Geographic phenotype variation in neotropical South American grasshoppers (Acrididae: Orthoptera): patterns along environmental gradients

Abstract

The understanding of phenotypic variability across a geographic space constitutes a central issue in present evolutionary biology. As a rule of thumb, body size and trait variation mirror the adaptation to the environment within a species range as a result of adaptation to local environments, or to neutral drift, or both. This study examines phenotypic variability, static allometry, and sexual size dimorphism (SSD) along environmental gradients in 3 South American grasshoppers with distinct ecological requirements: the terrestrial Dichroplus elongatus (Giglio-Tos) (winged) and Dichroplus vittatus (Bruner) (wing dimorphic), and the semiaquatic Cornops aquaticum (Bruner). Dichroplus elongatus follows Bergmann’s rule and a converse Allenian pattern, likely driven by developmental rates and biomechanical compensation in cooler environments. In D. vittatus, body size and scaled appendages (following Allen’s rule) are strongly influenced by precipitation and thermal range, showing sex-specific responses where females prioritize fecundity. Cornops aquaticum, which displays a converse Bergmann pattern, does not show any evidence for Allen’s rule or its converse, suggesting semiaquatic habitat shifts toward hydrodynamic constraints. SSD was female-biased in all species. However, static allometric patterns critically moulded the phenotype, adapting proportional investment in locomotor versus reproductive structures in a habitat-specific manner. Thoracic dimensions in D. vittatus followed the converse of Rensch’s rule due to high female sensitivity to environmental conditions, whereas the more stable semiaquatic habitat of C. aquaticum reduced sex-specific climatic responses. These results underscore that orthopteran phenotype evolution is a complex interaction between habitat use (terrestrial vs semi-aquatic), life history strategies, dispersal modes, and sex-specific selective pressures.

Introduction

The understanding of phenotypic variability across a geographic space constitutes a central issue in evolutionary biology. Body size may vary among geographic localities within a species range as a result of adaptation to local environments or neutral drift, with forces often acting synergistically to shape the phenotype (Cadena et al. 2012, Sun et al. 2013, Lee et al. 2016, Engen and Sæther 2017). Moreover, this variation may be a result of phenotypic plasticity (Berner et al. 2004). Insect species with widespread geographic ranges are helpful models to examine the phenotype variation patterns acting at the intraspecific level by assessing the relative impact of environmental variation on population differentiation (Bidau and Marti 2008, Oyedokun and Omoloye 2019, Juhász et al. 2020).

Phenotypic variation is also shaped by trait allometry, which defines the scaling relationship between a morphological trait and the body size of an organism. Scaling among individuals of a population at the same developmental stage (static allometry) allows researchers to explore how developmental processes relate to the evolution of phenotypic diversity within species (Huxley 1932, Gould 1966, Shingleton et al. 2007, Voje et al. 2021).

Morphological differences between males and females are an important source of phenotypic variation within species. Sexual size dimorphism (SSD), the difference in body size between sexes (Stillwell et al. 2010), exemplifies this variation. In all the species analyzed here, females are larger than males, representing the general trend within Orthoptera (Blanckenhorn et al. 2006, Hochkirch and Gröning 2008, Whitman 2008). A generally accepted hypothesis explaining SSD is that male and female body size may be shaped by differing selective pressures; this is referred to as the “differential equilibrium model” (Slatkin 1984, Fairbairn 1997, Blanckenhorn 2000). In female-biased SSD, the differential equilibrium hypothesis suggests that body size results from a balance between selection favoring larger size (female fecundity and male sexual selection) and negative viability selection on body size or extended development (Blanckenhorn 2000, Blanckenhorn et al. 2006, Stillwell et al. 2007, Hochkrich and Gröning 2008). Typically, geographic variation in adult body size and correlated morphological traits is not random, but results from clinal patterns of differentiation with respect to geographic (essentially latitude and altitude) and/or climatic variables (such as temperature and rainfall); henceforth, they may suggest environmental restrictions and probable local adaptations (Endler 1977, Stearns 1992, Berner and Blanckenhorn 2006, Huizenga et al. 2008, Whitman 2008, Savolainen et al. 2013).

Phenotypic clinal patterns, when associated with environmental gradients, are often described in terms of ecogeographical rules. Bergmann’s rule describes the inverse correlation between temperature and body size in nature; Allen’s rule states that animals have shorter appendages in colder climates, and Rensch’s rule states that sexual dimorphism tends to increase with body size if males are the largest sex, but tends to decrease if females are the largest (Allen 1877, Rensch 1959, Fairbairn 1997, Whitman 2008). These rules were first established for endotherms, although at present they may also be extended to ectotherms (Blanckenhorn and Demont 2004, Bidau and Marti 2008, Whitman 2008, Shelomi 2012, Romero et al. 2014, García-Navas et al. 2017, Shelomi and Zeuss 2017, Bidau and Martinez 2018, Rosetti et al. 2026). Allen’s and Bergmann’s rules were originally explained in relation to thermoregulation in birds and mammals; however, in orthopterans, body size patterns across geographic clines are better explained by the trade-offs between growing season length and developmental rate (Telfer and Hassall 1999, Chown and Gaston 2010, Shelomi 2012). At the intraspecific level, Rensch’s rule can be explained as variations in SSD among populations within the same species. This variation often drives clinal latitudinal or altitudinal patterns in SSD attributed to both adaptive and phenotypic plasticity scenarios (Blanckenhorn and Demont 2004, Fairbairn 2005, Teder and Tammaru 2005, Blanckenhorn et al. 2006, Stillwell et al. 2007).

Considerable researches have focused to analyze the relative importance of neutral and adaptive scenarios to explain observable variation. Compelling evidence of signatures of natural selection constitutes the occurrence of geographic patterns that appear independently in different geographic region (Endler 1977). In recent decades, many advances have been made to distinguish the levels of population differentiation at potentially adaptive traits (such as morphometric traits) with respect to neutral DNA marker variation (Roff and Mousseau 2005, Leinonen et al. 2008, Whitlock 2008, Ayala et al. 2011, Ortego et al. 2012, Rosetti and Remis 2013, Blanckenhorn et al. 2018, Colombo et al. 2021).

