 |
||||
|
||||
|
|
|
 |
||
|
||||
      |
Hypoxia Effects on the Living Resources of the Northern Gulf of Mexico Project
Collaborators:
Steve Brandt, Stuart Ludsin, Doran Mason
Great Lakes Environmental Research Laboratory (GLERL home)
Bill Boicourt, Mike Roman, Dave Kimmel
University of Maryland (UMD web site)
Ecosystem Forecasting Objective
To provide the ecological understanding needed to forecast the effects of hypoxia on pelagic food webs. Specifically, we seek to develop conceptual and quantitative models that will be incorporated into larger ecosystem forecasting efforts.
Executive Summary
During 2007, conducted both field and lab-based work in support of this project. Specifically, we completed our second full sampling season during the peak of hypoxia, processed diets for a whole range of Gulf of Mexico fishes, and began analyzing data for manuscripts.
Funding
CSCOR -Center for Sponsored Coastal Ocean Research (CSCOR web site)
Scientific Rationale
Of the numerous anthropogenic impairments experienced by aquatic ecosystems during this century, none has been more ubiquitous than excessive nutrient inputs, or eutrophication (Caddy 1993; Diaz and Rosenberg 1995; Cloern 2001). A typical consequence of eutrophication is reduced oxygen availability in bottom waters, owing to enhanced bacterial respiration in sediments, which oftentimes results in severe hypoxia (< 2 mg/l) or even anoxia (0 mg/l). Eutrophication-driven reductions in hypolimnetic oxygen levels have been documented in freshwater, marine, and estuarine systems throughout the world (Caddy 1993; Diaz and Rosenberg 1995; Breitburg et al. 2001; Rabalais and Turner 2001), which sometimes can occupy the majority of the water column. For example, in the "dead zone"of the Gulf of Mexico, hypoxia normally occurs in 20% to 50% of the water column, but can sometimes encompass 80% of the water column, depending on pycnocline depth (Rabalais and Turner 2001). Similarly, in the Chesapeake Bay, severe hypoxia can encompass more than two-thirds of the water column during summer in mesohaline portions of the bay (Officer et al. 1984).
In turn, because all metazoans, including both benthic and pelagic macroinvertebrates and fish, require oxygen to function, reduced oxygen availability can have dramatic effects on population demographics. The most obvious effect is reduced population size, owing to direct mortality. Population regulation via direct mortality is most prominent in sessile organisms (Diaz and Rosenberg 1995; Rabalais et al. 2001), but also can be important in highly mobile organisms, including fish (Coutant 1985; Wannamaker and Rice 2000; Breitburg et al. 2001; Marcus 2001). Less obvious, but likely more common, is population regulation via indirect effects that alter an organism's behavior (Breitburg et al. 2001; Marcus 2001; Marcus et al. 2004). Indeed, numerous investigations with both invertebrate and vertebrate species have demonstrated how oxygen availability can disrupt typical habitat use and movement patterns (Roman et al. 1993; Kolar and Rahel 1993; Aku et al. 1997; Keister et al. 2000; Wannamaker and Rice 2000). In turn, altered access to refuge, food, or thermal habitat, owing to reduced oxygen availability, can influence population production (positively or negatively, depending on the species; Breitburg et al. 1997, 2001) by altering predation mortality, feeding rates, metabolic rates, or fecundity (Aku and Tonn 1997, 1999; Breitburg et al. 2003; Kolar and Rahel 1993; Klumb et al. 2004; Marcus et al. 2004).
Changes in the performance or behavior of individual species in response to reduced oxygen levels can subsequently lead to fluctuations in community structure and composition. Shifts from assemblages comprised of species intolerant of low oxygen levels to those comprised "tolerant"species have been documented for zooplankton and benthic macroinvertebrate communities in systems undergoing eutrophication (Carr and Hiltunen 1965; Patalas 1972; Verdonschot 1996; Marcus 2001; Wetzel et al. 2001; Wu 2002; Kane et al. 2004). Likewise, changes in fish community composition have been linked to hypoxia-induced alteration of food and thermal habitat (Ketchum 1969; Tonn and Magnuson 1982; Nürnberg 1995; Ludsin et al. 2001). Unfortunately, delineating exactly how reduced oxygen levels influence a population or community is oftentimes difficult, especially when using in situ empirical data to do so. Clearly, owing to the numerous ways in which oxygen availability can directly and indirectly influence both physical and biological components of a system, such approaches will not only require quantification of numerous habitat variables in addition to dissolved oxygen levels (e.g., temperature, prey species behavior/availability), but also a framework to simultaneously account for direct and indirect effects, including their linear and non-linear interactions, on the focal organism(s).
