Multivariate classification of farming systems for use in integrated pest management studies

Abstract

. 1999. Multivariate classification of farming systems for use in integrated pest management studies. Can. J. Plant Sci. 79: 647-654. On-farm studies provide a realistic setting to examine the impact of interactions of management practices on weed communities within various farming systems. Clustering farm units into groups that use similar management practices enables the replication of farm management systems within on-farm studies. The goal of this study is to objectively classify farm units into management systems on the basis of quantitative variables describing aspects of cropping history and chemical input levels using multivariate techniques. Twenty-eight Saskatchewan farmers provided details of their management practices from 1990 to 1997 through a series of questionnaires. Twelve variables derived from the questionnaires were used to describe cropping history, pesticide, tillage and fertilizer use on each farm unit. These variables were used to cluster farm units using minimum variance classification and NMS ordination. Both techniques identified seven farm management systems. The greatest differences were observed between organic and non-organic systems. Farm management systems that used annual fallow and continuous annual cropping histories were most similar. The consistent results obtained by use of the two unrelated methodologies indicate the utility of this approach for the classification of farm management systems.

A typology of farm households for the Umutara Province in Rwanda

“…It is known that the adoption of new technology may vary among farm households because of differences in socio-economic characteristics (De Graaff 1996;Leeson et al 1999;Solano et al 2000;Mahapatra and Mitchell 2001;Asfaw and Admassie 2004;Somda et al 2005;Milán et al 2006). A farm typology study can be used to classify farm households based on socio-economic characteristics that may affect the adoption of new technology.…”

Methods to conduct and analyse weed surveys in arable farming: a review

“…Frequently, this could be overcome by installing plots with no herbicides, although this method restricts the survey site to small parts of the fields. Sampling after herbicide application is unproblematic in this regard and appropriate when the overall effect of farming systems reflecting the impact of all typical agronomic practices is to be examined (Leeson et al 1999;Pinke et al 2011Pinke et al , 2012. A time-consuming but informative alternative is to compare the weed flora in sprayed and unsprayed parts of the fields (Sutcliffe and Kay 2000) or before and after herbicide application (Saaverda et al 1990;Hidalgo et al 1990;Fried et al 2008).…”

“…The clustering of weed species into functional groups according to a variety of species traits enabled Fried et al (2009) to search for trait combinations that determine the success of a weed species under specific selection pressures. Other approaches that cluster management and environmental data rather than species data target the classification of farm management systems or eco-regions to be then related to weed vegetation (Leeson et al 1999;Hanzlik and Gerowitt 2011). While in other scientific fields (e.g.…”

“…Alternatives to haphazard sampling (often called 'random' though not following a randomisation scheme) are systematic or regular sampling with fields spaced at regular intervals (grid design) (Chancellor 1977;Firbank et al 1998) or subjective sampling with fields chosen to fit specific criteria, e.g. to 'represent the diversity of cultural practices and environmental conditions present' (Fried et al 2008) or fit desired farming systems (Leeson et al 1999). While caution is recommended when subjectively sampled data are used to make objective statements regarding the study area as a whole, systematic sampling of fields was found to yield data that are comparable to those of random sampling for large data sets and even more precise when sampling is done along (environmental) gradients (Hayek and Buzas 1997).…”

“…The results of a large number of different sites enable to assess these effects in a more realistic way (Andersson and Milberg 1998;Hallgren et al 1999;Gunton et al 2011) and thereby assist in any implementation of the experimental outcome. Such observational or 'on-farm' research proved eligible for the exploration of complex problems like the assessment of farming systems differing in cropping histories, chemical input level and tillage operations with respect to their impact on weed communities (Leeson et al 1999(Leeson et al , 2000.…”