Resumen
Aim of study: To better estimate the annual allowable cut reserve (AACR), taking into consideration the endemic windthrows (EW), we combined a series of existing algorithms into a coherent methodology to use the data available at district level, without any additional fieldworks.Area of study: The algorithm was tested on the EW occurred in the last 20 years in Brosteni FD (Eastern Carpathians, Romania) that covers 21,013 ha and we found that every year from an AAC of 37,000 m3 no more than 2,700 m3 shall be spared for EW that might occur next year.Material and methods: We considered three EW enabling factors (stand slenderness, location on pits and mounds, and the vicinity of canopy gap) and three contingency tables of the EW produced between 1999 and 2008, one for each 40-year age group. Then we calculated a Bayesian model for all six permutations of enabling factors, each of them being tested on the data referring to 2008-2017 periodResults: Plugging the posterior EW likelihoods into a Markov chains (MC) model, we produced a formula that enables a better estimation of the optimal AACR that could be replaced with salvage cuttings every next year. Other options of using the EW likelihoods are also presented at length, such as the type of age-class structure that requires no AACR, that is a ?U? shape age structure, as well as a rough assessment of the additional demand for seedlings needed to re-plant the stands affected by EW. The relatively short period of time the input data refer to, which is ten years, equals the time window of the forest planning and this parity allows a ten-year forecast period, enough for modeling the stationary age-structure of even-aged forests.Research highlights: A new model for optimizing the annual allowable cut (AAC) in even-age forests in the context of endemic windthrows (EW) scenario has been developed and evaluated.Keywords: Bayes? rule, forest management planning, endemic windthrows.