Recently, the analyses of phenotypes in populations offer a more effective resolution by considering both the potential effects of environmental factors and the spatial arrangement of samples and populations. Furthermore, the use of Linear Models (LM) on sets of quantitative variables through the landscape has improved the detection and description of phenotype variation patterns within and among groups or populations, and the unveiled evidence may help predict how environmental change might influence the future species distribution. The use of LM as an analysis tool provides a robust framework to quantify subtle phenotypic differences between populations by using multivariate linear data. This approach allows for the analysis of allometric effects—where proportional changes occur as a function of size—taken apart from variations driven by localized environmental pressures (Adams et al. 2013, Klingenberg 2016). For orthopterans, this distinction is critical, as the relative proportions of wings or hind leg segments (femur and tibia) may reflect dispersal ability or predator escape strategies that are independent of the overall body size clines predicted by environmental gradients (Tatsuta et al. 2004).

Insect species, especially those with a wider geographic distribution, usually exhibit considerable intraspecific molecular and body size variation, providing an appropriate model to study the relative contribution of stochastic and adaptive processes on population divergence. Here, we aim to analyze the phenotypic variability along geographic and climatic transects in a few widely distributed South American grasshopper species in both terrestrial and semiaquatic habitats: Dichroplus elongatus (Giglio-Tos), D. vittatus (Bruner), and Cornops aquaticum (Bruner).

Dichroplus elongatus is a ground-dwelling grasshopper with a wide geographic distribution, which has been reported as a dominant or codominant species in most grasshoppers’ outbreaks occurred in the last years (Mariottini et al. 2011). Dichroplus vittatus is a grasshopper that can display short-winged (brachypterous) or fully-winged (macropterous) morphs in the field, with the former being the most frequent (Turk and Barrera 1979, Cigliano and Otte 2003). In central-west Argentina, these species inhabit a transitional environmental gradient—shifting from the arid Monte scrubland to the more seasonal Espinal—characterized by significant fluctuations in thermal range and water availability (Rosetti et al. 2022). Cornops aquaticum is a semiaquatic Neotropical grasshopper that lives, feeds, and reproduces on plants of the family Pontederiaceae, the so called “water-hyacinths” (Franceschini et al. 2023). These plants are aquatic, and their floating mats are dragged downstream by the flow of the streams and rivers they live upon; Pontederia crassipes is, furthermore, an invading plant and has reached most tropical and subtropical freshwater environments, covering most bodies of water worldwide except, in South America, where they are native.

The analyses of these 3 species with diverse ecological requirements and dispersal strategies provide a unique comparative framework to explore how environmental gradients shape morphology. Therefore, we propose a major quantitative analysis of phenotype variation in widely distributed Neotropical grasshoppers with diverse biologic and demographic features. This variation is examined through the lens of ecological rules (Bergmann’s, Rensch’s, and Allen’s rules) and a suite of bioclimatic variables related to season length: temperature and precipitation. We attempted to contribute to the understanding about how environmental heterogeneity affects the phenotypic patterns of species with different adaptive strategies. Moreover, this study is particularly valuable as it highlights aspects of phenotypic variation in grasshoppers that have received little attention in Argentina.

Materials and Methods

Species and Measurements

The biological models whose phenotypic response to environmental patterns was assessed were 3 species native to South America, namely C. aquaticum, D. elongatus, and D. vittatus. The analyzed populations were sampled across multiple geographic and climatic gradients (Supplementary Table S1).

Previously published data were used here to fulfill the objectives of this paper. First, we considered available data from 292 individuals from 12 populations of D. elongatus and 168 individuals from 9 populations of D. vittatus distributed along latitudinal (700 km), longitudinal (165 km), and altitudinal (934 m) ranges in central-west Argentina (Rosetti and Remis 2018, Rosetti et al. 2026) (see Supplementary Table S1). The considered measurements involve: fastigium-third coxa length (Total), taken from the fastigium to the third femur-third coxa articulation, third femur length (Femur), third tibia length (Tibia), thorax length (Thorax L), and tegmen length (Tegmen). While D. vittatus is wing-dimorphic, the macropterous individuals are less common and may be absent in some populations. For this reason, in this work, we only analyzed brachypterous individuals, as they provided an adequate sample size for analysis.

Moreover, we also considered available data from 426 individuals of C. aquaticum from 13 populations along both the Paraná and Uruguay River Basins in South America (Colombo et al. 2021). In addition to the previously mentioned traits, we also measured thorax height (Thorax H) in this species.

Characterization of the Environmental Space

We retrieved data for 12 climatic variables relevant to grasshoppers’ geographic distribution (Bai et al. 2016, Li et al. 2024). These variables were obtained from the WorldClim v2.1 database (Fick and Hijmans 2017) at a resolution of 10 min. This database contains information about Annual Mean Temperature (BIO1), Mean Diurnal Range (MDR; BIO2), Temperature Annual Range (TAR; BIO7), Mean Temperature of Wettest Quarter (BIO8), Mean Temperature of Driest Quarter (MTDQ; BIO9), Mean Temperature of Warmest Quarter (BIO10), Mean Temperature of Coldest Quarter (BIO11), Annual Precipitation (BIO12), Precipitation of Wettest Quarter (BIO16), Precipitation of Driest Quarter (PPDQ; BIO 17), Precipitation of Warmest Quarter (BIO 18), Precipitation of Coldest Quarter (BIO19) (all temperatures in °C and all precipitations in mm).

As a preliminary approach for the study of the relationship between body size and environmental conditions in D. vittatus and C. aquaticum, a subset of those climatic variables was selected for each species. That subset included temperature and precipitation variables that were not highly correlated and had variance inflation factors (VIFs; Belsley et al. 1980) <7. VIFs were estimated with the car package (Fox and Weisberg 2020) in R 4.1.3 (R Core Team 2022). Three variables were chosen for D. vittatus and 5 for C. aquaticum. These variables and their respective VIF are shown in Supplementary Table S2.

Relationships Between Body Size and Shape With Respect to Climatic Variables

The following analyses were conducted on C. aquaticum and D. vittatus. For comparative purposes, we used previously conducted analyses for D. elongatus (Rosetti et al. 2026).

We fitted a linear mixed model (LMM) to predict body size, using total length (as a proxy for body size) as the response variable, climatic variables and sex as fixed effects, and population as a random effect. Since different environmental variables could be relevant for every species and traits, we used Analysis of Variance (ANOVA) and Akaike Information Criterion (AIC) to select the better combination of variables, that is, the simplest model that better fit the data. All LMM tests were performed with the lme4 (Bates et al. 2015) and lmerTest (Kuznetsova et al. 2017) packages, and model assumptions were tested with DHARMA (Hartig 2022) in R 4.1.3.

