Effects of forest composition and structure on breeding bird communities in Mount Edough (north-eastern Algeria)

Biodiversity Data Journal 13: e172987 doi: 10.3897/BDJ.13.e172987 Research Article Effects of forest composition and structure on breeding bird communities in Mount Edough (north-eastern Algeria) Zahra Brahmia , Rahma Machtoub , Slim Benyacoub , Nadia Ziane , Rachid Rouag ‡ Department of Biology, Faculty of Science, Chadli Bendjedid University, El Tarf, Algeria § Department of Biology, Faculty of Sciences, Badji Mokhtar University, Annaba, Algeria | 1Laboratory of Environmental Bio-surveillance, Department of Biology, Faculty of Sciences, Badji Mokhtar - Annaba University. 12, P.O.BOX. 2300, Annaba, Algeria ¶ Laboratory of Environmental Sciences and Agroecology, Department of Agronomy, Faculty of Sciences, Chadli Bendjedid University, El Tarf, Algeria Corresponding author: Rachid Rouag ([email protected]) Academic editor: Caio J. Carlos Received: 23 Sep 2025 | Accepted: 18 Nov 2025 | Published: 01 Dec 2025 Citation: Brahmia Z, Machtoub R, Benyacoub S, Ziane N, Rouag R (2025) Effects of forest composition and structure on breeding bird communities in Mount Edough (north-eastern Algeria). Biodiversity Data Journal 13: e172987. https://doi.org/10.3897/BDJ.13.e172987 Abstract This study examines the effects of forest vegetation structure on the richness, abundance and composition of bird communities in four different Mediterranean habitats in northeastern Algeria. Thirty bird species were recorded in the four sampled habitats, belonging to six orders and 14 families. The order Passeriformes is the most represented, with 21 species. Results indicate that cork oak (Quercus suber) and zeen oak (Quercus faginea) forests, characterised by a complex vertical structure and a closed canopy, support a rich and typically forest-dwelling bird community. In contrast, pine (Pinus pinaster) forests exhibit lower species richness and are dominated by generalist species, tolerant to habitat disturbance. Vegetation parameters such as tree height (r = 0.56), shrub cover (r = 0.49) and ecological vegetation volume (r = 0.58) were positively correlated with avian richness. With average species richness values of 18.5 and 16.3 species per plot, respectively, zeen oak forests and mixed forests had the highest levels of avian diversity, while pine forests had the lowest (average = 9.7 species). ‡ § § | ¶ © Brahmia Z et al. This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Keywords birds, assemblage, forest, diversity, Edough, Algeria Introduction Forest birds are widely recognised as reliable indicators of ecosystem integrity and environmental quality. Their sensitivity to variations in habitat structure, resource availability and human-induced disturbances makes them valuable tools for assessing the ecological condition of forest environments (Gregory et al. 2005; Devictor et al. 2008). Many ecological studies have highlighted the significant influence of vegetation structure and composition on biodiversity patterns (Ferry and Frochot 1990; Benyacoub 1993; Blondel 1995). Avian communities exhibit strong habitat specificity; the complexity and composition of vegetation significantly affect their diversity and spatial organisation (Caziani and Derlindati 2000; Brawn et al. 2001). This study aims to investigate the spatial distribution patterns of breeding bird communities in the high-altitude forests of the Edough massif across four distinct vegetation structures. Specifically, we seek to assess the richness, composition and structure of breeding avian communities within these habitats. Bird distribution in mountainous regions is particularly complex due to the wide variety of biotopes concentrated within small areas and pronounced altitudinal gradients (Dejonghe 1984). This region is considered a biodiversity hotspot within the Palaearctic ecozone (Comes 2004) and is also recognised as an area of high endemism for both flora (Hamel et al. 2013) and amphibians (Carranza and Arnold 2004; Veith et al. 2004). In recognition of its ecological value, several initiatives have been initiated with the objective of the classification of the Edough massif as a “Natural Park”, covering an area of 35,432 hectares. This research is motivated by the scarcity of