Selection of test sites

The number of sites needed for a successful study often depends on specific site characteristics such as the following  

Ideally, when small mammals are being evaluated, the treated area beyond the test site should be large enough that any recruitment of study species to the test site will be from a treated area. It is not possible to cover this recruitment area for most bird species and it is often impractical for small rodents, but should be considered. Owing to abundance, home range and recruitment dynamics, small passerines and small rodents are often selected as test species. 

It is common for site selection to require several months of intense evaluation. 

Selection criteria (see characteristics above) should be evaluated on-site, during the appropriate season, 1 year before research is to be performed. This timing also allows some time to develop partnerships with landowners, who must cooperate if research designs are to be successful. This is not a simple matter when conducting research with high-value crops such as fruits.  

---

The intrasample error for paper, vellum, and parchment is summarized in Table 1. Note that the coefficient of variance is significantly higher for vellum and parchment than for paper. The selection of test site was far more difficult on skin than on paper. For paper sampling, a fibril was selected which ran perpendicular to the ink line so that the electron beam would scan the surface of a single fibril without falling into the valley between fibers. With skin, the presence of hair follicles and anomalies in surface texture introduced more sampling error. 

---