To analyze the shape of static allometry, we also used LMM to evaluate the relative size of the following measured morphometric traits: Femur, Tibia, Thorax, Tegmen lengths, and Thorax height with respect to total length as a proxy of total size. The shape of allometries was modeled using the power-law relationship, which can be applied to traits that scale linearly on a log–log scale. Therefore, the generalized formula was:

$$\log \left(\mathrm{trait}\right)\mathrm{ }\sim \mathrm{ }\left(1\right|\mathrm{POP})\mathrm{ }+\mathrm{ }\mathrm{SEX}\mathrm{ }+\mathrm{ }(\mathrm{chosen}\mathrm{ }\mathrm{bioclimatic}\mathrm{ }\mathrm{variables})\mathrm{ }+\mathrm{ }\log \mathrm{ }(\mathrm{total}\mathrm{ }\mathrm{length})$$

Therefore, our analysis of static allometry investigates how the magnitude of a particular trait scales with body size. Slopes exceeding unity (b > 1) demonstrate positive allometry or hyperallometry (disproportionate increase in the trait), while slopes below unity (b < 1) indicate negative allometry or hypoallometry (disproportionate decrease relative to body size). Slopes that do not deviate significantly from 1 would indicate that a particular trait grows proportionally with body size (isometry) (Huxley 1932, Gould 1966).

Further, in order to assess whether each factor affected both sexes differentially, we carried out 2 more LMMs, one calculating the SEX:bioclimatic variables interaction and the other to evaluate the sex:total length relationship.

Therefore, the models could be written as:

Total length ∼ (1|POP) + SEX × (chosen bioclimatic ­variables) for total length and as:

log(trait) ∼ (1|POP) + SEX × (chosen bioclimatic variables) + log(total length) and

log(trait) ∼ (1|POP) + SEX × log(total length) + (chosen bioclimatic variables) for the remaining traits.

That way, the models that include the interactions SEX:bioclimatic variables compare the slope for each sex to assess whether the environmental conditions affect each sex differentially. Lastly, the models including the SEX:log (total length) investigate whether static allometric trajectories differ between males and females while controlling for climatic variation.

Again, the best model for each trait was selected using ANOVA and AIC. For the random effect, the proportion of variance that is attributable to the grouping structure was evaluated with the ICC (Intraclass Correlation Coefficient) and calculated by dividing the variance of the random effect (POP) by the total variance (the sum of the random effect variance and the residual variance). A higher ICC means that groups are quite different among them, while the individuals inside the sampling site are similar to each other.

For natural populations of the semiaquatic C. aquaticum, we examined the variables most likely to predict body size (Supplementary Table S2). The best model was selected based on the lowest AIC or prioritizing the simplest model when there were no significant differences. First, models for all the different combinations of bioclimatic variables were evaluated for each trait, and the ones with the lowest AIC were selected (Supplementary Tables S3–S8). Then, additive vs interaction models were evaluated, and since differences between AIC were negligible (Supplementary Table S9), the additive model was considered the one that better explained the results for all the traits. We evaluated multicollinearity among the bioclimatic variables using the VIF (Supplementary Table S12).

LMM models for brachypterous D. vittatus individuals for each trait, testing different combinations of the 3 bioclimatic variables previously selected (Supplementary Table S2), were evaluated with and without interaction for their factors, and the best model was selected as mentioned above. AIC results for all the models are shown in Supplementary Table S10, and the VIFs for the bioclimatic variables included in the best models are presented in Supplementary Table S11 to confirm that there was no major correlation between them. As explained before, the lack of significant correlation between the bioclimatic variables was assessed with VIF (Supplementary Table S12).

SSD Variation

To analyze the covariation of SSD in relation to overall body size, we quantified allometric relationships using the function y = ax^b, where x and y are body size, and a and b are constants. If x and y are measured in the same scale, y/x will be constant when b = 1. The presence of allometry for SSD can be estimated by regressing log (male size) on log (female size), using reduced major regression (RMA) (Fairbairn 1997).

Our analysis of SSD allometry investigates how the length of a particular trait in males scales with respect to the same trait in females. A regression slope b < 1 indicates an allometric pattern (SSD hypoallometry) in which female size exhibits greater variation than male size, whereas a slope b > 1 (SSD hyperallometry) reflects the opposite pattern, with male size varying more extensively than female size.

RMA regression was estimated using lmodel2 package (Legendre 2012) in R v.4.2.2 (R Core Team 2023).

Results and Discussion

Morphometric Variation Across Geographic and Climatic Clines

We present and discuss the analysis of phenotypic variation along geographic and climatic gradients in 3 native South American grasshoppers. Likewise, we will analyze the allometric relationships of phenotypic variables with respect to body size.

Dichroplus Elongatus

Previous studies in the ground-dwelling populations of D. elongatus located in the central-west Argentina have demonstrated latitudinal patterns of body size variation along a transect of about 900 km (Rosetti et al. 2026). Individuals become larger in cooler environments with a narrower thermal range, a pattern consistent with Bergmann’s rule. Dichroplus elongatus populations from colder, higher-latitude regions of central-west Argentina adapt to seasonal time constraints by evolving faster growth rates relative to populations at lower latitudes, in line with the hypothesis of Blanckenhorn and Demont (2004). This pattern also aligns with the temperature–size rule, which predicts smaller adults at higher temperatures and larger adults in cooler environments (Atkinson 1994, Angilletta et al. 2004).

Furthermore, in the same paper, we analyzed static allometry in D. elongatus from center-western Argentina and its association with climatic variables (see Figs 5 and 6 and Table 4 in Rosetti et al. 2026 for details). Most of the studied traits scale isometrically in both sexes (femur, tibia, and tegmen lengths), indicating that overall shape remains constant regardless of body size. However, some traits exhibited sex-dependent scaling. A significant negative Total×SEX interaction was detected, showing that tibia length increased at a higher rate with total body length in females than in males; this suggests that sexual dimorphism in this trait is, in part, driven by sex-specific allometric slopes (see Table 4 and Fig. 5B in Rosetti et al. 2026).

