Determining drought frequency and intensity on the basis of plant response: Wild hay in the sand hills of Nebraska, U.S.A.

Abstract

A definition of agricultural drought which relies on plant response to weather conditions was tested for the North Crop-Reporting District of north-central Nebraska, U.S.A. The period 1925–1979 was studied. The frequency of drought affecting production of wild hay, the principal forage, was determined. The occurrence of negative yield departures of more than 10% below that expected under normal weather conditions was used to define a drought year. This definition had previously been applied to the major grain-producing areas of the U.S.A. by McQuigg. (1973). The applicability of this type of drought definition to a forage crop in a sub-humid environment was considered uncertain when we began the analysis. Where no trend in yield exists, this frequency analysis procedure can be performed independent of climatic data. Drought events were segregated on the basis of their severity. No trend in wild hay yield data was determined to exist. The incidence of negative yield departures greater than 10%, but less than 20%, from that expected under normal weather conditions defined the ‘moderate drought’ years. Negative yield departures of 20% or more defined the ‘severe drought’ years. The single-year probabilities of drought occurrence for wild hay were 0.15 for severe drought and 0.05 for moderate drought on a district-wide basis. A county-level analysis of drought frequency for wild hay production revealed less sensitivity to drought in the drier western portion of the district. The probability of severe drought in a single year was as much as six times greater in the more humid eastern sections of the district than in the drier west. This seeming anomaly is due to the fact that a high proportion of wild hay is produced on sub-irrigated wet meadowland in the western counties. This distinct ecological zone disrupts the expected drought pattern for the region and suggests that agricultural drought frequency is best assessed at the smallest geographical unit for which yield data are available.