2007 Research Accomplishments
Field Program
A successful 10-day cruise was conducted during late July through early August in support of this study. Three 24-hr (diel) stations were sampled, as were three long north-south transects and a square-wave sampling design, using a combination of acoustics and bottom trawling (the mid-water trawl was not operational).
Diet and Laboratory Analysis
During FY07, 1,942 diets were processed in support of this study , across a range of trophic guilds. These designations into trophic guilds were based on 2006 diet data. Benthivores and planktivores had an overall diet comprised of over 60% benthic organisms (gastropods, bivalves, polychaetes, nematodes...) and plankton (copepods, shrimp, crabs...), respectively. Omnivores had neither benthic, planktonic, or fish prey in over 60% of their diets. Piscivores were species whose diets comprised 30-50% fish. Ultimately, these diets will support our efforts to build food webs for the Gulf of Mexico, and in turn, allow us to do EcoPath modeling.
1.Descriptive work on variables:
a. Data of all variables (water temperature, dissolved oxygen, salinity, chlorophyll a, zooplankton biomass, relative fish biomass) of each transect have been integrated into a matrix data file, which is ready for statistic analysis, graph plotting, or other studies.
b. We have made contour plots of variables of all transects to visualize their distributions.
c. To test the impacts of hypoxia on pelagic fish distribution, each transect (with hypoxia) was partitioned into four general regions: hypoxic zone (< 2.0 mg O2/l), normoxic zone (area above hypoxic zone), edge zones (within 1 km, 3 km or 5 km on each side of the hypoxic zone), and remote zone (> 5 km from the hypoxic zone). Preliminary results showed that Dissolved oxygen levels varied among years with hypoxia occupying from 0% to 29.5% of the water column along transects. Likewise, fish biomass and zooplankton biomass varied among years and transects. As expected, we observed lower biomass of fish in the hypoxic zone relative to other regions. Unexpectedly, however, we did not observe statistically significant accumulation of fish biomass at the edges of hypoxia; fish biomass equally as high above the hypoxic zone as it was adjacent to it (i.e., at the edges). Overall, our findings indicate that hypoxia in the northern Gulf of Mexico reduces available habitat for fish, which may also affect the energy flow from zooplankton to pelagic fish.
2.Growth rate potential model:
a. Growth rate potential (GRP) model has been used to assess ecosystem capacity, qualify habitats, responses of fish to stress factors. Some studies show it is a successful tool to estimate fish production. Others show its results were inconsistent with fisheries catch. This study uses GRP model with bay anchovy parameters to investigate how well it is as a habitat index in the northern Gulf of Mexico. GRP model results clearly show that fish avoid hypoxic water (<2.0 mg O2/l) as observed in the contour plots mentioned above. Habitat quality predicted by GRP model highly correlated with relative fish biomass distribution for most of transects, and contributed as high as 89% variance in the relative fish biomass distribution. Model sensitivity analysis showed that most parameters are linearly related to the growth rate potential and are not sensitive to habitat quality predictions. Regional sensitivity analysis was applied to parameters that are nonlinearly related to growth rate potential and showed that maximum and optimal temperatures for consumption are the most sensitive ones. Out study indicates that GRP model is a good habitat index and can be extended to predict habitat quality for a fish community, when fish community composition is relatively stable.
3. Spatial regression:
Two transects were selected: F2003 and MFT2006. Spatial analysis were carried out on averages of top 2m water, bottom 2m water, and whole water column. Stepwise selection was used to find the significant variables to predict fish biomass. Then collinearity is removed by removing one of the variables if two variable's correlation is greater than 0.7. Correlation coefficients of linear regression between fish biomass and predictive variables varied from 0.34 to 0.92. Autocorrelation analysis (Geoda) showed that fish biomass is highly autocorrelated. We are going to explicitly partition the contributions to variance in fish biomass by spatial structure (SP), environmental variables (EV), SP+EV, unexplained error.
