Effects of water table, clipping, and species interactions on Carex nebrascensis and Poa pratensis in riparian meadows

In this study, we hypothesized that the primary variable determining specie s responses and interactions within wet/mesic riparian meadows in central N evada, USA was the water table but that the direct and indirect effects of livestock grazing modified both species responses and interactions. We test ed this hypothesis for two widespread riparian species, Carex nebrascensis and Poa pratensis ssp. pratensis, that co-occur but have maximum expression at different water tables. Carex nebrascensis has widely spread tillers ty pical of 'guerilla' plant architecture, while P. pratensis has closely spac ed and compact tillers typical of 'phalanx' plant architecture. Individuals of both species were grown at mid- and low water tables with or without ne ighbors and were either clipped or not clipped at the end of the first grow ing season. For the study site with the most complete record, rnean water-t able depth during the growing season (May through August) on the mid sites was -32 cm in year 1 and -7cm in year 2. Water-table depth on the low sites was -69 cm in year 1 and -31 cm in year 2. Plant survival tillering, bioma ss, and seed production over a 2-yr period were used to quantify the specie s responses. Water table had no effect on tillering or biomass of C nebrasc ensis, indicating that it is adapted for growth and persistence over the ra nge of water tables examined. In contrast, growth and tillering of P. prate nsis was severely restricted at more shallow water tables. Poa pratensis ha d about 50% fewer tillers and lower biomass (9.0 vs 46.5 g) for neighbor-re moved plots on mid- than low-water-table sites at one of the study meadows. Further, tiller numbers of P. pratensis increased over time on the low-wat er-table plots, but decreased on the mid-water-table plots. Clipping had no effect on the survival, tillering, or growth of either species. The clippi ng treatment may not have removed sufficient leaf mass or may have been app lied too near the end of the growing season to elicit a response. Neighborh ood removal resulted in a 3- to 10-fold increase in tillering and higher pl ant mass (1.3 vs. 9.5 g) for C. nebrascensis. Poa pratensis showed an even greater response to neighbor removal. Tillering was 6-to 100-fold greater a nd mass was 15 to 50 times greater on neighbor removed than neighbor-intact plots. Comparisons between single species and mixed species plots indicate d that there was an interaction between the species that limited tiller pro duction in C. nebrascensis. These results indicate that P. pratensis, the ' phalanx' species, is capable of more rapidly responding to disturbances tha t remove neighbors and increase available space than C. nebrascensis, the ' guerilla' species. Contrary to previous studies, the interactions between t he two species do not seem to be related to plant architecture and can be b est attributed to generally greater growth rates and increased competitive ability for P. pratensis at lower water tables. Grazing may further alter t he relative competitive ability of the two species in favor of P. pratensis .