studies on the avifauna of this massif. Existing studies mainly focus on the biology and distribution of specific species, such as the African Blue Tit (Cyanistes teneriffae) and the Eurasian Jay (Garrulus glandarius) (Chabi et al. 1995; Chabi et al. 2000; Ziane et al. 2006; Belabed et al. 2018; Sakraoui 2019; Zeghdani et al. 2024). Conversely, few investigations have undertaken a comprehensive inventory of bird species and evaluated the conservation status of the main forest formations of the massif (Zeghdani et al. 2024). The inventory and assessment of bird communities, particularly breeding species, provide reliable indicators of forest ecosystem health. Within this context, the present study aims to enhance current knowledge on the breeding birds of the Edough region and to assess their long-term conservation status. To achieve these objectives, we implemented a methodological approach combining standardised ornithological surveys with an analysis of forest habitat characteristics. 2Brahmia Z et al Materials and Methods Study site This study was conducted in Edough massif, situated in north-eastern Algeria. Edough massif is part of the grand Tell Atlas chain running through northern Maghreb. Maximum elevation is 1008 m at Mont Bouzizi. The whole massif is marked by steep gradient from its peak to sea level and is characterised by an important forest cover of Mediterranean Cork Oak Quercus suber, Zeen Oak Quercus faginea forests and rich scrubs (Boulahbal et al. 2022). The climate of the region is typically warm Mediterranean, with an important rainfall difference between the mountains and the coast (1200 and 700 mm), respectively. Bird sampling was carried out on the northern slope of the massif, at an altitude between 363 and 510 metres (36°55'38"N 7°41'31"E). Sampling took place in the forest on the 8 km of the Col du Chacal track, which is approximately 5.5 km long and runs along the W16 road between Annaba and the village of Seraïdi. This track crosses a dense forest. The slopes are relatively steep. Facing north, the landscape is carved out by numerous valleys that are home to intermittent torrential streams. Vegetation structure The variation in vegetation structure was used to study the community structure of bird species. Four forest habitats were identified along a 200 m wide, 5 km long sampling transect: - A pine forest (Pinus pinaster) was found to correspond with a rugged terrain formation where bedrock was occasionally visible. The sparse undergrowth is dominated by Retama sphaerocarpa, Calycotome spinosa, Erica arborea and Cistus monspelliensis. - A cork oak forest (Quercus suber), whose undergrowth comprises Calycotome vilosa, Erica arborea, Cistus monspelliensis, Arbutus unedo, Viburnum tinus, Cytisus triflorus, Phillyrea angustifolia and Myrtus communis. - A mixed formation whose tree layer includes zeen oak (Quercus faginea), cork oak (Quercus suber) and maritime pine (Pinus pinaster), with dense undergrowth consisting of Calycotome vilosa, Erica arborea, Cistus monspelliensis and Cistus salviifolius. - A pure zeen oak (Quercus faginea) forest, characterised by trees sometimes exceeding 30 metres in height and relatively dense, tall undergrowth consisting of Calycotome vilosa, Erica arborea, Cistus salviifolius, Arbutus unedo, Viburnum tinus and Phillyrea angustifolia. (Fig. 1). Environmental variables To assess the influence of habitat structure on breeding bird richness, a set of vegetation parameters was measured within each sampling plot. The average height (TLH) and cover (TLC) of trees and the average height (SLH) and cover (SLC) of shrub vegetation th Effects of forest composition and structure on breeding bird communities ... 