Relative tegmen length showed interactions with bioclimatic variables not associated with total length: males had longer tegmen at higher mean temperatures of the coldest quarter, while females had longer tegmen at higher precipitation of the wettest quarter (PPWQ), revealing sex-specific environmental effects (see Fig. 6C and D in Rosetti et al. 2026). These sex-specific environmental responses suggest that sexual dimorphism in tegmen size is ecologically plastic and potentially driven by divergent selective pressures acting on each sex.

Thorax length is the only trait that exhibited hypoallometry, with a stronger reduction in males. Additionally, thorax length increased at a higher rate with total body length in females than in males, demonstrating again a sex-specific effect (see Table 4 and Fig. 5D in Rosetti et al. 2026). Regarding climate variables, we detected significant interactions with sex. Specifically, the interaction between sex and TAR was positive, whereas the interaction with mean temperature of the warmest quarter was negative, highlighting distinct sex-specific differences in thorax length in response to environmental conditions (see Fig. 6E and F in Rosetti et al. 2026). Overall, these results indicate that some traits display sexual dimorphism and are influenced by environmental conditions in a sex-dependent manner.

Dichroplus Vittatus

Morphometric variation of brachypterous individuals in terrestrial populations of D. vittatus from central-west Argentina, across 700 km, demonstrated that the best LMM model that explains the total length (as proxy of body size) variability is the one that includes sex (SEX), PPWQ, TAR, and interaction between TAR and SEX (Table 1; Fig. 1A and B). Sex dimorphism biased toward females is evident in D. vittatus (Table 1; Fig. 1). Total length shows positive associations with PPWQ in both males and females, with no significant interaction. Thus, individuals tend to be larger under higher precipitation conditions (Fig. 1A). However, the response to thermal variability was sex-specific: while females showed a larger body size with respect to TAR, males become smaller. (Fig. 1B). This divergent pattern suggests a trade-off between development time and adult body size driven by different selective pressures. As noted by Horne et al. (2019), such sexual divergence can occur when females follow a converse Bergmann pattern to maximize fecundity, while males may prioritize protandry—maturing earlier to maximize mating opportunities—at the expense of final body size. In the transitional landscape studied here, higher thermal ranges might compress the favorable season, forcing males to accelerate development (becoming smaller) while females maintain or increase size to ensure reproductive output.

Table 1.

Summarized results of LMM of total length as response variable with scaled climatic variables and sex as fixed effects in Dichroplus vittatus and Cornops aquaticum

	b	95% CI		t	P value
		Lower	Upper
i) Dichroplus vittatus
SEX	−1.079	−1.208	−0.944	−16.05	<10−4
PPWQ	0.462	0.125	0.799	2.46	0.04
SEX × TAR	−0.292	−0.513	−0.075	−2.59	0.01
ii) Cornops aquaticum
SEX	−2.0110	−2.193	−1.815	−20.87	<10−4
MDR	0.2530	0.097	0.408	2.88	0.02
MTDQ	0.3129	0.125	0.501	2.93	0.02
PPDQ	0.3072	0.147	0.466	3.39	0.01

b, regression slope; t: t-statistic.

Fig. 1.

Predicted associations based on the linear mixed model (LMM) testing the influence of environmental variables PPWQ (A) and TAR (B) on total length (in mm) as a proxy of body size in males and females of Dichroplus vittatus (red line for females and blue line for males).

Static allometric relationships for the lengths of the third femur, third tibia, tegmen, and thorax were analyzed relative to total length (used as a proxy for body size through LMMs) (Table 2). The former 3 traits scaled hypoallometrically (with regression slopes significantly lower than 1), revealing that these traits increase disproportionately more slowly relative to body size across the studied area (Table 2; Fig. 2A–C). We also detected a significant negative SEX × Total interaction for tibia length, suggesting that the relative scaling rate of this trait significantly decreases in males compared to females, resulting in a more severe hypoallometry in males (Table 2; Fig. 2C). In line with this, parallel trends were observed in D. elongatus of the same geographic area (Rosetti et al. 2026). Notably, the significant SEX × Total interaction for tibia length was shared by both Dichroplus species. These results suggest that, in both species, the sex-specific selective pressures modulate hindleg scaling, even though the specific allometric scaling (isometric vs hypoallometric) differs between them.

Table 2.

Summarized results of the LMMs of the relationship between 5 morphometric traits with respect to body size in Dichroplus vittatus

	b	95% CI		t	P value
		Lower	Upper
Thorax L-total	0.0346	−0.0737	0.1574	0.59	0.55
SEX	−0.1390	−0.1479	−0.1290	−28.89	<10−4
TAR	0.0368	0.0197	0.0539	3.83	0.006
PPWQ	0.0474	0.0283	0.0666	4.42	0.003
SEX × TAR	−0.0289	−0.0396	−0.0184	−5.29	<10−4
SEX × PPWQ	−0.0291	−0.0395	−0.0190	−5.53	<10−4
Third femur-total	0.2318	0.1305	0.3454	4.21	<10−4
SEX	−0.1010	−0.1094	−0.0917	−22.36	<10−4
PPWQ	0.0323	0.0192	0.0454	4.36	0.003
TAR × SEX	0.0113	0.0010	0.0211	2.18	0.03
Third tibia-total	0.4140	0.2030	0.6632	3.58	0.001
SEX	0.2076	0.0011	0.4212	1.91	0.06
PPWQ	0.0263	0.0149	0.0377	4.06	0.006
Total × SEX	−0.3619	−0.5939	−0.1355	−3.06	0.003
Tegmen-total	0.3435	0.1916	0.5125	4.23	<10−4
SEX	−0.0926	−0.1052	−0.0788	−13.85	<10−4
TAR	0.0319	0.0046	0.0590	2.09	0.08
PPWQ	0.0598	0.0290	0.0905	3.47	0.01

b, regression slope; t: t-statistic.

Fig. 2.

Allometric scaling of femur (A), tibia (B), and tegmen (C) of Dichroplus vittatus. Predicted relationships are indicated by solid lines for each sex (red line for females and blue line for males).

In contrast, thorax length in D. vittatus was independent of total body length, as the regression slope was not significantly different from zero (Table 2). Rather than a typical allometric constraint, this lack of significant relationship suggests a morphological decoupling between thorax size and overall body length (Emlen and Nijhout 2000, Whitman 2008), indicating that thorax development may be highly canalized in brachypterous individuals. This indicates that thorax development may be highly canalized regarding internal physiological scaling, yet highly responsive to external climatic cues. Concurrently, in brachypterous individuals, the reduction of flight constraints likely relaxes the genetic or functional integration between thoracic dimensions and total body size, allowing this trait to vary independently.

