Variation in Pheidole nodus (Hymenoptera: Formicidae) functional morphology across urban parks

Introduction

The paradigm of morphological shifting remains a central topic in evolutionary biology. Traditionally, ecogeographical rules (e.g., Bergmann’s, Allen’s and Hesse’s rules) have often been used to explain spatial and/ or temporal patterns of morphological variation for endothermic (e.g., Cui et al., 2020; Ryding et al., 2021) or ectothermic (e.g., Arnan, Cerdá & Retana, 2015; Guilherme et al., 2019; Sosiak & Barden, 2021) species in response to gradients in environmental conditions, such as temperature. Variations in morphological traits may also result from phenotypic plasticity; that is, the capacity of a single genotype to produce a range of phenotypes under different environmental conditions (Gibbin et al., 2017); however, plasticity can be adaptive, non-adaptive or neutral depending on its relation to the optimal fitness in the changed environment (Ghalambor et al., 2007). Adaptive responses occur when the plastic response is favored by directional selection.

Environments may differ in their ecological optima (i.e., a certain combination of ambient factors that is optimal for the growth, existence and reproduction of an organism), leading to directional selection and shifts in a population’s position within ecological space, or environments may differ in the range of ecological variation they can support (i.e., in the strength of environmental filtering they impose), constraining the volume of ecological space occupied by a species, or population (Algar & López-Darias, 2016). These two options are not mutually exclusive and could act in concert or opposition. Morphological shifts can occur at different taxonomic levels, but typically involve the adaptation of populations to localized environmental conditions (Cohen, Haran & Dor, 2019; Cohen et al., 2021), especially when driven by ecological isolation.

Urbanization is an increasingly important driver of global change (Szulkin, Munshi-South & Charmantier, 2020; Des Roches et al., 2021). Urban ecological factors (e.g., food availability, physiological demands, habitat modification, pollution, thermal landscape) create selective pressures that can have a strong effect on remnant populations, resulting in phenotypic traits that are divergent from conspecifics living in natural areas (i.e., urban-derived phenotypes, Diamond & Martin, 2021; Diamond, Prileson & Martin, 2022). For example, in urban areas, neotropical lizard (Anolis cristatellus) populations exhibit longer limbs relative to body size, more sub-digital scales and a greater heat tolerance, although it remains to be established if this is genetically based (Winchell et al., 2016; Campbell-Staton et al., 2020). Importantly, phenotypic changes may occur within the lifespan of the individual (Cohen et al., 2021).

Within an urban ecosystem, isolated ‘green spaces’ represent ‘green islands’ with the potential to preserve faunistic and floristic assemblages. City parks often differ from one another in size, structure and complexity, and thus provide highly informative crucibles for studying how individual organisms adapt their tolerances, resulting in differentiation among populations enabling them to persist within them. Evidence of differentiation is scarce, but has been found both in invertebrates (De Carvalho et al., 2017) and vertebrates (Littleford-Colquhoun et al., 2017). Furthermore, Winchell et al. (2023) have shown that genomic parallelism could underlie phenotypic parallelism (i.e., convergence), in response to consistent urban environmental trends.

Ants are an ecologically dominant faunal group in most terrestrial ecosystems, which play key ecological roles as nutrient cyclers, predators, soil engineers and regulators of plant growth and reproduction (Czechowski et al., 2012). Studies on ants have informed understanding on ecological responses to disturbance and land management (Andersen, 2019), and how they can provide a bioindicator of environmental degradation (Andersen, 1997; Andersen & Majer, 2004).

There is an extensive literature describing how ants adapt to urban park conditions; however, this mainly focuses on diversity and the structure of ant community assemblages (Clarke, Fisher & LeBuhn, 2008; Ślipiński, Zmihorski & Czechowski, 2012; Carpintero & Reyes-López, 2014; Reyes-López & Carpintero, 2014; Liu et al., 2019; Nooten et al., 2019). In contrast, there is a scarcity of research examining whether environmental heterogeneity among isolated urban parks can cause morphological shifts within fragmented ant populations. Given, however, that morphological shifts among worker ants from various species have been attributed to various environmental drivers in populations in natural habitats, morphological shifts adapting to urban habitats seem highly plausible (Heinze et al., 2003; Clémencet & Doums, 2007; Bernadou et al., 2016; Antonov, 2017; but see Gaudard, Robertson & Bishop, 2019). In these previous studies, homospecific populations tended to be characterized on the basis of the mean value of their functional traits, as though these represent a static unit in space and time. Ignoring intraspecific trait variation across paratypes could underestimate a species’ ability to compete with other community members (Ashton et al., 2010), the degree of niche and trait overlap (Courbaud, Vieilledent & Kunstler, 2012), resource use (Bolnick et al., 2003; Eggenberger et al., 2019) and potential responses to new conditions found in urban environments.

