**2.2. Perennial and semi-perennial crops**

Extensive root systems, long-term soil cover and protection, and reduced need for tillage and weed suppression, give semi-perennial crops excellent choices for bioenergy feedstocks. Crops such as sugarcane, perennial grasses like switchgrass, *Miscanthus* spp. and elephant grass, and trees grown in short rotations tend to have lower water quality impacts than conventional crops [22–24]. While many perennial crops considered for bioenergy have relatively high water use efficiency, their total water requirements can also be relatively large. Such crops are ideally suited to areas with high water availability and flows where water quality can be easily managed [25]. For example, one analysis indicated that that *Miscanthus* spp. could replace 50% of corn acreage in most areas of the Midwest US without adversely affecting the hydrologic cycle. In drier regions, *Miscanthus* spp. should be limited to 25% of the area [26]. Additionally, it has been suggested that the use of perennial grasses may increase seasonal evapotranspiration (ET) compared to grains due to the access of these grasses to moisture deeper in soil profiles [27].

#### **2.3. Forest woody biomass**

Forests provide important regulation of both water quality and seasonally available water quantity in most large watersheds. Forest bioenergy systems are judged compatible with maintaining high-quality water supplies in forested catchments. This general statement is true as long as BMPs that are designed for environment and resource protection, and include nutrient management principles, are followed [28–30]. While short term water impacts, including increased sediment, nitrates, phosphates, and cations can occur, there is no evidence of long term adverse impacts in forest catchments subject to normal management operations [12]. However, more research is needed to guide BMPs concerning special activities in forest management (e.g. stump extraction, weed control, and forest fertilization [29, 31]. Quantitative water flows in forest stands are affected if stands are subject to operations involving significant basal area reductions. But since a forest estate typically is a mosaic of stands of different ages, where only a small share of all stands are harvested in a particular year, water flow regimes on the larger landscape level typically are not affected significantly by stand level operations. Exceptions occur where forests are replaced with other land covers or more intensively managed tree crops.
