Periphyton biomass and ecological stoichiometry in streams within an urban to rural land-use gradient

This study examined the effects land use on biomass and ecological stoichiometry of periphyton in 36 streams in southeastern New York State (USA). We quantified in-stream and land-use variables along a N-S land-use gradient at varying distances from New York City (NYC). Streams draining different landscapes had fundamentally different physical, chemical, and biological properties. Human population density significantly decreased (r = -0.739; P < 0.00001), while % agricultural land significantly increased (r = 0.347; P = 0.0379) with northing. Turbidity, temperature, conductivity, and dissolved Mg, Ca, SRP, pH, DOC, and Si significantly increased in more urban locations, but NO3 (-) and NH4 (+) did vary not significantly along the gradient. Periphyton biomass (as AFDM and Chl-a) in rural streams averaged one-third to one-fifth that measured in urban locations. Periphyton biomass in urban streams averaged 18.8 +/- A 6.0 g/m(2) AFDM and 75.6 +/- A 28.5 mg/m(2) Chl-a. Urban Chl-a levels ranging between 100 and 200 mg/m(2), are comparable to quantities measured in polluted agricultural streams in other regions, but in our study area was not correlated with % agricultural land. Periphyton nutrient content also varied widely; algal C varied > 20-fold (0.06-1.7 mu mol/mm(2)) while N and P content varied > 6-fold among sites. Algal C, N, and P correlated negatively with distance from NYC, suggesting that periphyton in urban streams may provide greater nutrition for benthic consumers. C:N ratios averaged 7.6 among streams, with 91% very close to 7.5, a value suggested as the optimum for algal growth. In contrast, periphyton C:P ratios ranged from 122 to > 700 (mean = 248, twice Redfield). Algal-P concentrations were significantly greater in urban streams, but data suggest algal growth was P-limited in most streams regardless of degree of urbanization. GIS models indicate that land-use effects did not easily fit into strict categories, but varied continuously from rural to urban conditions. We propose that the gradient approach is the most effective method to characterize the influence of land use and urbanization on periphyton and stream function.