This study provides evidence for the shifts in boh taxonomic and functional diversity within the copepod communities across the ecoregions of the Central Atlantic Ocean. Our results reveal a high correlation between alpha taxonomic and functional metrics, indicating that modeling functional diversity hinges upon the species composition of copepod communities. Spatial gradients, including latitude, longitude, and depth, emerge as the primary drivers influencing these diversity shifts. In contrast, the analysis of β-diversity revealed considerable variations in species composition across ecoregions, despite exhibiting similar functions, resulting in a low turnover of functional β-diversity.
Comparing community composition with functional groups
The functional groups described in the present study differed from those identified by Benedetti et al. (2023). This discrepancy could be attributed to differences in spatial scale (regional versus global), species composition, and the functional analysis method applied. While Benedetti et al. (2023) utilized a factor analysis of mixed data (FAMD), our approach relied on methodologies based on multidimensional space. used a factor analysis of mixed data (FAMD), our approach relied on multidimensional space methodology (Villeger et al., 2008). For instance, in the study of Benedetti et al. (2023) Clausocalanus was placed in FG1 and Oncaea in FG2, reflecting their trophic regime disparity (omnivore-herbivore and detritivore, respectively). However, our findings grouped them together within the same functional group (FG1), prioritizing their classification as generalist species and their cruising feeding strategy, despite their low species count. Additionally, we observed a higher number of species within the FG5, which were dispersed in Benedetti et al. (2023), separating Calocalanus and Paracalanus from Scaphocalanus. Furthermore, we grouped Pleuromamma with Lucicutia, and Heterorrhabdus in the FG8, whereas Benedetti et al. (2023) associated this genus with Gaetanus and Metridia. Therefore, conducting a direct comparison between the two studies might not be the most adequate approach, and instead, the most relevant issue is the ecological patterns derived from each study.
The abundance and distribution of copepods across the Atlantic ecoregions have been described in several studies (Bottger-Schnack, 1997; Benedetti et al., 2023; Fernández de Puelles et al., 2023). They concluded that Oncaea and Oithona are the most common genera, with especially Oncea venusta and Oithona plumifera, followed by Clausocalanus with C. furcatus. Oncaeid species dominates the mesopelagic and bathypelagic layers across all ecoregions, with the highest abundance observed in the upwelling area. In contrast, Clausocalanus spp. inhabit the epipelagic layer, and although common, reach greater abundance in the southern areas (Peralba & Mazzocchi, 2004; Bendetti et al., 2022, Fernández de Puelles et al., 2023). The FG5 was the most represented group, encompassing 61 species from various genera such as Scaphocalanus, Calocalanus, Scolecithricella, and Paracalanus spp. While most genera maintained a consistent and low abundance across ecoregions (< 2%), the genus Paracalanus, represented by P. parvus and P. indicus, exhibited a high abundance in the upper layers of cold and nutrient-rich waters, particularly in Cape Verde and upwelling ecoregions (Fernández de Puelles et al., 2023). This could be linked to filter-feeding mode and the broadcaster spawning strategy, which requires lower energy demand. Certain functional groups may be more closely associated with specific oceanographic conditions. For example, FG8, represented by Pleuromamma spp. and Temora spp., was more abundant in the upwelling ecoregion (Fernandez de Puelles et al., 2023). However, this group comprised omnivore-detritivore and omnivore-herbivore species, allowing them to have a wider distribution and be better adapted to different environments. FG4, characterized by Oithona spp. and Corycaeus spp., showed slightly greater abundance in the oligotrophic and equatorial ecoregions (Fernández de Puelles et al., 2023). This distribution is influenced by competition under resource-limited conditions, leading to a higher presence of carnivore species utilizing ambush-feeding modes, which requires less energy (Kiørboe et al., 2011). Moreover, these conditions can promote carnivorous strategies, including cannibalism (Ohman & Hirche, 2001). Therefore, shifts in community composition seems to be associated with a greater frequency of certain functional traits (Benedetti et al., 2018, 2023; Becker et al., 2021; Tang et al., 2022).
We did not find that body size was significant for any of the principal components. However, previous studies have shown that this functional trait is related to temperature, with smaller copepods associated with warmer waters and larger copepods in cold waters (Becker et al., 2021; Tang et al., 2022). In contrast, we identified myelination as key trait, which may be linked to habitat preference. This could be explained by the fact that myelination is frequently associated with feeding mode and size (Benedetti et al., 2023). Myelinated copepods have a lipid-rich myelin sheath around their nerves, enabling faster reaction times and thus more efficient feeding or escape behaviours (Lenz, 2012). The myelination exhibits the same spatial patterns as body size and feeding mode. Amyelinated and small species mainly occurred in the tropical gyres, whereas large, myelinated copepods predominated in polar regions (Benedetti et al., 2023).
