There are many reasons for the paucity of integrated crop/livestock research and associated publications. Integrated/crop livestock experiments that involve adequate treatments and replications, as perceived by both crop and animal scientists, require large numbers of hectares, many animals, considerable labor to conduct the research, substantial financial resources, and a commitment by Federal and State Research Agencies to fund such long-term research projects. To be truly integrated, crop/livestock research must be multidisciplinary, involving scientists of diverse training and experience with expertise to address various aspects of the research problem, and scientists must function as a cohesive unit or team. The prevailing attitude that all experimental data must be statistically analyzed to be of any value is also a detriment to integrated research. Statistical analyses of these projects may be quite challenging and require new or unusual approaches. Related to the prevailing need for statistical analysis is also the need for scientists to publish senior authored publications for career advancement. Conducting integrated research may not facilitate scientists' publishing the number and quality of publications required for them to meet these criteria. A further obstacle to integrated research alluded to above, involves the many experimental design compromises that must be made by cooperating scientists. Crop and soil scientists for example, use many treatments and replications with small plots, while animal scientists, by necessity, have experiments that involve relatively large numbers of hectares and animal numbers with relatively few treatments and replications. It is therefore extremely difficult to initiate such projects given these inherent differences in crop versus livestock research protocol, as well as to design effective experiments that will provide publishable data. Making compromises on the many factors relevant to integrated crop/livestock research while designing experiments that will provide solutions to pertinent producer problems as well as useful data that can be statistically analyzed and published is, therefore, extremely difficult.
