ESTIMATING SEED PRODUCTION OF COMMON PLANTS IN SEASONALLY FLOODED WETLANDS

We developed a technique to quickly estimate seed production of common moist-soil plants because previously reported methods were too time consuming to be of value to waterfowl resource managers. Eleven regression equations were developed for 13 plant species in the upper Mississippi Alluvial Valley and the Rio Grande Valley, New Mexico. Estimated time to collect a sample was 1.5 minutes. Easily measured vegetation characteristics such as inflorescence number, inflorescence length, and plant height were used as independent variables to estimate seed mass of known mass samples. Coefficients of determination (R2) ranged from 0.79 for rice flatsedge (Cyperus iria) to 0.96 for smartweeds (Polygonum spp.). The accuracy and precision of equations tested using independent data indicate that the technique can be used to detect changes in seed mass of moist-soil plants in seasonally flooded impoundments. Because of the small sample area per plot used (0.0625 m2) and changes in the density of plants within an impoundment, we recommend that as many samples as economically feasible be collected to reliably estimate seed production.