Phytoplankton primary production and photosynthetic parameters in reservoirs along a gradient of watershed land use

We investigated how watershed land use (a gradient of agricultural vs. forested land) relates to phytoplankton primary production (PPr) and photosynthetic parameters in 12 reservoirs in Ohio and examined spatial variation in these parameters. Shallow sites near stream inflows had higher light attenuation, total phosphorus (TP), chlorophyll, nonvolatile suspended solids (NVSS), light-saturated photosynthesis (P ), and volumetric PPr than deeper sites near B m dam outflows, but areal PPr and the initial slope of the photosynthesis–irradiance curve (a) were not significantly different between sites. Mean mixed layer irradiance and the severity of light limitation did not differ between sites because shallower depths compensated for higher light attenuation at inflow sites. Watershed land use (percent agriculture) was only weakly (but significantly) related to mean annual PPr, TP, and chlorophyll, but there was a well-defined upper limit to the effect of land use on all three of these parameters. Multiple regression showed that inclusion of additional watershed factors (the ratio of watershed land area to reservoir volume and the ratio of cropland area to number of livestock) greatly increased the variance explained compared to land use alone. TP and chlorophyll were highly correlated with each other and with PPr. Comparison of our TP–chlorophyll, TP–PPr, and chlorophyll–PPr regressions with those of other studies suggests that reservoirs have lower PPr per unit TP than natural lakes, probably because of lower light intensity and higher concentrations of nonalgal P in reservoirs. Watershed land use affects the amount of nutrients exported into lakes and reservoirs via stream inflows. Watersheds dominated by agricultural or urban lands typically export nutrients at higher rates than undisturbed watersheds (Vitousek et al. 1979; Beaulac and Reckhow 1982; Puckett 1995). However, considerable variation exists in the relationship between land use and watershed nutrient export (Mueller et al. 1995; Puckett 1995), as well as in the relationship between nutrient loading rate and eutrophication indicators such as nutrient concentrations, algal biomass, and algal productivity (Carpenter et al. 1998; Correll 1998; Smith 1998). Export of sediments from watersheds is also a function of land use. In particular, agricultural watersheds export considerable quantities of sediment as well as nutri1 To whom correspondence should be addressed. Present address: Department of Fisheries and Wildlife, Michigan State University, East Lansing, Michigan 48824 ([email protected]). Acknowledgments We thank K. Arend, H. Boone, A. Bowling, B. Bunnell, A. Christian, J. Headworth, E. Johannes, J. Pyzoha, and L. Theis for assistance in the field and lab; E. Fee for advice on estimating primary production; S. Harper for help with primary production computer programs; and A. Bowling, C. Glaholt, S. Harper, M. Horgan, K. Sigler, and two anonymous reviewers for comments on an earlier draft of the manuscript. Funding was provided by NSF grant DEB 9726877 to M.J.V. and W.H.R. and a Miami University Undergraduate Research Grant to L.B.K. ents. Loading of sediments can reduce algal productivity by decreasing light intensity (Hoyer and Jones 1983; Kimmel et al. 1990; Knowlton and Jones 1995). Quantitative relationships between watershed land use and indicators of lake eutrophication are exceedingly scarce (e.g., Field et al. 1996; Siver et al. 1999; Arbuckle and Downing 2001; Prepas et al. 2001). This is surprising, considering that land use patterns are changing rapidly in many parts of the world (Sala et al. 2000) and that lakes integrate watershed processes. To our knowledge, no multiple-lake studies have explicitly quantified the relationship between land use and phytoplankton primary productivity. In addition, we know of no studies explicitly relating land use to any eutrophication indicators in reservoirs. In this paper, we explore how agricultural land use in watersheds is related to phytoplankton primary productivity, photosynthetic parameters, and associated eutrophication indicators in reservoirs. Reservoirs can be especially influenced by inputs of nutrients and sediments because they have relatively large watersheds compared to natural lakes (Kimmel et al. 1990; Wetzel 1990). In addition, many reservoirs in the U.S. Midwest reside in agricultural watersheds and likely experience high rates of sediment and nutrient inputs (Kimmel et al. 1990; Knowlton and Jones 1995; Knowlton and Jones 2000). Because of the influence of stream inflows, reservoirs often exhibit pronounced longitudinal gradients in light and nutrients, which can result in gradients in the physiological pro609 Land use and reservoir primary production Fig. 1. Representation of photosynthetic parameters and data from a test of holding time on photosynthetic rate using Acton Lake phytoplankton. Table 1. Characteristics of study lakes and their watersheds and the number of times lakes were sampled for primary production, photosynthetic parameters, and potential predictor variables. Total P and chlorophyll values represent simple means of all dates sampled during the 3 yr. Lake Surface area (ha) Mean depth (m) Total P (mg L21) Chlorophyll (mg L21) Watershed area (km2) Watershed land use (% of watershed area) Agriculture Forest Other No. of dates sampled (1998, 1999, 2000) Acton (A) Alum Creek (M) Berlin (E) Burr Oak (B) Caesar Creek (C) Delaware (D) Knox (K) La Due (L) Piedmont (I) Pleasant Hill (P) Stonelick (S) Tappan (T) 232 935 1560 266 719 453 195 566 982 312 58 964 3.9 7.8 4.5 4.5 11.0 3.5 2.1 3.7 4.5 4.8 2.1 3.3 129.9 27.6 25.9 43.0 48.9 91.7 77.6 50.4 49.8 85.5 153.1 59.1 56.3 11.5 4.9 10.8 10.0 31.8 25.6 15.9 12.2 32.9 14.2 20.8 257 321 640 86 608 1060 80 93 222 512 60 184 88.8 72.4 53.5 13.6 84.3 84.2 67.3 35.2 36.7 50.8 78.4 28.6 9.3 21.8 38.9 80.9 12.5 13.5 28.0 47.1 56.2 44.8 19.4 63.7 2.0 5.9 7.6 5.6 3.3 2.2 4.6 17.7 7.1 4.4 2.2 7.7 7,8,10 2,2,1 2,*0,0 3,*5,7 4,0,0 3,2,1 2,*0,0 2,2,1 1,0,0 2,3,7 4,0,0 2,0,0 * On all dates in Berlin, one date in Burr oak, and one date in Knox, only the outflow site was sampled. cesses underlying photosynthesis and, hence, primary production rates (Kimmel et al. 1990). aB, the slope of the light-limited part of the photosynthesis–irradiance (P-I) curve, and P , the light-saturated photosynthesis rate (Fig. B m 1) are parameters often used to characterize phytoplankton photosynthetic physiology (Fee et al. 1992). aB and P can B m vary with lake trophic status (Fee et al. 1987) and along gradients of light and nutrients. However, we know of no studies that have explicitly investigated the relationships between land use and these photosynthetic parameters. Here, we examine how agricultural land use in watersheds affects primary productivity and photosynthetic parameters among and within 12 reservoirs and relate these variables to light and nutrient concentrations. We sampled sites near stream inflows, where nutrient and sediment inputs are hypothesized to be relatively high, and near the dam where these inputs are hypothesized to have fewer effects. We hypothesize that as watersheds become increasingly dominated by agriculture, primary production, aB, and P will increase B m because of increased nutrient inputs. Furthermore, we hypothesize that both photosynthetic parameters will be higher at inflow sites than at dam sites; P because nutrient availB m ability is higher at inflow sites and aB because phytoplankton at inflow sites should be better adapted to low light conditions.