Improving Pacific Northwest prairie and oak savanna vegetation data collection techniques to increase consistency between groups and time.
Summary
This discussion concerns the use, accuracy, and calibration of ocular estimate techniques within and between organizations collecting PNW prairie and oak savanna vegetation data, and through time. Participants generally agreed that ocular estimates for percent cover are inherently biased and comparing data across institutions and time is an extremely ambitious undertaking. However, several participants were able to provide insight and potential solutions to the problem. These suggestions included: 1) recording change analysis in GIS of aerial photographs from those years with available imagery; 2) recording observer identity which could then be used as a random variable in models to account for consistent bias; 3) good training and collaborative data collection; 4) using different dimensions of pre-cut paper/plastic materials to test and compare observers’ percent cover estimates, with the crew leader knowing in advance what cover the pieces constitute. It was evident from the discussion that there exists a desire to stray from ocular estimates and develop collection techniques that are more objective such as point-intercept methodology, but until these techniques become more monetarily and phenologically feasible collaboration and steady improvements to current protocols will be necessary.