Rotation experiment & on-farm trials

Mother and Baby research design

Rotation experiments have to be multi-year and tend to be complex in terms of numbers of treatments.  Our rotation study is no exception.  We were interested in comparing rotation length (2 versus 4 years), choice of crops grown between strawberry crops (broccoli or lettuce based rotations) and different fertility/disease management strategies (legume/cereal winter cover crop; legume/cereal winter cover crop + compost + feathermeal as needed based on PSNT soil test; cereal cover crop + mustard seed meal; and an untreated bare winter fallow). This makes 16 treatments, and when replicated 4 times a total of 64 plots.  Such a replicated trial tells us a lot about treatment effects, but we can realistically only do this at one location.  Hence the Mother and Baby design.  The replicated trial is the "Mother" trial, and a subset of the treatments are then reproduced on multiple farms as unreplicated treatments - the "Baby" trials.  This allows us to see how the sub set of treatments performs at a farm scale, across locations and under the management of many farmers.  In this project our farmer partners chose a set of 4 treatments to test on their farms and compare against their standard practices.



Reference for Mother and Baby design

For a discussion of the Mother and Baby design please see:

Snapp, S., G. Kanyama-Phiri, B. Kamanga, R. Gilbert, and K. Wellard. 2002. Farmer and researcher partnerships in Malawi: Developing soil fertility technologies for the near-term and far-term. Experimental Agriculture 38:411-431.