Modeling the Population Effects of Hypoxia on Atlantic Croaker (Micropogonias undulatus) in the Northwestern Gulf of Mexico: Part 2—Realistic Hypoxia and Eutrophication

Abstract

Quantifying the population-level effects of hypoxia on coastal fish species has been challenging. In the companion paper (part 1), we described an individual-based population model (IBM) for Atlantic croaker in the northwestern Gulf of Mexico (NWGOM) designed to quantify the long-term population responses to low dissolved oxygen (DO) concentrations during the summer. Here in part 2, we replace the idealized hypoxia conditions with realistic DO concentrations generated from a 3-dimensional water quality model. Three years were used and randomly arranged into a time series based on the historical occurrence of mild, intermediate, and severe hypoxia year types. We also used another water quality model to generate multipliers of the chlorophyll concentrations to reflect that croaker food can be correlated to the severity of hypoxia. Simulations used 100 years under normoxia and hypoxia conditions to examine croaker population responses to the following: (1) hypoxia with food uncoupled and coupled to the severity of hypoxia, (2) hypoxia reducing benthos due to direct mortality, (3) how much hypoxia would need to be reduced to offset decreased croaker food expected under 25 and 50% reduction in nutrient loadings, and (4) key assumptions about avoidance movement. Direct mortality on benthos had no effect on long-term simulated croaker abundance, and the effect of hypoxia (about a 25% reduction in abundance) was consistent whether chlorophyll (food) varied with hypoxia or not. Reductions in hypoxia needed with a 25% reduction in nutrient loadings to result in minimal loss of croaker is feasible, and the croaker population will likely do as well as possible (approach abundance under normoxia) under the 50% reduction in nutrient loadings. We conclude with a discussion of why we consider our simulation-based estimates of hypoxia causing a 25% reduction the long-term population abundance of croaker in the NWGOM to be realistic and robust.

Modeling the Population Effects of Hypoxia on Atlantic Croaker (Micropogonias undulatus) in the Northwestern Gulf of Mexico: Part 1—Model Description and Idealized Hypoxia

“…When we combine the results presented here and the results in the companion paper (Rose et al 2017a), we conclude that the long-term effect of hypoxia on croaker is about 25% decrease in abundance. We based this on the results presented here for the time series simulations that showed an 11% decrease.…”

“…However, the approach is ideal for isolating any potential populationlevel effects of hypoxia, which would then provide a minimum estimate of population effects and also provide information on the types of responses to be expected within the model and therefore easier to detect and quantify as more and more realistic conditions are added to the population model. In a companion paper (Rose et al 2017a), we explore population responses of croaker under increased variability by using DO output from a hydrodynamics-water quality model, perform sensitivity analysis that varies some key model assumptions and values of inputs, and also perform simulations that also couple the severity of hypoxia to the availability of food for croaker. In a third paper (Rose et al 2017b), we impose interannual variability on the simulated croaker population response to hypoxia to evaluate the ability of field sampling to detect a known population-level effect.…”

“…However, quantitative evidence that reducing hypoxia has positive population-level effects on coastal fish species is equivocal except for several well-studied systems ). The reasons that detecting hypoxia effects is difficult include high variation in population dynamics from other sources, uncertainties in establishing hypoxia exposure and effects on individuals in the field, and general challenges in monitoring and modeling multiple life stages that use different habitats (Rose 2000;Rose et al 2017a). We use the term Bpopulation level^here in the fisheries management sense of a fish stock: a stock is a group of reproducing individuals of the same species (at least semi-closed) that have similar life history parameters (Begg and Waldman 1999).…”

“…Adult croaker may be exposed to hypoxia due to their seasonal nearshore-offshore migrations on the shelf associated with the summer growing and fall-winter spawning seasons (Darnell et al 1983). Croaker generally prefer near-bottom conditions, and evidence suggests they move horizontally rather than vertically to avoid very low DO concentrations (Craig 2012; see summary in Rose et al 2017a). Low DO has been shown to alter croaker densities and foraging dynamics in ways that may impair growth (Eby et al 2005;Powers et al 2005), and summer exposure to hypoxia can affect fecundity during the following fall spawning season (Thomas et al 2007, Thomas andRahman 2012).…”

“…Mean lengths at age and stage durations were used as target values during model calibration of the overlying surface waters (Redalje and Fahnenstiel 1994;Wysocki et al 2006), suggesting that surface chlorophyll-a may be a crude index of benthic secondary production that is then directly utilized by croaker in the GOM. Rose et al (2017a) further discuss the use of surface chlorophyll as indicator of food for croaker. Three hypoxia scenarios, classified by the areal extent (km 2 ) of hypoxia (mild, intermediate, and severe), were simulated (Fig.…”