Relationships between environmental conditions and the zooplankton community
In order to evaluate the extent to which physical environmental variables may shape the spatial and temporal structure of the zooplankton community in the GoM, the relationships between environmental data and taxa-abundance matrices were evaluated through DistLM. These analyses were conducted using both the genetic information of each cruise and in a comprehensive analysis in which all stations (from all cruises) were included. For the latter, the results of the DistLM marginal tests indicated that each of the environmental predictors (when considered in isolation) explained a fraction of the variation observed in the zooplankton community structure (ranging from 2% up to 8.5%; Table 1A and 1B). However, the sequential tests showed that up to 18% of zooplankton diversity may be explained by the synergistic combination of DO, temperature, and longitude for both loci (Table 1C and 1D; R2 < 0.19, P < 0.0004). Moreover, for only 18S, an additional 1.68% of zooplankton variability may be explained by salinity (Table 1C). Similar results were also obtained with the individual cruise datasets (Supplementary Information Table S5).
The relationship between environmental parameters and abundance at the family level detected with both loci was visualized with a dbRDA plot. The first axis captured up to 43.1% of the fitted variability and 8.4% of the total variation among both loci and was mainly correlated with two hydrographic variables (oxygen and temperature; Fig. 3A and 3B). The second axis captured up to 35.2% of the fitted variability and ~6% of the total variation, and the strongest correlation was with longitude. Notably, zooplankton abundance scaled negatively with DO, temperature, and longitude for the majority of stations, indicating that a higher degree of complexity in the zooplankton community could be present in waters that are relatively cold and that have low dissolved oxygen concentrations, which were located in the western portion of the GoM.
In order to further evaluate this result and explore the potential differences in community structure between stations, we clustered each station considering its hydrographic features and geographic position (Fig. 4). Our finding suggests the presence of three different “ecoregions” in the deep water region of the southern GoM. The first ecoregion mostly included stations off the eastern coast of the Yucatan platform (sampling line Y, hereafter Y) located at < 86.30 °W (Fig. 5 and Fig. 6), which were mainly characterized by high oxygen concentrations and temperatures. Moreover, among the western stations of the GoM (> 86.30 °W), an additional latitudinal boundary was observed in the proximity of the 22 °N parallel (Fig. 6). Indeed, the northern stations (sampling lines: A, B, and C; hereafter N) generally presented higher concentrations of oxygen and higher water temperatures than those of the southern stations (sampling lines: D, E, F, G, H, and J; hereafter S). However, the proposed partition was less clear for the stations located along Lines C, D, and E, which suggests that environmental differences are most notable between the extreme southern and northern sectors of the surveyed area and less marked in between (Fig. 6).
Overall, the hydrographic measurements followed trends that are expected with seasonal change; salinity and temperature were higher during summer (and highest in August-September 2017 cruise XIXIMI-06) while oxygen, fluorescence, and density were higher in late spring (June 2016 cruise XIXIMI-05). However, no significant inter-annual differences were observed with regard to hydrographic predictors with the exception of dissolved oxygen, which was significantly higher in late spring (XIXIMI-05) compared to that in summer (XIXIMI-04 and XIXIMI-06; one-way ANOVA; F = 9.97, p < 0.0002). Additionally, we observed significant differences in the environmental variables among the N, S, and Y ecoregions (PERMANOVA; pseudo-f > 4.7623; p = 0.0002).