3 were evaluated in every station. The DBH (diameter at breast height) was also calculated. This variable is used to monitor tree growth and calculate both net production and basal area of forest plots. During each bird census, we used a tape measure to measure the diameter at DBH of about ten trees in the station. An index of Ecological Vegetation Volume (EVV) was then calculated to quantify the vertical complexity of the vegetation. EVV combines the mean height and the cover of each vegetation layer according to the formula: - where Hi and Ci represent the mean height and cover percentage of layer i, respectively. This index provides an integrated measure of structural complexity reflecting both vertical and horizontal vegetation development. Avifauna sampling The point count method is a widely used technique for sampling birds across the world due to its effectiveness and easy implementation (Diefenbach et al. 2003). This method allows us to estimate species richness amongst different habitats (Barlow et al. 2007, Castaño-Villa et al. 2014). We estimated species densities using conversion coefficients calculated by Benyacoub (1993). The study site was surveyed weekly from April 2019 to mid-May 2019, for all habitat types. We performed 24-point counts (Suppl. material 1). For each point count, we used a timed count of 20 min in a radius of 10 m. All counts were made by the same person (ZB) to avoid inter-observer bias. Sampling was done mainly during the morning (6:30 to 10:30 am) when birds are found to be most active. Each point was separated by at least 150 m from all other points to minimise the probability of Figure 1. Location of the study site in the Edough massif.  4Brahmia Z et al sampling the same bird more than once. We estimated species densities using conversion coefficients calculated (Benyacoub 1993) by converting IPA values to the number of pairs/10 ha to allow comparison between habitats. Data Analyses To examine the bird community composition, we used species richness “S” that is defined as the total number of species found in a given habitat. To compare the composition of breeding bird communities amongst habitats, a Venn diagram was performed to plot common and biotope-specific bird species of each biotope (Heberle et al. 2015). The diversity was calculated using the diversity index of Shannon-Weaver. Richness was correlated with structural parameters using Pearson’s correlation coefficient (r). Pearson’s correlation coefficients were computed using SPSS software (version 11.0.1). We calculated the asymmetrical Hellinger distance between sampling sites using R 4.5.0. Hellinger distance is a highly recommended distance for analysis, based on abundance data (Legendre and Legendre 2012). To explore the relationship between bird community composition and forest structure, a redundancy analysis (RDA) was performed by using Statistica (14.0.1.) software. It allows the direct identification ecological interpretation of species-habitat relationships (ter Braak 1986). Results Structure of vegetation The four study stations are presented as a succession of vegetation where the degree of coverage and height evolve more or less positively and regularly from one station to another (Table 1). PIN SUB MIX ZEEN TLH (m) 10.97±3.71 14.25±6.27 17.2±6.70 20.01±3.68 TLC (%) 35±14.14 38.38±5.02 46.25±12.48 45.71±7.50 SLH (m) 2.35±0.07 3.14±0.06 3.07±1.26 3.12±1.23 SLC (%) 73.5±9.19 62.85±21.57 56.3±24.18 52.46±17.14 DBH (cm) 52.59±15.13 43.91±12.12 48.59±12,20 56.94±12.26 VEV 414.13±13.97 659.89±206.41 926±146.68 1328.92±294.38 TLH: Tree layer height; TLC: Tree layer cover; SLH: Shrub layer height; SLC: Shrub layer cover; DBH: Diameter at breast height; EVV: Ecological vegetation volume. Table 1. Vegetation structure parameters at the four sampled habitats (Mean ± SD). Maritime pine forest (PIN); cork oak forest (SUB); Mixed formation (MIX); zeen oak forest (ZEEN). Effects of forest composition and structure on breeding bird communities ... 5 The analysis of vegetation structural parameters revealed differences amongst the four forest types studied. Tree layer height (TLH) varied considerably, being lowest in pine forests (10.97 ± 3.71 m) and highest in Zeen oak forests (20.01 ± 3.68 m). Tree layer cover (TLC) followed a similar trend, with higher values in mixed and Zeen oak forests (46.25 ± 12.48% and 45.71 ± 7.50%) and lower in pine forests (35 ± 14.14%). Shrub layer height (SLH) was lowest in pine forests (2.35 ± 0.07 m) and highest in cork and Zeen oak forests (≈ 3.1 m). Conversely, shrub layer cover (SLC) was greatest in pine forests (73.5 ± 9.19%) and decreased in denser forests, especially Zeen oak (52.46 ± 17.14%). Mean tree diameter at breast height (DBH) ranged from 43.91 ± 2.12 cm in cork oak to 56.94 ± 12.26 cm in Zeen oak forests. The Ecological Vegetation Volume (EVV) increased steadily from pine (414.13 ± 