The prevalence of some ant generalist species in urban environments may be attributable to high functional morphology plasticity, enabling them to utilize different habitats. Examining workers of a local dominant ant species, Pheidole nodus, from urban parks in Nanchong city, China, here we tested for differences in how individuals from segregated populations fill morphological space (morphospace), and whether these differences arise due to the strength of environmental filtering (environmental filter-strength hypothesis). This would be implied if different phenotypic optima between park environments drive directional selection and shifts in a population’s position within morphospace. Alternatively, parks may differ in the range of ecological variation they can support (optimum-shift hypothesis), constraining the volume of ecological space occupied by each ant population. Either of these mechanisms in isolation, or combined, could potentially lead to shifts in the positions of populations in morphospace between habitats.

Materials & Methods

Study area

This study was conducted in seven urban parks around downtown Nanchong (30°46′–31°85′N, 105°44′–106°96′E), located in Sichuan (China) (Fig. 1). This downtown area covers c. 160 square kilometers, occupied c. 1.5 million residents. The downtown core stretches north-south along the Jialing River. Nanchong has an East Asia monsoon climate, with an annual rainfall of 980 to 1,150 mm, a temperature range of 15.8 to 17.8 °C, and relative humidity from 76.0% to 86.0% (Liu et al., 2018; Luo et al., 2023).

Results

A total of 163 P. nodus workers were collected and measured; 19 from Baita, 15 from Chenshou, 15 from Huohua, 18 from Nanhu, 38 from Qixiang, 15 from Shengtai, and 43 from Xishan. Workers from a single park uniquely occupied 79.7% of total population morphospace dissimilarity (MD) (Table 3), which was significant based on our null model. The contribution of each individual park to MD ranged from 0.9 to 28.6%, where unique morphospace volume differed significantly from the null exception for each park except for Baita (Table 3). Among park pairs, MD ranged from 89.6 to 99.5%, and only six park population pairs (of 21; Chenshou-Qixiang, Chenchou-Xishan, Huohua-Qixiang, Huohua-Xishan, Qixiang-Shengtai and Shengtai-Xishan) occupied similar convex hull volumes (Table 4, Fig. S3).

Morphospace (convex hull) volume clustering was significant at four of these seven parks, Chenshou (median value = 0.80), Huohua (1.14), Qixiang (2.76) and Shengtai (1.66), but was not significant for Xishan (3.86, P = 0.975); significantly larger convex hull volumes were evident at Baita (4.57) and Nanhu (8.50) parks. (Fig. 2A). Centroid displacement was significantly greater than the null expectation value at Baita (median population distance from pooled centroid across all populations = 0.98), Nanhu (1.25) and Qixiang (1.11) parks, but was not significant for Chenshou (0.74, P = 0.100) and Xishan (0.67, P = 0.624); significantly lower centroid displacements were evident for Huohua (0.57) and Shengtai (0.50) (Fig. 2B).

Discussion

We found that P. nodus populations from urban parks spaced several kilometers apart exhibited distinct morphometric traits variations. Although these seven urban populations were likely entirely segregated from each other, morphospace volumes at four of the seven urban parks were smaller than the null expectation, and centroid displacements at three of these urban parks exceeded the null expectation, suggesting trait clustering and a shift in trait optima. These results are highly consistent with the environmental filter-strength hypothesis, which proposes that the strength of environmental filtering varies across environments, leading to differential trait-clustering (phenotypic variance). Simultaneously, these results also support the optimum-shift hypothesis, which proposes that the phenotypic optimum differs between environments, leading to divergence in phenotypic morphospace position between populations in different environments (see: Algar & López-Darias, 2016). From this we deduce that generalist ant species, such as P. nodus, can exhibit both substantial intraspecific variation in phenotypic diversity and optima in response to differences in habitat conditions.

Furthermore, we also found substantial segregation for both phenotypic clustering and optima between populations. This suggests that ant inter-population dispersal via spillover between these urban parks was severely limited, due to habitat fragmentation. This is typical in urban landscapes, where residential, commercial and industrial development, along with roads and paved pedestrian areas present effective dispersal barriers, especially for species with restricted dispersal distances (Fletcher Jr, Reichert & Holmes, 2018). Indeed, the urban landscape habitat matrix presents a generally inhospitable and ubiquitous barrier to colonization by many urban-avoiding terrestrial arthropods (Ricketts, 2001; Schmidt, 2008; Brühl & Eltz, 2010). Recent studies of ants conducted across Manhattan’s urban habitat mosaic (Savage et al., 2015), as well as in the suburban riparian corridor in Ku-ring-gai Local Government Area, Sydney, Australia (Ives et al., 2011) and across urban parks in Taichung, China (Liu et al., 2019) all found similar evidence of impeded spillover between fragmented populations. As for P. nodus specifically, its dispersal potential is unknown but, in general, ants are poor dispersers, with genetically viscous population structures (Sundström, Seppä & Pamilo, 2005; Hakala, Perttu & Helanterä, 2019).