4. Patch size:
Two transects were selected: F2003 and MFT2006 to study the size of longitudinal and latitudinal patches (Zhang et al. 2006). Spatial autocorrelation coefficient, Moran's I, is used to determine patch sizes. The results suggest that a linear gradient dominates the biotic and abiotic variables of transect F from nearshore toward offshore. All variables have high correlations with spatial variable, latitude, linearly or quadratic. Variables along transect MFT show patchy spatial structures. They have weak correlations with spatial variable, longitude, except salinity. The patch size of each variable varied with detrending techniques. Proper trend-removing technique can greatly decrease patch sizes, while the patch size of each variable was generally larger using the original data. For transect F, a quadratic trend dominated the data set, while day-night cycle played an important role in the data of transect MFT. This high-resolution quantification of the variables' spatial variability and patch size can be used for more realistic sampling design and spatial-explicit model setup.
Previous Research Accomplishments
During summer 2003 and 2004, we participated in a research cruise in the northern Gulf of Mexico wherein continuous collections of fish acoustics data (using GLERL acoustics gear), as well as physical (dissolved oxygen, temperature, salinity, density) and lower trophic level (chlorophyll, zooplankton) data, were made across the entire dead zone. During 2006 we also included trawling and diel experiments.
Figure 1. Sampling in the dead zone of the Gulf of Mexico during 2003, 2004 and 2006.
These collections were conducted over the same areas that historically (1985-present) have been sampled by Nancy Rabalais (LUMCON), immediately following her annual surveys, such that we could coordinate data collections in future papers and research proposals. Because our collections were made continuously between Rabalais’ CTD cast sites, we may be better able to define patchiness in hypoxia/anoxia, and its impact on physical and biological attributes.
Research Products
Rae, C., S. Ludsin, K. Hozyash, and D. Kimmel. 2007. Hypoxia effects on fish in the Northern Gulf of Mexico. 21st National Conference on Undergraduate Research, Dominican University of California, San Rafael, CA (contributed poster)
Ludsin, S.A. 2006. Ecological Consequences of Hypoxia in Coastal Systems: Case Studies of Lake Erie, Chesapeake Bay, and the Northern Gulf of Mexico. NOAA-GLERL, Ann Arbor, MI (invited seminar)
References
Aku, PMK, Rudstam LG, Tonn, WM (1997) Impact of hypolimnetic oxygenation on the vertical distribution of cisco (Coregonus artedi) in Amisk Lake, Alberta. Can J Fish Aquat Sci 54:2182-2195.
Aku, PMK, Tonn WM (1997) Changes in population structure, growth, and biomass of cisco (Coregonus artedi) during hypolimnetic oxygenation of a deep, eutrophic lake, Amisk Lake, Alberta. Can J Fish Aquat Sci 54:2196-2206.
Aku, PMK, Tonn WM (1999) Effects of hypolimnetic oxygenation on the food resources and feeding ecology of cisco in Amisk Lake, Alberta. Trans Am Fish Soc 128:17-30.
Breitburg DL, Loher T, Pacey CA, Gerstein A (1997) Varying effects of low dissolved oxygen on trophic interactions in an estuarine food web. Ecol Monogr 67:489-507.
Breitburg, DL, Pihl L, Kolesar SE (2001) Effects of low dissolved oxygen on the behavior, ecology and harvest of fishes: a comparison of the Chesapeake and Baltic systems. In: Rabalais, NN, Turner RE (eds) Coastal hypoxia: consequences for living resources and ecosystems. American Geophysical Union, Washington, DC, p 241-267.
Breitburg DL, Adamack A, Rose KA, Kolesar SE, Decker MB, Purcell JE, Keister JE, Cowan JH (2003) The pattern and influence of low dissolved oxygen in the Patuxent River, a seasonally hypoxic estuary. Estuaries 26:280-297.
Caddy J (1993) Toward a comparative evaluation of human impacts on fishery ecosystems of enclosed and semi-enclosed seas. Rev Fish Sci 1:57-96.
Carr JF, Hiltunen JK (1965) Changes in the bottom fauna of western Lake Erie from 1930 to 1961. Limnol Ocean 10:551-569.
Cloern JE (2001) Our evolving conceptual model of the coastal eutrophication problem. Mar Ecol Prog Ser 210:223-253
Coutant CC (1985) Striped bass, temperature, and dissolved oxygen: a speculative hypothesis for environmental risk. Trans Am Fish Soc 114:31-62.
Diaz RJ, Rosenberg R (1995) Marine benthic hypoxia: a review of its ecological effects and the behavioral responses of benthic macrofauna. Ocean Mar Biol Ann Rev 33:245-303.
