decreasing island size classes, namely for islands below 50,000 km2, below 10,000 km2, and below 5,000 km2. The selection of the largest threshold follows from Meiri et al. (2006), who found that similar patterns of size evolution are obtained when the area of the largest islands included is less that 50,000 km2.
Finally, we used a model-selection approach to determine the variables [distribution, climatic variables and/or taxonomic effect (Family)] that best explain geographical variation in body size among species using the log10 maximum body length and width as response variables. Due to high correlation between length and width for males and females, we performed four separate modeling exercises, one for each variable. We started with a full model to arrive at an adequate one. The removal of predictor terms was based on the maximum decrease in AIC and, in the end, the overall increase in model fit for the removal of each remaining term. We stopped removing terms when no further deletion of a term produced a decrease in AIC. In each step we performed a Shapiro test to control the normality of residuals and confirm the good adjustment of the model to our data.
All numeric variables were log transformed when they weren’t normally distributed. R.2.10.1 and Statgraphics Plus 5.0 were used in all analyses.
Bergmann’s and Rensch’s rules
Maximum body length and width of terrestrial isopods increases with the increase of absolute values of latitude (Table 1). So, there is an increase in body size when departing from the Equator in both the North and South hemispheres, suggesting a pattern consistent with Bergmann’s rule. Moreover, the slope of the regression of body size on latitude is steeper for females, for all species belonging to Ligiidae, as well as for those belonging to Synocheta and Crinocheta. This results in greater among-species variation in female than male body size. So, we found no evidence in favor of Rensch’s rule.