In relation to bioclimatic variables, the best LMMs identified positive associations with PPWQ for femur, tibia, and tegmen lengths (Fig. 3C, E, and F), with tegmen length additionally showing a marginally significant positive relationship with TAR (Table 2). Moreover, the static allometry of femur length was significantly affected by the interaction between TAR and sex, evidencing a sex-dependent effect (Fig. 3D).

Fig. 3.

Predicted association between log–log scaled morphometric traits and environmental variables in Dichroplus vittatus. Thorax length vs PPWQ (A) and TAR (B), Femur length vs PPWQ (C) and TAR (D), Tibia length vs PPWQ (E), and Tegmen length vs PPWQ (F).

The best LMM model for scaled thorax length includes a significant positive association with TAR and PPWQ, as well as significant interactions of both variables with sex (Table 2). Individuals have a larger scaled thorax at higher thermal amplitudes and precipitation level, though this effect is more pronounced in females (Fig. 3A and B). The significant association of this trait with climatic variables suggests a trait-specific sensitivity to environmental fluctuations that is independent of overall body size changes.

Regarding the random effects, population-level variation, confirmed by the ICC for each trait, demonstrated some divergence among grasshoppers from different populations, with the tibia being the trait displaying the lowest value (Supplementary Table S13).

In general, in D. vittatus, there were evident associations between body size and scaled morphometric traits with 2 climatic variables, PPWQ and TAR, which are key drivers of orthopteran development. Interestingly, the more pronounced response in females regarding total body length and thorax and tibia scaled lengths suggests that fecundity selection exerts a strong influence on female phenotype, while the more pronounced effect of climatic variables on femur length in males underscores the pivotal role of sex-specific selection in shaping male locomotory morphology.

Cornops Aquaticum

Morphometric variation in the semiaquatic species C. aquaticum along the Medium Paraná and Uruguay Rivers follows the converse Bergmann’s rule (Colombo et al. 2021). In the present study, we examined the variables most likely to predict total length. The LMM that most effectively explains variation in total length includes the fixed effects sex (SEX), MDR, MTDQ, and PPDQ. Based on these results, sexual dimorphism appears to be strongly biased toward females (Table 1; Fig. 4). All these climatic variables show positive associations with total length in both males and females, with no significant interaction. Consequently, individuals tend to be larger under higher temperature and rainfall conditions (Fig. 4). Earlier research in this species revealed disparity between neutral (molecular) marker distribution when compared with geographical data, thus strongly suggesting the likely adaptiveness of this clinal morphological pattern (Colombo et al. 2021).

Fig. 4.

Predicted associations based on the linear mixed model (LMM) testing the influence of environmental variables MDR (A), MTDQ (B), and PPDQ (C) on total length (in mm) as a proxy of body size, in males and females of Cornops aquaticum (red line for females and blue line for males).

Static allometric relationships for the lengths of the third femur, third tibia, tegmen, and thorax, as well as thorax height, were examined relative to total length, used as a proxy for body size, using LMMs (Table 3). The 5 morphometric traits displayed hypoallometry or negative allometry (slope < 1), suggesting that their growth is disproportionately low relative to body size throughout the examined geographic region (Table 3, Fig. 5). The best models for each morphometric trait include different climatic variables, which are shown in Table 3. Only 3 significant effects were detected for bioclimatic variables: scaled tibia length is associated positively with PPDQ, while scaled thorax height showed positive relationships with MDR and MTDQ (Fig. 6). These results indicate that, in both sexes, at higher precipitations, temperatures and temperature amplitudes scaled tibia length and thorax height grow less in proportion to body size. Interpreting the negative static allometry in light of the converse Bergmann’s pattern that detected in C. aquaticum, we can suggest that populations at lower latitudes experience higher temperatures and precipitations, leading to larger body sizes, while the body size-related morphometric traits analyzed here increase to a lesser extent.

Table 3.

Summarized results of the LMMs of the relationship between 5 morphometric traits with respect to body size in Cornops aquaticum

	b	95% CI		t	P value
		Lower	Upper
Thorax L-total	0.2438	0.1743	0.3107	6.98	<10−4
SEX	−0.1610	−0.1785	−0.1446	−18.63	<10−4
MTWQ	0.0097	−0.0015	0.0209	1.60	0.14
PPDQ	0.036	−0.0027	0.0202	1.41	0.19
Thorax H-total	0.2446	0.1547	0.3271	5.54	<10−4
SEX	−0.1458	−0.1688	−0.1257	−13.25	<10−4
MDR	0.0238	0.0126	0.0338	3.71	0.005
MTDQ	0.0177	0.0060	0.0290	2.48	0.05
PPDQ	0.0063	−0.004	0.0164	0.98	0.36
Third femur-total	0.3241	0.2444	0.3980	8.31	<10−4
SEX	−0.1215	−0.1414	−0.1034	−12.62	<10−4
MDR	0.0050	−0.0090	0.0190	0.66	0.52
PPDQ	0.0066	−0.0087	0.0218	0.79	0.45
Third tibia-total	0.2878	0.2155	0.3604	7.78	<10−4
SEX	−0.1357	−0.1531	−0.1177	−15.18	<10−4
PPDQ	0.0162	0.0095	0.0228	4.83	0.001
Tegmen-total	0.2667	0.1871	0.3455	6.60	<10−4
SEX	−0.1260	−0.1458	−0.1066	−12.59	<10−4
PPDQ	0.0117	−0.0112	0.0348	0.99	0.34

b, regression slope; t: t-statistic.

Fig. 5.

Allometric scaling of thorax length (A), thorax height (B), and femur (C), tibia (D), and tegmen lengths (E) in Cornops aquaticum. Predicted relationships are indicated by solid lines for each sex (red line for females and blue line for males).

Fig. 6.

Predicted association between log–log scaled morphometric traits, and environmental variables in Cornops aquaticum. Thorax height vs MDR (A) and MTDQ (B) and tibia length vs PPDQ (C).

Regarding the random effects for C. aquaticum, population-level variation demonstrated a similar trend to that detected in D. vittatus, with tibia length displaying the lowest ICC value, almost zero in this case (Supplementary Table S13). This trait is the most conserved in both species. For all the evaluated traits, the proportion of inter population variation was smaller for C. aquaticum than for D. vittatus, probably reflecting the higher environmental stability of aquatic habitat in the former.

