Meta-analysis of dead tree volume variable in Hyrcanian forests

In close to nature silviculture, dead trees have great importance because of their essential role in the conservation of biodiversity in forest ecosystems. In this regard, the amount (volume) of dead trees in pristine natural forests can be a guide for its maintenance in managed forests. One of the most important questions in recent years is how much dead trees can be recommended to retention in the managed natural forests. Related resources show different results for this variable. Therefore, in a meta-analysis, studies in Hyrcanian forests were collected and analyzed. This is the first meta-analysis for important structural factor in close to nature silviculture, dead tree volume, in the Hyrcanian forests.

This study includes design and search strategy, collecting articles and extracting data from them, include and exclude criteria and finally statistical analysis of extracted data. Dead tree related research was searched in national and international data base, including SID, Science Direct, Web of Science, Scopus, Google Scholar and IranDoc until May 2022. In this study, after review of several articles, data (mean, standard deviation and sample size) of 16 original studies related to the dead tree volume variable in Hyrcanian forests were collected that had 163 samples and the total area of the sample plots was 92.44 hectares. After extracting the data, Cochran's Q test and I2 index were used to evaluate the heterogeneity the studies and select the appropriate model for data analysis. To find the source of heterogeneity between studies, subgroup analysis was performed based on geographical areas (province) and tree composition. In this study, Beggs and Egger tests were used to investigate the publication bias. Data analysis was done with Comprehensive meta-analysis ver. 3.7.

The mean of dead tree volume was estimated to be 35.293 m3/h (Confidence interval 95%: 24.769 - 45.817) for Hyrcanian forests according to the significance of Cochrane Q test (p = 0.0001) and I2 index (99.735) based on the random effects model. Subgroup analysis showed that the highest dead tree volume was estimated based on a random model in the Golestan province forests as 50.520 m3/h (95% CI: 45.669 - 55.372). In this study, the publication bias test was not significant (p-value for Egger and Beggs tests was 0.28 and 0.24, respectively).

The average of dead tree volume in the Hyrcanian forests was estimated to be "average". Accordingly, the maintenance of at least 35 m3/h of dead tree in Hyrcanian-managed forests can be a guarantee for the preservation of the biological functions of these trees.