Slide 24: Species distribution datasets...3
It is usually useful to supplement data with new data where there are gaps, either temporally or spatially. Gradsect sampling is a very useful technique for this purpose, since it maximises sampling across gradients, as described in Wessels et al, 1998.
A recent study of focussed vs random sampling by Hirzel and Guisan in 2002 highlighted some useful pointers. Obviously, a larger sample size is likely to produce more robust results. However, systematic sampling that covers more ground is generally more useful than the traditional random sampling techniques. Furthermore, when environmental information about the region is already known, it is better to include this in the experiment design in order to incorporate as much information as possible.