Diversity and community structure of cyanobacteria and other microbes in recycling irrigation reservoirs.

Recycling irrigation reservoirs (RIRs) are emerging aquatic environments of global significance to crop production, water conservation and environmental sustainability. This study characterized the diversity and population structure of cyanobacteria and other detected microbes in water samples from eight RIRs and one adjacent runoff-free stream at three ornamental crop nurseries in eastern (VA1 and VA3) and central (VA2) Virginia after cloning and sequencing the 16S rRNA gene targeting cyanobacteria and chloroplast of eukaryotic phytoplankton. VA1 and VA2 utilize a multi-reservoir recycling irrigation system with runoff channeled to a sedimentation reservoir which then overflows into transition and retention reservoirs where water was pumped for irrigation. VA3 has a single sedimentation reservoir which was also used for irrigation. A total of 208 operational taxonomic units (OTU) were identified from clone libraries of the water samples. Among them, 53 OTUs (358 clones) were cyanobacteria comprising at least 12 genera dominated by Synechococcus species; 59 OTUs (387 clones) were eukaryotic phytoplankton including green algae and diatoms; and 96 were other bacteria (111 clones). Overall, cyanobacteria were dominant in sedimentation reservoirs, while eukaryotic phytoplankton and other bacteria were dominant in transition/retention reservoirs and the stream, respectively. These results are direct evidence demonstrating the negative impact of nutrient-rich horticultural runoff, if not contained, on natural water resources. They also help in understanding the dynamics of water quality in RIRs and have practical implications. Although both single- and multi-reservoir recycling irrigation systems reduce the environmental footprint of horticultural production, the former is expected to have more cyanobacterial blooming, and consequently water quality issues, than the latter. Thus, a multi-reservoir recycling irrigation system should be preferred where feasible.