Modelling α-diversity
Diversity appeared to be closely linked to a strong stratified water column (Longhurst, 1985; Medellín-Mora et al., 2021) and to the composition of microzooplankton (Pierrot-Bults, 2003). Our findings revealed greater values of species richness, taxonomic distinctiveness and functional specialization in the shallower layers (0-200 m depth), with this trend being more pronounced in the oligotrophic waters. However, changes in DVM were only detected with the functional disparity, suggesting that these species occupy a wider ecological niche. In fact, Fernández de Puelles et al. (2023) pointed out the presence of non-migrant species within the epipelagic layer (e.g., Paracalanus, Clausocalanus, Calocalanus, and A. danae), alongside deepwater species that migrate to this layer during the night (e.g., Pleuromamma, Euchirella, Subeucalanus, Rhincalanus). The non-migrant species were grouped together (FG5) except for Acartia spp. (FG7). In the case of migrant species, they were assigned to different FGs. Euchirella was the most dominant genera in FG2, Rhincalanus and Subeucalanus were in FG5, and Pleurommama was important in FG8. Additionally, the functional evenness was higher in deeper layers for the oligotrophic ecoregion, indicating a greater uniformity in this particular ecoregion.
It is a well-established fact that the zooplankton distribution is influenced by factors such as food availability (Simões et al., 2013; Couret et al., 2023), environmental conditions (oxygen concentration) (Vinogradov, 1970), water column stratification (Longhutst, 1985; Braghin et al., 2021), and the distribution of the water masses (Bonecker et al., 2014). The mixed GAMs models pointed out spatial factors (latitude, longitude, and depth) as the key for understanding both the taxonomic and functional composition for most indices, aligned with these influential factors before mentioned. Interestingly, while the functional originality and specialization models did not show a clear geographical pattern, they did exhibit a sensitivity to oxygen fluctuations, particularly within the OMZ. Previous studies observed aggregations of copepods (e.g., Eucalanus, Subeucalanus, Paraeucalanus, and Pleuromamma) in OMZs (Karstensen et al., 2008; Jackson and Smith, 2016) suggesting a metabolic slowdown under such conditions. This indicates that the Cape Verde ecoregion harbours more specialized copepods. Conversely, chlorophyll did not emerge as a significant variable in any of the functional indices but showed significance in two of the three taxonomic indices (SR and Δ+). This suggests that regions with higher productivity may influence species richness but have lower impact determining the functional diversity at the local scale (Li et al., 2022; Tang et al., 2022). This occurrence might be attributed to the thorough examination of correlations among environmental variables and the integration of spatial variables, a facet often overlooked in alternative studies.
Taxonomic and functional β-diversity
The taxonomic and functional diversity exhibited opposite patterns, with taxonomic diversity being driven by species replacement and functional diversity by nestedness. This pattern is commonly observed in zooplankton assemblages (Braghin et al., 2018; Diniz et al., 2021; Liu et al., 2023) and certain fish communities (Villéger et al., 2012, 2013; Bender et al., 2017). It implies a damping effect within the system, where some species are replaced without significant functional differentiation, allowing the new species to perform similar ecological roles, even with extreme trait combinations. The functional space of each ecoregion was delineated by species with the most extreme trait combination, positioned at the extremes of the convex hull, whereas species with intermediate trait values affected the level of overlap between convex hulls (Woodward, 2009; Villéger et al., 2013; Braghin et al., 2018). Thus, ecoregions with low functional diversity became a subset of sites with high functional diversity, resulting in a high nesting pattern (Heino & Tolonen, 2017). It is also plausible that an extensive biogeographical scale might affect to potential local environmental constraints or the scale of species competition (Villéger et al., 2013).
Studies on zooplankton, including copepods, have demonstrated that environmental factors shape differences in functional traits (Pomerleau et al., 2015; Becker et al.,2021; Tang et al., 2021; Benedetti et al., 2023). In events of environmental stress such as El Niño or the seasonal variation in the southern Yellow Sea (Pomerleau et al., 2015; Li et al., 2022) a low functional dispersal was observed, suggesting a convergence of functional traits within the community. These environmental conditions affect the zooplankton distribution, leading to a pattern where each functional group is distributed according to its tolerance to specific environmental conditions. Theoretically, when environmental conditions shift, the most suited species to the new conditions may thrive, while others must respond through adaptive plasticity (Hutchinson, 1991; Chase & Myers, 2011). This process could result in the restructuring of communities, that leads either to homogenization or differentiation of assemblages, and causing rapid spatial changes (Cavender-Bares et al., 2009). Additionally, such shifts could lead to a gain or loss of biodiversity (Hooper et al., 2012). Interestingly, communities with high species richness can exhibit a loss of functional diversity due to an increase in functional redundancy (Villéger et al., 2013). Previous studies found a positive correlation between redundancy, resilience, and stability (Cardinale et al., 2006; Biggs et al., 2020). This suggests that ecological communities with more functionally redundant species are more likely to exhibit higher resilience or greater stability in both community structure and ecological function over time. Biggs et al. (2020) highlighted the significance of functional redundancy as a key component in biodiversity research and conservation, as it enables the prediction of overall ecosystem function despite variations in community structure, including species loss. However, the measures of functional redundancy used in previous studies were not demonstrated to be the most quantitatively representative (species richness within a functional group and functional space overlap) (Biggs et al., 2020). In this context, we evaluated both metrics and argued that copepods are already highly resilient compared with other marine organisms (Bakun and Weeks, 2004; Brander et al., 2007; Bates et al., 2013) under normal environmental conditions.