Assessing coarse woody debris by integrating full area sampling and line intersect sampling: Combining the best of both worlds

Effective and comprehensive monitoring of the quantity of deadwood has become an important aspect of forest inventories for studies on structural and demographic dynamics, biodiversity and carbon stocks. Assessing dead wood quantity, however, is challenging and time consuming due to the structural complexity of deadwood. To monitor coarse woody debris (CWD) in a 10-year remeasuring cycle, we propose to combine full area sampling, to assess and position large-diameter CWD, with line intersect sampling, to estimate the volume of small-diameter CWD. The aim was to simultaneously I) lower the work load in the field, II) ensure low variability in average CWD volume and III) enable remeasurement of a high share of individual CWD objects to study dead wood dynamics and related biodiversity. Using data from 1601 circular plots measured with full area sampling in 16 temperate forest reserves (Flanders, northern Belgium), we simulated line intersect transects and tested threshold diameters between 10 and 130 cm to subdivide CWD to be measured with full area or line intersect sampling. The work load of the combined sampling with a threshold diameter of 20 to 40 cm was about 50 to 90% of the full area sampling work load respectively. Yet, no significant increase in the coefficient of variation (CV) of the average CWD volume was registered for threshold diameters up to 30 cm (133 %) compared to the full area sampling (125 %). The probability to relocate a 30 cm diameter CWD object after 10 years was 50 % and the probability for remeasuring increased with diameter. Using full area sampling with 10 cm threshold results in only 32 % of the objects relocated and remeasured after 10 years; using the combined sampling, only positioning logs over 30 cm increases this figure to 67 %, thus avoiding idle work. We conclude that combining full area and line intersect sampling has the advantage of lowering the work load, increasing the share of relocated objects over time, while not significantly increasing the variability in average CWD volume. An optimal threshold diameter between both methods was comprised between 20 and 40 cm but might further depend on relative importance of work load, need for relocation, CV, the study setup (e.g. number and size of plots), stand characteristics (e.g. CWD volume, dominant tree species) and decay rate. We propose to use a 30 cm threshold diameter for measuring CWD in temperate forest reserves.