Static Allometric Comparison

The differential allometric growth patterns in all 3 species observed in certain traits may result from strategic differences in resource allocation (Whitman 2008, Voje et al. 2021). Most studied traits in D. elongatus, except for thorax length, showed isometry, whereas in D. vittatus and C. aquaticum, hypoallometry was evident across most studied traits.

Dichroplus elongatus and C. aquaticum exhibited static hypoallometry in thorax length, whereas D. vittatus showed that this scaled trait varies independently of total body size. These patterns may reflect contrasting resource investment strategies driven by flight capability. For monomorphic macropterous, hypoallometric species (D. elongatus, C. aquaticum) under environmental constraints, individuals often prioritize overall body size over specific structures like thorax length and/or may also reflect selection for flight efficiency (Bonduriansky 2007, Shingleton et al. 2007). In the brachypterous individuals of D. vittatus, thorax length varies in response to climatic variables but remains decoupled from overall body size variation. In this species, resource investment in the thorax is modulated directly by climatic variables rather than by total body size constraints, which is likely the result of the loss of flight-related functional demands in brachypterous morphs.

Environmental factors, such as temperature and/or precipitation, induce species-specific changes in thoracic growth (decreasing with climatic variables in D. elongatus and increasing in D. vittatus and C. aquaticum). Despite these contrasting thermal responses, a more pronounced pattern is typically observed in females of terrestrial species (appearing as static hypoallometry in D. elongatus and as a decoupled scaling in D. vittatus)—thus reflecting an evolutionary trade-off to maintain fecundity and maximize egg production (Roff 1992, Stearns 1992, Preziosi et al. 1996). Meanwhile, in semi-aquatic C. aquaticum, environmental constraints were similar in both sexes, with no clear evidence of sex-specific selection or female-biased investment.

Allen’s rule states that animals have shorter appendages in colder climates to reduce heat loss (Allen 1877, Whitman 2008). Allen’s rule research specifically conducted in arthropods has been comparatively less consistent and far reaching when compared with the massive evidence coming from vertebrate study. Orthopterans may display an Allenian or non Allenian-like pattern due to allometric, ecological, developmental, or reproductive effects, thermoregulation not being necessary for this explanation (Willott and Hassall 1998, Bidau and Marti 2008, Bai et al. 2016).

The protruding body parts in D. elongatus (third femur, third tibia, and tegmen lengths) scaled isometrically with body size (Rosetti et al. 2026). Empirical evidence indicates that D. elongatus populations from central-west Argentina follow a converse Allenian pattern for tegmen and limbs. Since these traits are negatively associated with temperature, this suggests that thermoregulatory constraints are not the primary driver of limb evolution in this terrestrial grasshopper. This pattern may be explained by biomechanical compensation. In the colder environments of central-west Argentina, converse Allenian pattern may compensate for lower muscle performance due to thermal constraints favoring longer hind limbs (femur and tibia) and would maintain effective jumping distances and escape maneuvers through increased mechanical leverage (Mousseau 1997, Bidau and Marti 2008). The simultaneous occurrence of Bergmann’s rule and a converse Allenian pattern in D. elongatus suggests an interesting adaptation. While larger body sizes in cooler regions support Bergmann’s rule, probably favoring fecundity, longer limbs (converse Allen pattern) may compensate for temperature-induced muscular constraints at lower temperatures, thereby maintaining locomotor efficiency (Katz and Gosline 1993). In the terrestrial winged species D. elongatus, the isometric scaling of the tegmen and limbs suggests that maintaining dispersal capacity is a primary selective driver. In the arid and thermally seasonal environments of central-west Argentina, the ability to colonize optimal microhabitats and locate patchy resources before winter likely outweighs the thermoregulatory pressures typically associated with Allen’s rule (Bidau and Marti 2008, Whitman 2008).

In contrast to D. elongatus, brachypterous individuals of D. vittatus exhibit an Allenian pattern, where protruding parts such as the third femur, third tibia, and tegmen scale hypometrically with body size, particularly in colder environments. These results showed that tegmen and limbs grow lower than body size due to functional constraints, a pattern that is more pronounced in colder environments due to reduced growth season and thermal-developmental constraints (Allen’s Rule pattern). Hypoallometric scaling of hind limbs may represent a biomechanical adjustment to increased body mass, maintaining structural robustness and locomotor efficiency. The relative reduction in hind limb size may also reflect a shift in resource allocation. The decreased investment in locomotor structures could allow greater allocation to life-history traits (fecundity, sexual selection) (Whitman 2008, Voje et al. 2021).

The observed hypoallometry in D. vittatus aligns with recent studies on flightless and brachypterous orthopterans, where a reduction in flight capacity triggers a shift in morphological constraints and energetic investment (Meza-Joya et al. 2022). In the absence of flight, appendage scaling is often reorganized to prioritize terrestrial locomotion and environmental adaptation (Chang et al. 2021). For D. vittatus, this reorganization manifests a stricter adherence to Allen’s Rule, where cold-climate populations optimize their energy budget through a more compact morphology. This suggests that the loss of flight relaxes the tight integration required for dispersal, allowing limb and tegmen lengths to scale hypoallometrically (Katz and Gosline 1993, Hale 2014). Conversely, in fully winged counterparts like D. elongatus, these traits must remain tightly integrated with overall body size (isometry) to preserve locomotor performance and dispersal capacity.

The protruding parts (femur, tibia, and tegmen length), in the semiaquatic C. aquaticum, scaled hypoallometrically but did not show any association with temperature variables, so there is no evidence of Allenian or converse Allenian pattern. The lack of an Allenian or converse Allenian pattern in C. aquaticum highlights the fundamental difference between terrestrial and semi-aquatic life histories. For this species, the protruding parts scale hypometrically and appear decoupled from temperature variables, suggesting that environmental thermal gradients are not the primary drivers of its morphology (Danks 2007).

This independence from temperature can be attributed to the thermal buffering capacity of water and the humid microhabitats provided by its host plants (Haque et al. 2025). Unlike the terrestrial Dichroplus species, which are exposed to high thermal seasonality and aridity, C. aquaticum inhabits a more stable environment where the selective pressures related to climatic variability may be less pronounced than in terrestrial environments.

