Comparing a new model-based method to fixed-area sampling for estimating the abundance of standing dead trees

Abstract

Standing dead trees (SDTs) are an important component of forest ecosystems, but reliable estimates of SDT population parameters may be costly, due to spatial and temporal variation in dead tree abundance, which often causes an excess number of zero observations (a.k.a. 'zero inflation') in survey data. The objective of this study was to compare a new estimation method, the Expected-Zero Hurdle (EZ-Hurdle) method, with a typical fixed-area sampling (FAS) method, as methods for estimating the abundance of SDTs under different forest conditions. This study demonstrated that the EZ-Hurdle method provided more precise estimates than FAS when data describing the distance from the plot centre to the nearest SDT were incorporated into a model-based estimator that reduced estimation error associated with zero-inflation in the data. However, the EZ-Hurdle method was determined to be a less cost-efficient sampling method than FAS with smaller (∼7 m radius) plots, because the additional time required to find the nearest dead tree to each plot centre was large, relative to the improvement in precision. On the other hand, the EZ-Hurdle method was a more cost-efficient sampling method than using larger (∼18 m radius) plots. The EZ-Hurdle method appears to be a useful method for improving the precision of estimates of SDT abundance whenever there is a restriction on changing the sampling design or adding more plots to a survey. © Institute of Chartered Foresters, 2012. All rights reserved..