Identification of main variables in forest management for recovery of endemic species habitat in burned areas of southern Chile

Research Article

Identification of main variables in forest management for recovery of endemic species habitat in burned areas of southern Chile

DOI: 10.2989/20702620.2024.2325984
Author(s): Roberto Moreno Instituto de Investigación Sistema Tierra (IISTA), Universidad de Córdoba, España , Ricardo Zamora Universidad de Córdoba, España , Norman Moreno-García Universidad Mayor, Chile , Carlos Esse Instituto Iberoamericano de Desarrollo Sostenible (IIDS), Universidad Autónoma de Chile, Chile , Carmen Galán Instituto de Investigación Sistema Tierra (IISTA), Universidad de Córdoba, España , Juan Ramón Molina Universidad de Córdoba, España

Abstract

Wildfires are among the most serious disturbances that affect forest biodiversity, damaging both flora and fauna. In some burned areas, vegetation recovery is a long and complex process that requires the support of restorative actions. This study evaluated changes in forest structural features after a fire that affect habitat suitability for Rhinocryptids, an endemic group of bird species that are indicators of biodiversity in the temperate forests of South America. The study area was the Malleco National Reserve, Chile, where the main ecosystems are temperate forests. Four species from the Rhinocryptidae family of birds were used as bioindicator species. This group of species is endemic to temperate forests in South America. After fires occurred in 2002 and 2015, affecting temperate forests dominated by several Nothofagus species, two field inventories were carried out in 2016 and 2018. These inventories recorded the presence of Rhinocryptids and some forests structural features. Differences between the two stages of forest regeneration were assessed using statistical tests (T-test and Wilcoxon–Mann–Whitney test) and a classification model was proposed to determine the presence of Rhinocryptidae species. The statistical analysis of forest structural features showed that the variables more relevant in the habitat recuperation are vertical stratification (p-value < 0.001) and understory cover (p-value = 0.053). The trained decision tree to predict the presence of Rhinocryptidae species had an accuracy of 65.22%. This model showed that the coexistence of more than one Rhinocryptidae species occurs when vertical stratification is not mono-stratified and the density is lower than 335 trees ha−1. Our results can help forest managers to consider forest structure as a suitable habitat for certain major wildlife species when implementing post-fire recovery plans.

Get new issue alerts for Southern Forests: a Journal of Forest Science