Furthermore, the observed hypoallometry in tegmen and limbs likely reflect functional and hydrodynamic constraints (Vogel 1996). In a semi-aquatic medium, excessive appendage length could increase drag or hinder movement within dense vegetation. Thus, the morphometric variation in C. aquaticum appears to be shaped by the mechanical requirements of its specialized habitat rather than the macroclimatic factors that shape the phenotype of its terrestrial counterparts (Adis et al. 2008, Pincebourde and Woods 2012).

Variation in SSD

Morphological differences between males and females are an important source of phenotypic variation within species. All species analyzed here show female-biased SSD, which represents the general trend within Orthoptera. The evolution of larger females is frequently attributed to the positive correlation between size and ovariole/egg production (Bellinger and Pienkowski 1985). As detected previously in D. elongatus, females may increase their reproductive potential, reaching higher adult body size, which allows them to produce more eggs (Rosetti and Remis 2015).

Static allometry, growth, and development represent other main factors shaping SSD (Roff 1992, Bidau and Marti 2007). The higher number of nymphal instars in females, and so earlier emergence of adult males (protandry), leads to a higher body size in females, whereas males tend to be smaller sized (Bellinger and Pienkowski 1989, Esperk et al. 2007). In D. elongatus, D. vittatus, and C. aquaticum, the number of nymphal instars is 5 in males and 5 or 6 in females (Turk and Barrera 1979, Adis et al. 2004). These results pointed out the relative importance of development time molding SSD.

To test the relative importance of allometry in SSD, we regressed log (male size) on log (female size) in all 3 studied species (Table 4). In D. elongatus, the RMA slopes for femur, tibia, and tegmen length did not differ significantly from 1 (isometry). For thorax length, the results suggest a possible trend toward static hypoallometry for this trait; while the statistical significance is marginal (P = 0.07), the confidence interval for the slope fails to include 1 (Table 4). These results agree with previous findings, in which populations of this species from central-east Argentina showed a similar trend toward a decrease in thorax length SSD with increasing total body size (Rosetti and Remis 2015; Table 4).

Table 4.

Reduced major-axis regression (RMA) analysis of log (mean male size) on log (mean female size) from populations of Dichroplus elongatus, Dichroplus vittatus, and Cornops aquaticum

	Trait	b	T	df	P value	95% CI
Dichroplus elongatus
i) Central-east  a	Thorax L	0.408	−2.0120	9.9020	0.07	0.2369 to 0.9019
	Femur	0.769	−1.0510	8.5790	0.32	0.4489 to 1.2580
	Tibia	0.740	−1.0250	9.0940	0.33	0.3976 to 1.2864
	Tegmen	0.838	−0.642	8.1390	0.53	0.4078 to 1.3085
ii) Central-west	Thorax L	0.6010	−1.9917	11.1182	0.07	0.3739 to 0.9658
	Femur	0.8739	−0.6599	10.9301	0.52	0.5566 to 1.3719
	Tibia	0.8724	−0.5996	11.3980	0.56	0.5299 to 1.4362
	Tegmen	0.7007	−1.1262	13.0069	0.28	0.3807 to 1.2897
Dichroplus vittatus
	Thorax L	0.4977	−2.2724	8.5927	0.05	0.2836 to 0.8735
	Femur	0.8032	−0.8832	8.6162	0.40	0.4558 to 1.4154
	Tibia	0.7950	−0.7763	9.2736	0.46	0.4111 to 1.5370
	Tegmen	1.157	1.1353	7.7786	0.29	0.8051 to 1.6620
Cornops aquaticum  b
	Total	1.0481	0.2099	12.9452	0.84	0.6207 to 1.7699
	Femur	1.1681	0.7109	12.0241	0.50	0.6744 to 2.0232
	Tibia	0.8551	0.5543	14.2010	0.59	0.4739 to 1.5431
	Thorax L	0.9556	0.1611	14.9175	0.87	0.5143 to 1.7755
	Thorax H	0.5268	1.8048	14.2406	0.09	0.2821 to 0.9838
	Tegmen	1.2784	1.1178	13.6783	0.28	0.7263 to 2.2504

b, slope; T, Clarke’s T statistic; df, degrees of freedom.

^aFrom Rosetti and Remis (2015).

^bFrom Romero et al. (2014).

Similar results were also found in populations of D. vittatus where all SSD traits scaled isometrically, except for thorax length, which seems to scale hypoallometrically (Table 4). Our results indicate that males and females do not vary in a similar fashion across the studied areas, thorax length tends to increase less than proportionally in males compared to females following the converse of Rensch’s rule.

Consequently, while strict isometry is the prevailing pattern for most traits across both Dichroplus species, thorax dimensions consistently exhibit lower slopes (ranging from 0.40 to 0.60). This hypoallometric pattern was statistically significant in D. vittatus, while a parallel trend was evident in D. elongatus, indicating a decrease in thorax SSD as body size increases across these terrestrial species.

Our results suggest that female size is more sensitive to the aforementioned environmental conditions than male size, as previously noted, females showed a larger relative increase in body size than males as conditions improved. The reduced scaling of thorax SSD in terrestrial grasshoppers may be shaped by a combination of sex-specific allometric constraints and environmental factors.

Regarding C. aquaticum, the regression of log (male size) on log (female size) to test allometry in SSD demonstrated that the RMA slopes for femur, tibia, and thorax lengths did not depart from 1, indicating that these traits scale isometrically in both males and females.

Thorax height showed the slope value that departed furthest from unity; although its P value was 0.07, this constitutes a suggestive result pointing toward a trend of static hypoallometry, which would be an interesting avenue for future research to confirm with larger datasets (Table 4).

The allometric slope for tegmen length SSD was greater than 1, although it was not statistically significant when it comes to reject the hypothesis of isometry. Earlier research analyzing sexual dimorphism in C. aquaticum along a far wider latitudinal gradient (between Trinidad and Tobago, lat +10° and Punta del Este, lat −35°) showed significant hyperallometry in tegmen length (Romero et al. 2014). Even though in the present study, within a far smaller geographical range (between lat −27° and −34°), we failed to detect significant hyperallometry for tegmen length, previous research encompassing a larger geographic area reported significant hyperallometry for this trait. This contrast underscores that despite local variations, the species inherently exhibits hyperallometry when viewed at a broader spatial scale. This detected allometric pattern is consistent with Rensch’s rule, which predicts greater evolutionary divergence in male size than in female size. This can be explained by the fragmented nature of its host plant in aquatic landscapes (Adis and Junk 2003), where disproportionately longer wings in larger males may enhance dispersal capability and mate-finding success.

