Analysis of within location non-replicated multi-locational on-farm yield trials for testing adaptability of crop varieties

Abstract

Within location replication is necessary for efficient analysis of multi-locational on-farm yield trials for testing wide adaptability of crop varieties using presently available methods. However, within location replication is practically difficult, comparatively more expensive and cumbersome even with few varieties in the test. A method to efficiently analyze within location non-replicated multi-locational onfarm yield trials is proposed. In the proposed method, non-replicated trial is conducted in each location and yield deviations of varieties from the maximum yield in each location are computed and an analysis of variance is performed on variety yield deviations pooled over all locations. Mean deviation (d ) cand variance of deviations (s2) across locations are computed for each variety. Varieties are selected simultaneously for significantly lower and lower s2 and the variety with the lowest d and the lowest 2 is the most adaptable varietyin the test. The efficiency, effectiveness and the reliability of the proposed method is illustrated in comparison to the most appropriate presently available method with within location replication taking rice (Oryza sativa L.) as an example.