Some studies developed in reservoirs of tropical have demonstrated that hydrology and quali-quantitative structures of macrophytes are strongly altered by variations of hydrological regimen. Among the 10th and 20th days of June of 2010, the metrological station of Tapacurá – PE registered 307 mm of precipitation, that characterized a unimodal flood pulse of short duration. Thus, the present study had the following aims: (i) to evaluate the influence of a flood pulse over the richness, floristic composition and biomass of macrophytes in two reservoirs from the of Northeastern Brazil (Cursai and Tapacurá); (ii) and to identify which hydrological variables explained the quail-quantitative patterns of those plants. In order to answering those questions, biotic (floristic composition and biomass) and abiotic (hydrological variables) data were collected semiannually in the periods before and after the flood pulse. For floristic sampling, it was plotted (in each collection) a sample plot of 5 m wide by 40 m long, with the starting point located 0,5 m from the shore. For biomass sampling it was used the technique of sample squares (0,25 x 0,25 m), distributed along of four parallels transects (equidistant in 10 m) and perpendiculars to the shore. For each transect were released three sample squares, which were located at 0,5, 2,0 and 4,0 m from the shore. In the laboratory, biomass samples were washed in running water, dried in a oven at 50oC until constant dry weight and, subsequently, weighed in digital scales, the values expressed in gPS/m2. For each field expedition, water samples were collected for laboratorial analysis of hydrological variables pH and turbidity, and the concentrations of nitrogen and phosphate nutrients. Also in field, hydrological variables were measured: depth, conductivity, dissolved oxygen, temperature and transparency. The Student t test was used to verify significant differences in the values of richness and biomass between pre and post-flood. The evaluation of floristic similarity between pre and post-flood was performed using a cluster analysis. To identify the hydrological variables that best explained the richness variation or biomass between the periods before and after the flood pulse were performed proceedings of multiple linear regression (with partial correlation a posteriori) and canonical correspondence analysis (CCA). 14 species were inventoried, of which 13 spp. occurred in the period after the flood pulse and eight before flood. The Student test t analysis indicated significant differences (p< 0,05) for richness and biomass of macrophytes in Cursai and Tapacurá before and after the flood pulse. Through the Monte Carlo test (0,84) and the percentages of similarity (S), it was found significant differences (p< 0,05) for the specific composition of aquatic plants from the periods pre and post-flood, both for the reservoir of Cursai (S< 30%) and for Tapacurá (S< 50%). The hydrological variables of transparency and dissolved oxygen were identified as predictors of richness variation between the pre and post-flood periods of Cursai reservoir, while in Tapacurá predictors variables were not found. The correlations (obtained by the CCA analysis) showed that the variables dissolved oxygen, temperature, nitrate and total phosphorus were considered explanatory for biomass of macrophytes in the analyzed reservoirs.