13.97) to Zeen oak forests (1328.92 ± 294.38). Birds diversity The avifaunal composition consists of 30 bird species spread across six orders and 14 families. The order Passeriformes is by far the most represented, with 21 species accounting for 70% of the total. Four orders have only one species each: Piciformes (three species; 10%), Columbiformes (two species; 6.67%) and Accipitriformes (two species; 6.67%). Coraciiformes and Galliformes were the least represented taxa, accounting for just 3.33% of the total diversity. Fringillidae is the most species-rich family with six species (20%). Paridae, Muscicapidae and Picidae follow with three species each (10%). The other families are represented by one or two species each. According to their phenological status, 3.33% of the birds were unintentional visitors, 6.67% were winter visitors, 6.67% were passage migrants and 63.33% were resident breeders. According to the IUCN Red List, all of species (100.00%) were categorised as least concern. The mixed formation (26.25%) and the Pine forest (13.75%) had the lowest richness and were followed by the Cork oak forest and Zeen oak forest, which had 30.00% and 30.00% of habitat-specific records, respectively. The Zeen oak forest was the richest habitat, with 24 bird species (30.00% of all occurrences) (Table 2). Species distribution across habitats The Venn diagram below shows the distribution of bird species across four forest habitats. It reveals clear patterns of species overlap. Of the 30 species observed, 30.0% were found in all four habitats. This core group comprises generalist, including the European Robin (Erithacus rubecula), the Common Blackbird (Turdus merula) and the Great Tit (Parus major). Some species are strongly associated with the Cork oak habitat, as evidenced by its 10.0% contribution of unique species, including the Western Bonelli's Warbler (Phylloscopus bonelli) and the Orphean Warbler (Curruca hortensis). The Common Chaffinch (Fringilla coelebs) and the European Greenfinch (Chloris chloris), for Table 2. Composition of bird species in the Edough massif by family, habitat, conservation status and phenological status (LC = Least Concern, VU = Vulnerable). 6Brahmia Z et al example, demonstrate wide ecological tolerance by inhabiting all or most habitats (Fig. 2). Communities structure The structural indices of the various forest habitats clearly show a gradient ranging from evergreen formations, represented by cork oak, to deciduous formations, represented by zeen oak. In general, pine forests are the least diverse, with only eight species. There is a steady increase in richness, diversity index and evenness index in the three other climax formations, from the cork oak forest to the zeen oak forest (Table 3). PIN SUB MIX ZEEN S 8 21 19 24 H' 2,28 3,44 3,57 3,63 Forest environments with multiple tree layers have the highest species densities. Deciduous environments have significantly higher densities than evergreen environments. Zeen oak has the highest total density, which is 25% higher than Cork oak and almost three times higher than Pine forest. Figure 2. Venn diagram of bird species percentage occurring over the four habitats.  Table 3. Species richness and Shannon diversity and equitability per forest types. Effects of forest composition and structure on breeding bird communities ... 7 The density of birds varied significantly depending on the type of habitat. Zeen oak forest, with its structurally complex vegetation, had the highest total breeding pair density (88.05 pairs/10 hectares). Mixed forest (78.46 pairs per 10 hectares) and (Cork oak 70.09 pairs per 10 hectares) followed. Pine forest had the lowest density (40.94 pairs/10 ha) (Fig. 3). Clear patterns of species dominance and composition existed within the bird populations of the four forest habitats (measured in pairs per 10 hectares). The European robin (Erithacus rubecula) was the most common species in all habitats. In pine forests, eight pairs were found per 10 hectares, whereas in Quecus suber environments, this figure was 14.6 pairs per 10 hectares. This demonstrates its ability to adapt to a wide range of environments, indicating a generalist approach. In contrast, the Sardinian warbler (Sylvia melanocephala) had a high density in pine forests (19.5 pairs per 10 