Kane DD, Gannon JE, Culver DA (2004) The status of Limnocalanus macrurus (Copepoda: Calanoida: Centropagidae) in Lake Erie. J Great Lakes Res 30:22-30.
Keister JE, Houde ED, Breitburg DL (2000) Effects of bottom-layer hypoxia on abundances and depth distributions of organisms in Patuxent River, Chesapeake Bay. Mar Ecol Prog Ser 205:43-59.
Ketchum, BH (1969) Eutrophication of estuaries. Pages 197-209 in Eutrophication: Causes, Consequences, Correctives. National Academy of Sciences, Washington, DC.
Klumb RA, Bunch KL, Mills EL, Rudstam LG, Brown G, Knauf C, Burton R, Arrhenius F (2004 ) Establishment of a metalimnetic oxygen refuge for zooplankton in a productive Lake Ontario embayment. Ecol Appl 14:113–131.
Kolar CS, Rahel FJ (1993) Interaction of a biotic factor (predator presence) and an abiotic factor (low oxygen) as an influence on benthic invertebrate communities. Oecologia 95:210-219.
Ludsin, SA, Kershner MW, Blocksom KA, Knight RL, Stein RA (2001) Life after death in Lake Erie: nutrient controls drive fish species richness, rehabilitation. Ecol Appl11:731-746.
Marcus NH, Richmond C, Sedlacek C, Miller GA, Oppert C (2004) Impact of hypoxia on the survival, egg production and population dynamics of Acartia tonsa Dana. J Exp Mar Biol Ecol. 301:111-128.
Marcus NH (2001) Zooplankton: responses to and consequences of hypoxia. In: Rabalais, NN, Turner RE (eds) Coastal hypoxia: consequences for living resources and ecosystems. American Geophysical Union, Washington, DC, p 49-60.
Nürnberg GK (1995) The anoxic factor, a quantitative measure of anoxia and fish species richness in central Ontario lakes. Trans Am Fish Soc 124:677-686.
Officer CB, Biggs RB, Taft JL, Cronin LE, Tyler MA, Boynton WR (1984) Chesapeake Bay anoxia: Origin, development, and significance. Science 223:22-26.
Patalas K (1972) Crustacean plankton and the eutrophication of St. Lawrence Great Lakes. J Fish Res Board Can 29:1451-1462.
Rabalais NN, Harper DE, Turner RE (2001) Responses of nekton and demersal and benthic fauna to decreasing oxygen concentrations. In: Rabalais, NN, Turner RE (eds) Coastal hypoxia: consequences for living resources and ecosystems. American Geophysical Union, Washington, DC, p 115-128.
Rabalais NN, Turner RE. 2001. Hypoxia in the northern Gulf of Mexico: description, causes and change. In: Rabalais, NN, Turner RE (eds) Coastal hypoxia: consequences for living resources and ecosystems. American Geophysical Union, Washington, DC, p 1-36.
Roman MR, Gauzens AL, Rhinehart WK, White JR (1993) Effects of low oxygen waters on Chesapeake Bay zooplankton. Limnol. Oceanogr. 38:1603-1614.
Tonn WM, Magnuson JJ (1982) Patterns in the species composition and richness of fish assemblages in northern Wisconsin lakes. Ecol. 63:1149-1166.
Verdonschot PFM (1996) Oligochaetes and eutrophication: an experiment over four years in outdoor mesocosms. Hydrobiologia 334:169-183.
Wannamaker CM, Rice JA (2000) Effects of hypoxia on movements and behavior of selected estuarine organisms from the southeastern United States. J. Expt Mar Biol Ecol 249:145-163.
Wetzel MA, Fleeger JW, Powers SP (2001) Effects of hypoxia and anoxia on meiofauna: a review with new data from the Gulf of Mexico. In: Rabalais, NN, Turner RE (eds) Coastal hypoxia: consequences for living resources and ecosystems. American Geophysical Union, Washington, DC, p 165-184.
Wu RSS (2002) Hypoxia: from molecular responses to ecosystem responses. Marine Pollution Bull 45:35-45.
Zhang, X., M. Roman, D. Kimmel, C. McGilliard, and W. Boicourt. 2006. Spatial variability in plankton biomass and hydrographic variables along an axial transect in Chesapeake Bay. Journal of Geophysical Research -Ocean 111(C5):C05S11.
last updated: 2007-10-18 mbl