Land-use disturbance arising due to urbanization generally intensifies selective pressures that reduce diversity and causes biotic homogenization (Rocha-Ortega & Castaño Meneses, 2015; Morelli et al., 2016). Consequently, locally invasive non-native biotas often steadily replace native biotas. Our finding that ant populations in Chenshou, Huohua, Qixiang and Shengtai parks had more homogeneous morphological trait patterns suggests that selective pressures in these urban sites favor reduced phenotypic plasticity in corroboration of this homogenization scenario. Similar intra-specific canalization of phenotypic diversity has been noted for other taxa in urban settings, such as mosquitos (Culex nigripalpus) (De Carvalho et al., 2017), Puerto Rican Crested Anoles lizards (Anolis cristatellus) (Winchell et al., 2016; Winchell et al., 2018) and neotropical Wedge-billed Woodcreepers (Glyphorynchus spirurus) (Thompson et al., 2022). In contrast, urban conditions appear to drive phenotypic diversity in bumblebees (Bombus pascuorum and Bombus lapidarius), because urban landscaping and management can lead to cities having a greater morphological variety of flowering species in any one season compared to rural areas (Eggenberger et al., 2019).

In our study, the population at Xishan Park was exceptional, where neither morphospace volume nor centroid displacement were distinct from that expected based on random sampling across all the collected individuals. Plausibly, this may be because this park is much larger than the others we sampled and continues to support a remnant of more natural native broad-leaved evergreen forest, providing habitat diversity and a broader range of food types. That the intra-specific morphological diversity among P. nodus workers under Xishan Park’s more natural conditions was around average, compared to the full meta-population sampled, suggests that these ants have evolved an optimal base morphology suited to their most natural habitat.

Mechanistically, it remains difficult at this stage to resolve whether these functional trait shifts are (i) an adaptive response to urban conditions, (ii) simply an example of background capacity for phenotypic plasticity, or (iii) based on genetics or epigenetics. Although P. nodus belongs to the generalized Myrmicinae functional group, which prefer open habitats and are often favored by disturbance (Andersen, 2019), the tendency for phenotypic clustering and optimum displacement in urban populations suggests that this species does not exhibit pre-adaptations (exaptation; sensu Gould & Vrba, 1982) to urban conditions, unlike (for example) from cockroaches, pigeons and rats. Plausibly phenotypic characteristics differ among urban populations due to myriad novel environmental effects on ontological development. To test for a heritable genetic component would require testing the fitness of F1 and F2 offspring reared in a controlled environment (McDonnell & Hahs, 2015). Alternatively, if workers inherit traits from the queen that contribute to colony fitness, this would provide an adaptive explanation. Any heritable genotypic adaptation to novel environmental conditions would arise through either of two selective mechanisms: (1) the expression of phenotypic plasticity (the ability of one genotype to express varying phenotypes when exposed to different environmental conditions); and (2) evolution via directed selection for particular phenotypes, resulting in the modification of the population gene pool (Clusella-Trullas, Terblanche & Chown, 2010; Fox et al., 2019; Desmond, 2021; Jacquier et al., 2021; Lambert et al., 2021). Even in the absence of genetic data, the reduced morphospace and centroid displacement we detected are consistent with urban habitat fragmentation reducing genetic diversity within urban populations, while simultaneously increasing populations genetic differentiation from a standard neutral model. Such neutral processes are typically associated with detrimental fitness consequences, in part arising from fragmentation and population isolation of that leads to increased drift and elevated probability of inbreeding depression (Combs et al., 2018).

Ours is the first preliminary study to report differences in variance of ant functional morphology at species level among urban parks. Our study involved a relatively small sample size, due to the low abundance of ants in this highly disturbed region collected using a pitfall sampling method, chosen to sample morphological traits from only foraging workers experiencing urban habitat conditions, and to exclude colony workers. On the basis of this sample size, we caution that those differences in the morphospaces of each park population we detected, while indicative, should not be considered conclusive until further work builds on this foundation to more fully quantify the effects of ecological filters on intraspecific morphological variation among ants in urban micro-habitats.

Conclusions

The evidence we found for morphological distinctiveness and segregation for both phenotypic clustering and optima between populations implies that filtering and selective mechanisms may both be operating on ant functional morphological traits. Ultimately, however, it remains to be seen whether P. nodus will have sufficient plasticity to continue to persist faced with the novel challenges of urban habitats, especially with concomitant climate change pressures (Goh, 2020), and will be robust to those Allee effects that can perturb isolated populations (Schoereder et al., 2004). Nevertheless, our results begin to demonstrate how novel urban environments present challenges to ant generalist populations, to which species respond through phenotypic, and possibly adaptive, change.

Supplemental Information