The Great Slaty Woodpecker (Mulleripicus pulverulentus) has experienced a rapid population decline due to the loss of primary forest habitats across its range. Despite being classified as globally Vulnerable, detailed information regarding its status and distribution is largely insufficient and outdated. To address this, we conducted surveys from 2019 to 2021 in the western Terai Arc Landscape of Nepal, covering 29 transects, each 5 kilometers long, to estimate the present population status, nature of excavated cavities, and overall distribution of the species in Nepal. We measured the diameter at breast height (DBH) within 15 meter circular plots at each woodpecker sighting location to explore the relationship between tree diameter and woodpecker presence. Additionally, we modeled the potential distribution of the Great Slaty Woodpecker across Nepal using available occurrence points. A total of 81 individuals were recorded across 14 transects, with 66 individuals within protected areas and 15 outside. Our finding demonstrated a direct correlation between tree DBH and woodpecker presence, indicating that large trees are critical for the species, with an average DBH of 61.26 cm for trees where woodpeckers excavated cavities. Furthermore, we found that the total suitable habitat for the species in Nepal is approximately 6,738 km², with a significant portion located outside protected areas. The habitat in community forests, outside protected areas is particularly vulnerable to selective logging, posing a threat to the species. Therefore, further studies on the impact of logging on the Great Slaty Woodpecker are essential for effective conservation strategies.

Data Collection

Potential habitats for M. pulverulentus were identified through using a combination of literature reviews, expert recommendations, and consultations with local birding groups.Field data were collected through transect surveys, each 5 km length, across five blocks in protected and outside the protected areas.

Surveys were conducted during the breeding season, spanning from March-July, in the years 2019 and 2021. No surveys were conducted in 2020 due to COVID-19 restrictions. Surveys were conducted from 0700 hrs to 1200 hrs in the morning and 1500 hrs to 1700 hrs in the evening, with surveyors walking at a pace of 1 km/hr. We recorded observations of M. pulverulentus individuals sighted or heard during these surveys, defined as “per sighting”. termed per sighting later. Additional observations made outside the designated transects and opportunistic sightings after 2021 were noted and utilized as presence points for distribution modeling and to better understand species’ overall distribution.

We compiled 156 occurrence points of M. pulverulentus collected between March 2017 and May 2024. These data points were sourced from our field surveys, opportunistic sightings, and published records accessed through the Global Biodiversity Information Facility (GBIF).

Habitat Data

At each detection point, a circular plot of 15 m was established to measure the Diameter at Breast Height (DBH) of trees, along with tree height and forest type. Furthermore, when the excavated cavities were identified the DBH of the excavated tree was measured, the number of excavated cavities were counted. The cavity height from the ground was also measured.

Habitat Distribution

We used MaxEnt (version 3.4.1) with bioclimatic variables and occurrence data to model predicted habitat distributions for M. pulverulentus in Nepal. All variables were included after confirming acceptable multicollinearity levels (|r| < 0.70) (Dormann et al., 2013). Presence points were spaced at least 1 km apart to minimize spatial autocorrelation. Model settings included 1,000 maximum iterations and 10 replicates. Model performance was evaluated using threshold-independent and threshold-dependent methods.  AUC values below 0.7 indicate poor performance, 0.7–0.9 suggest moderate performance, and values above 0.9 denote excellent performance (Pearce and Ferrier, 2000). True Skill Statistics (TSS), calculated as Sensitivity + Specificity - 1, assessed threshold-dependent performance, with values ranging from -1 to 1 (Allouche et al., 2006). We averaged TSS scores from all 10 model runs using R software (R Core Team, 2018). Additionally, the habitat map was clipped by land use change to identify the current habitat status across various land use forms in Nepal.