Cavity-nesting birds show threshold responses to stand structure in native oak forests of northwestern Tunisia

• Published in: Forest ecology and management Vol. 325; pp. 1 - 7
• Main Authors: Touihri, Moez, Villard, Marc-André, Charfi, Faouzia
• Format: Journal Article
• Language: English
• Published: Kidlington Elsevier B.V 01.08.2014; Elsevier
• Subjects: Alterations; Animal and plant ecology; Animal, plant and microbial ecology; Biological and medical sciences; Birds; Certhia brachydactyla; Conservation targets; Dead wood; Dendrocopos major; Density; Ecological thresholds; Ecology; Ficedula; Forest harvesting; Forest management; Forestry; Forests; Fundamental and applied biological sciences. Psychology; Mathematical models; Mediterranean forest; Synecology; Terrestrial ecosystems; Thresholds; Woodpeckers; Tunisia; North Africa; Africa; Forest harvesting; Conservation targets; Dead wood; Woodpeckers; Ecological thresholds; Mediterranean forest; Organic matter; Conservation; Cavity; Dicotyledones; Angiospermae; Hardwood forest tree; Aves; Primary forest; Structure; Threshold; Nesting; Mediterranean vegetation; Harvesting; Forest ecology; Woody plant; Fagaceae; Response; Vertebrata; Northwest; Quercus; Forestry; Spermatophyta
• ISSN: 0378-1127; 1872-7042
• DOI: 10.1016/j.foreco.2014.03.050

•Conservation of forest birds requires a quantitative knowledge of their ecological requirements.•We surveyed forest birds at stations representing a gradient in human disturbance (fire, grazing, firewood collection).•We detected habitat thresholds corresponding to a 90% probability of detection for five target species.•We propose conservation targets for the conservation of sensitive oak forest birds. The detection of thresholds in forest bird response to gradients in forest alteration is a powerful approach to quantify their ecological requirements and to develop evidence-based targets for conservation. For this purpose, we analyzed the response of 31 forest bird species to a gradient in forest alteration by human activities (e.g. fire; grazing; firewood collection) in the Kroumirie region of northwestern Tunisia. We surveyed forest birds using point counts at 48 stations located throughout Feija National Park. From a data set of 12 vegetation variables, we obtained a gradient in forest alteration using a principal component analysis. Then, we modeled the probability of detection of bird species along this gradient using logistic regression. We selected the five most sensitive species as target species as indicated by the area under the receiver-operating characteristic curve (AUC). Then, we quantified thresholds in their probability of detection as a function of four habitat structure variables: density of large-diameter trees (dbh>30cm), canopy closure, density of snags, and downed woody material. Nineteen of 31 species exhibited a significant response to forest alteration (PCA1), including 7 positive and 12 negative responses. Among these, the five best models (AUC>0.8) corresponded to Levaillant’s Woodpecker (Picus vaillantii), Great Spotted Woodpecker (Dendrocopos major), Lesser Spotted Woodpecker (Dendrocopsminor), Short-toed Treecreeper (Certhia brachydactyla), and Atlas Flycatcher (Ficedula speculgera). On the basis of their threshold values, we recommend to protect stands with densities of large-diameter trees of at least 650stemsha−1, and 207stemsha−1 of snags, to meet the structural requirements (90% probability of detection) of the most demanding species, respectively Levaillant’s Woodpecker and Atlas Flycatcher.