hectares), suggesting a preference for open or conifer-dominated habitats. Conversely, its low density (0.7 pairs per 10 hectares) in zeen oak areas suggests that it avoids these habitats. Blackcaps (Sylvia atricapilla) and blue tits (Cyanistes teneriffae) are found in large numbers in many types of habitat, but are most prevalent in zeen oak and cork oak woodlands. This shows that they favour wooded environments with complex structures. Mixed forests with lots of undergrowth also have high densities of chiffchaffs (Phylloscopus collybita) and wrens (Troglodytes troglodytes). Overall, Pine forests hosted fewer species with high densities, notably the Sardinian Warbler (Curruca melanocephala) and the Robin (Erithacus rubecula), while Zeen forests exhibited the highest densities for most species, indicating a richer and more structurally diverse environment. The structure of bird communities was also explored using a heatmap, based on the Hellinger transformation to visualise variations in relative abundance amongst different Figure 3. Total breeding density (pairs/10 ha) across habitat types.  8Brahmia Z et al forest habitats. The heatmap clearly shows differences in species composition across habitats. Mixed forest and Zeen oak forest exhibit the highest relative abundances and species diversity. Generalist species, such as Turdus merula, Cyanistes teneriffae, Sylvia atricapilla and Erithacus rubecula, maintain consistently high abundances across all habitats, indicating ecological plasticity. Conversely, forest-specialist species, such as Periparus ater, Certhia brachydactyla, Regulus ignicapilla and Columba palumbus, are more strongly associated with denser, more structurally complex habitats. Pine forest and cork oak forests display lower species richness and are dominated by Fringillidae and Sylviidae taxa, such as Sylvia melanocephala, Fringilla coelebs and Serinus serinus. Overall, the heatmap shows an increasing gradient of species richness and abundance from pine to Zeen oak forests, which is consistent with the progressive growth in forest structural complexity (Fig. 4). Vegetation structure and relationships with bird community Our analysis revealed clear patterns linking vegetation structure to bird species richness across the sampled Mediterranean forest habitats. The scatterplots highlighted that vertical structural features of vegetation are strong predictors of avian diversity. The Figure 4. Heatmap with Hellinger-transformed bird abundance values across forest habitat types (PIN, SUB, MIX, ZEEN).  Effects of forest composition and structure on breeding bird communities ... 9 • Rocamora G (1987) Biogéographie et écologie de l’avifaune nicheuse des massifs péri méditerranéens d’Europe occidentale. Ecole Nationale Supérieure Agronomique. Montpellier. • Sakraoui W, et al. (2019) Does increasing nestboxes size affects breeding success of blue tits (Cyanistes caeruleus)? study in a Locality of Seraidi (Northeast, Algeria). Ecology, Environment and Conservation 25 (2): 609‑613. • ter Braak CF (1986) Canonical Correspondence Analysis: A New Eigenvector Technique for Multivariate Direct Gradient Analysis. Ecology 67 (5): 1167‑1179. https://doi.org/ 10.2307/1938672 • Veith M, Mayer C, Samraoui B, Barroso DD, Bogaerts S (2004) From Europe to Africa and vice versa: evidence for multiple intercontinental dispersal in ribbed salamanders (genus Pleurodeles). Journal of Biogeography 31 (1): 159‑171. https://doi.org/10.1111/j. 1365-2699.2004.00957.x • Willson M, Comet T (1996) Bird Communities of Northern Forests: Ecological Correlates of Diversity and Abundance in the Understory. The Condor 98 (2): 350‑362. https://doi.org/ 10.2307/1369153 • Zeghdani N, Prodon R, Benyacoub S (2024) Resilience of Bird Communities After Wildfires in Algerian Cork Oak Forests. Acta Ornithologica 58 (2). https://doi.org/ 10.3161/00016454ao2023.58.2.004 • Ziane N, Chabi Y, Lambrechts M (2006) Breeding Performance of Blue Tits Cyanistes caeruleus ultramarinus in Relation to Habitat Richness of Oak Forest Patches in NorthEastern Algeria. Acta Ornithologica 41 (2). https://doi.org/10.3161/068.041.0201 Supplementary material Suppl. material 1: Abundance and density of birds by habitat Authors: Rouag R. Data type: Abundance Download file (15.88 kb) 16 Brahmia Z et al