Macrobenthic communities in a shallow normoxia to hypoxia gradient in the Humboldt upwelling ecosystem

Abstract

Hypoxia is one of the most important stressors affecting the health conditions of coastal ecosystems. In highly productive ecosystems such as the Humboldt Current ecosystem, the oxygen minimum zone is an important abiotic factor modulating the structure of benthic communities over the continental shelf. Herein, we study soft-bottom macrobenthic communities along a depth gradient-at 10, 20, 30 and 50 m-for two years to understand how hypoxia affects the structure of shallow communities at two sites in Mejillones Bay (23°S) in northern Chile. We test the hypothesis that, during months with shallow hypoxic zones, community structure will be much more dissimilar, thereby depicting a clear structural gradient with depth and correlated abiotic variables (e.g. organic matter, temperature and salinity). Likewise, during conditions of deeper hypoxic zones, communities will be similar among habitats as they could develop structure via succession in conditions with less stress. Throughout the sampling period (October 2015 to October 2017), the water column was hypoxic (from 2 to 0.5ml/l O2) most of the time, reaching shallow depths of 20 to 10 m. Only one episode of oxygenation was detected in June 2016, where normoxia (>2ml/l O2) reached down to 50 m. The structure of the communities depicted a clear pattern of increasing dissimilarity from shallow normoxic and deep hypoxic habitat. This pattern was persistent throughout time despite the occurrence of an oxygenation episode. Contrasting species abundance and biomass distribution explained the gradient in structure, arguably reflecting variable levels of hypoxia adaptation, i.e. few polychaetes such as Magelona physilia and Paraprionospio pinnata were only located in low oxygen habitats. The multivariable dispersion of community composition as a proxy of beta diversity decreased significantly with depth, suggesting loss of community structure and variability when transitioning from normoxic to hypoxic conditions. Our results show the presence of semi-permanent shallow hypoxia at Mejillones Bay, constraining diverse and more variable communities at a very shallow depth (10-20 m). These results must be considered in the context of the current decline of dissolved oxygen in most oceans and coastal regions and their impact on seabed biota.

Fig 1. Map of the study site.

Location of the two study sites at Mejillones Bay in northern Chile.

Fig 2. Oceanographic parameters.

Time series of temperature of dissolved oxygen, temperature and salinity at during the study period and sites. Blank spaces indicate months with no sampling.

Fig 3. Variability in taxonomic richness.

Temporal fluctuations in taxonomic richness (mean ± SD) of the macrobenthos per sampling site during the study period in Mejillones Bay. Color lines represent depth in m.

Fig 4. Variability in abundance.

Temporal fluctuations in abundance (mean ± SD) of the macrobenthos per sampling site during the study period in Mejillones Bay. Color lines represent depth in m.

Fig 5. Variability in biomass.

Temporal fluctuations in biomass (mean ± SD) of the macrobenthos per sampling site during the study period in Mejillones Bay. Color lines represent depth in m.

Fig 6. Patterns of variability in community structure.

nMDS of the community structure estimated from the abundance data.

Fig 7. Patterns of variability in community structure.

nMDS of the community structure estimated from the biomass data. Symbols are centroids estimated for the depth x month factor.

Fig 8. Species contribution to dissimilarity in community structure.

nMDS bubble plots showing the contribution of species abundance (selected by the BVSTEP analysis) to the patterns of dissimilarity observed in the environmental gradient and time. The size of the bubble represents the relative abundance of species. Left row contains species for Punta Chacaya and the right contains species for Playa Blanca.

Fig 9. Relationships between community and environmental parameters.

CAP ordination plots used as a discriminant function to show main grouping of community structure in function to the environmental parameters.

Fig 10. Relationships between community and environmental parameters.

Draftsman plots showing the gradient generated between the community structure and environmental parameters.