The analysis in allometric scaling for SSD showed that isometry is a common factor in the analyzed species, indicating that both sexes vary in the same fashion across the studied area. However, specific traits in certain species departed from isometry and exhibited significant allometric pattern shifts.

This work showed that D. vittatus follows the converse of Rensch’s Rule for thoracic traits, likely favoring female reproductive output in stable grasslands, whereas the hyperallometric investment in flight structures (tegmen) previously detected in C. aquaticum males underscores the high cost of mate-searching in fragmented aquatic landscapes. These findings suggest that SSD in South American grasshoppers is not a uniform trait, but a plastic response to the balance between fecundity selection and dispersal demands.

Concluding Remarks

• Species-Specific Clinal Variation: Body size in Argentine grasshoppers responds diversely to climatic gradients. While D. elongatus follows Bergmann’s rule (larger in cooler environments), the semi-aquatic C. aquaticum conforms to its converse, suggesting that the thermal buffering of aquatic habitats alters the selective pressures that are typical of terrestrial species. These patterns are associated with thermic amplitude, temperature, and precipitations and may be related to the duration of the season length.

• Patterns of Thoracic Variation: Allometric static patterns may reflect contrasting resource investment strategies driven by flight capability. Macropterous species (D. elongatus and C. aquaticum) showed static hypoallometry of thorax length, indicating a strong selective pressure to keep an optimal thoracic size—likely linked to muscle efficiency or structural stability—regardless of overall body size fluctuations. Furthermore, this pattern may reflect a resource allocation trade-off, where energy is prioritized toward fecundity or sexual selection traits rather than further thoracic growth. In the brachypterous individuals of D. vittatus, the relaxation of flight-related trade-offs allows resource investment in the thorax to be modulated directly by climatic variables rather than by total body size constraints.

• Sex-dependent Clinal Variation: In terrestrial species, some scaled traits are influenced by environmental conditions in a sex-dependent manner. On the contrary, the semiaquatic species living in more homogeneous habitats and with limited mobility probably experiences similar pressures on males and females, reducing sex-specific differences in the response to climatic variables.

• Adaptive Trade-offs and Allen’s Rule: Dichroplus elongatus (winged) exhibits a converse Allenian pattern, where longer limbs in cold environments may compensate for constraints on muscle performance related to thermal seasonality to maintain jumping and escape efficiency. In contrast, in the wing-dimorphic D. vittatus, the brachypterous individuals here analyzed follow Allen’s rule. In these flightless phenotypes, reduced appendages in colder climates likely reflect a shift in energy allocation toward other life-history traits (such as fecundity) and a stricter adherence to thermal-developmental constraints, since the functional requirement of longer wings for dispersal is absent. Finally, C. aquaticum shows neither Allenian nor converse Allenian patterns, highlighting that its morphology is governed by hydrodynamic and habitat-specific constraints (such as navigating dense aquatic vegetation) rather than macroclimatic thermal gradients.

• Developmental Drivers of SSD: Female-biased SSD is a stable trait across these species, primarily driven by differential nymphal development (protandry in males and additional instars in females), which allows females to maximize fecundity through larger body sizes. Static allometric patterns may also play a critical role in molding the final phenotype, adapting the proportional investment in locomotor versus reproductive structures in each sex. Additionally, the phenotypic plasticity in response to bioclimatic variables indicates that the degree of dimorphism is not fixed, but rather modulated by environmental pressures that affect each sex differently.

• Rensch’s Rule and Environmental Sensitivity: Across almost all evaluated traits and species in our dataset, strict isometry in SSD was the prevailing pattern, indicating a highly coordinated, proportional morphological scaling between the sexes. However, terrestrial D. vittatus tends to follow the converse of Rensch’s rule for thoracic traits, with female size appearing more sensitive to environmental improvements than male size. In the semi-aquatic C. aquaticum, our study failed to detect significant hyperallometry for tegmen length, thus maintaining strict isometry within this geographical range despite a slope greater than 1. However, previous research encompassing a larger geographic area reported significant hyperallometry for this trait, confirming that males are more sensitive to the environmental conditions than females under wider spatial scales.

• Ecological Synthesis: These results underscore that orthopteran phenotype evolution is a complex interaction between habitat use (terrestrial vs semi-aquatic), life history strategies, dispersal modes, and sex-specific selective pressures. Ultimately, our comprehensive framework demonstrates that macroclimatic gradients shape orthopteran morphology through a delicate balance between adaptive eco-geographical strategies and structural constraints. While environmental seasonality drives species-specific clinal shifts in body size (following Bergmann’s rule or its converse) and appendage scaling (in line with Allen’s rule or its converse) to optimize locomotor or reproductive efficiency, these selective pressures operate primarily within a highly conserved, isometric framework of SSD. The predictable buffering capacity of the semi-aquatic habitat versus the fluctuating pressures of terrestrial grasslands further determines whether phenotypes undergo strict integrated scaling or localized morphological decoupling—such as the thoracic variations linked to flight capability. In sum, this study shifts the paradigm from rigid macroevolutionary rules to a more nuanced view, where geographic phenotypic variation is dynamically sculpted by the functional demands of the ecological niche.

Author Contributions

Maria Isabel Remis (Conceptualization [lead], Data curation [equal], Formal analysis [equal], Funding acquisition [lead], Investigation [equal], Methodology [equal], Project administration [lead], Resources [equal], Supervision [lead], Validation [equal], Visualization [equal], Writing—original draft [equal], Writing—review & editing [equal]), Paula Gomez Cendra (Data curation [equal], Formal analysis [equal], Software [lead], Validation [equal], Visualization [equal], Writing—original draft [equal], Writing—review & editing [equal]), Natalia Rosetti (Data curation [equal], Investigation [equal], Validation [equal], Writing—original draft [equal], Writing—review & editing [equal]), and Pablo Cesar Colombo (Conceptualization [equal], Investigation [equal], Validation [equal], Writing—original draft [equal], Writing—review & editing [equal])

Supplementary Material

Supplementary material is available at Journal of Insect Science online.

Funding

Financial support from Consejo Nacional de Investigaciones Científicas y Técnicas (Argentina) (PIP 11220220100132CO) to MIR is gratefully acknowledged.

Conflicts of Interest

None declared.