Converting visual census data into absolute abundance estimates: a method for calibrating timed counts of a sedentary insect population

Abstract.  1. Visual surveys for small organisms on complex substrates often yield serious underestimates of true counts. When both visual counts (relative estimates of abundance) and absolute counts can be obtained from the same sample, however, the visual counts can be calibrated such that absolute estimates can be obtained in the future from visual surveys alone.
2. A method is presented for converting quick, timed, visual counts of a sedentary insect on a shrub into absolute estimates of abundance.
3. Analogies were drawn from simple, well-known predation theories to develop a two-parameter non-linear model. Parameter estimates were obtained by both inverse prediction and direct estimation methods; the latter were found to yield more accurate predictions of absolute abundance.
4. The calibration model is mechanistic in its approach, and thus has potential for application in other systems in which all individuals are visible, but able to be missed during timed counts.