Amy, Sounds like an interesting field situation. if you need to quantify vegetation patchiness within your 15m circular plots, then I recommend you do some exploratory analysis first before designing your protocol. For example, you could perform multiple variogram analyses, using a range or sequence of quadrat sizes (e.g. 10, 20, 30, 40cm, etc) to determine the scale at which patches of herbaceous vegetation are most apparent. This would allow you to choose the most suitable quadrat size for your sampling protocol. Also, you may need to divide your protocol into forested, meadow, and riparian classes as the processes that determine patch size are likely to be different for the different vegetation types.
Have you considered that physiognomy may be more important to owl foraging than the herbaceous community composition? Does the owl use forest trees as a perch to scan for small mammals in the meadows? If so, then might not average tree height or distance between nearest neighbors be important variables? Likewise, there may be a minimum (or optimal) meadow area for the owls (related to prey carrying capacity or length of the field of vision of the predator). Another possibility is that foraging quality is governed by the length of the forest-meadow ecotone. From your description it seems you have yet to decide on how you will analyze your data statistically. I recommend you determine this first, and let your decision guide your data collection protocol. I recommend you look at the following texts for additional guidance on measuring and analyzing vegetation patterns: * Dale, MR.T. 1999. Spatial pattern analysis in plant ecology. Cambridge University Press, New York. * Legendre and Legendre. 1998. Numerical Ecology. Elsevier Scientific Publ. Co. Amsterdam. Good luck with your field techs! Mike Amy Williams wrote: > I am putting together a vegetation protocol for measuring plant > communities > and general habitat characteristics associated with known > observations of a > Sierra Nevada owl. My question is: how best can one capture patchy > vegetation in 15-m radius circular plots (associated with foraging > points)? > Foraging habitat for this owl, while often at meadow edges, is quite > variable (e.g. sometimes in very small drainages further in the > forest, on > dry upland logged areas, larger riparian non-meadow areas, etc.). > Keeping in > mind that vegetation within the plot is not homogeneous (the plot > being > often, but not always, half meadow and half forest) is there a > statistically > sound method for capturing vegetation, especially herbaceous? I > would like > the protocol to be user-friendly for non-botanist field crew. > Because of > this I would like to avoid visual cover estimation or any other > subjective > method of data collection. Would it be feasible to use a 10-cm square > sub-quadrat and record presence/absence of individual species (or > where > botany skills falter, type: forb, grass, sedge) and average height > class at > each meter point on each of the four cardinal direction transects > (= 60 > points)?
