Structural and compositional differences in gallery and spiny forests of Southern Madagascar: Implications for conservation of lemur and tree species

Abstract

Madagascar's unique dry forests, particularly gallery and spiny forests, face severe threats and are significantly understudied, leaving only a fraction of the original extent intact. Thus, there is a critical need for characterizing, conserving, and restoring this diverse forest ecosystem. Conducting extensive floristic surveys and environmental analyses, we investigated structural and compositional differences between the gallery and spiny forests, as well as within distinct gallery forest sites in Berenty Reserve in the south of the island. We also evaluated differences in habitat quality between the spiny and gallery forests for three species of diurnal lemurs in the reserve, and analyzed the current population trend of the tamarind trees, a species of ecological and cultural importance in Madagascar. Our findings revealed that the spiny and gallery forests differed in composition and structure, confirming the unique ecological characteristics of gallery forests and the underexplored richness of spiny forests. Spiny forests exhibited higher species richness despite a comparatively lower sampling effort, emphasizing the need for focused conservation efforts in these overlooked ecosystems. Tamarind populations, vital for lemur nutrition, showed signs of inadequate regeneration suggesting a recruitment bottleneck, possibly due to factors like a lowering water table, brown lemur foraging habits, or shifts in environmental conditions. Urgent interventions, including enrichment plantations, were recommended to ensure the survival of this keystone species. Contrasting botanical and lemur-centric perspectives revealed that while spiny and gallery forests differed botanically, they offered comparable habitat quality for ring-tailed and sifaka lemurs. However, brown lemurs exhibit a preference for the gallery forest, highlighting the intricate relationship between plant composition and lemur habitat choices. Our study underscores the urgency of expanding our knowledge of Madagascar´s dry forests, and Berenty Reserve, as one of the few remaining protected areas with gallery and spiny forests, serves as a reference for future research in Madagascar's understudied ecosystems.

Fig 1. Gallery and spiny forests at Berenty Reserve, Madagascar.

(a) Forest locations at Berenty reserve (Modified from Rambeloarivony & Jolly 2013b). (b) Aerial photograph of Berenty Reserve forests, showing forest plot locations. Spiny forest plots are indicated with yellow circles, and gallery forest plots are indicated with green circles (Image: Modified Copernicus Sentinel data [2024]). (c) Photo of the spiny forest, and (d) Photo of the gallery forest (Photos: Ariadna Mondragón-Botero).

Fig 2. Number and overlap in species in the different forest areas in Berenty Reserve.

The Venn diagrams show the overlap between a) the gallery and the spiny forest, and b) among the three sites of the gallery forest. Only 10 species were shared between the spiny and gallery forests, and 12 were common to the three gallery forest sites.

Fig 3. Species accumulation curves for the spiny and gallery forests in Berenty Reserve.

a) gallery and spiny forests where the yellow line (dots) corresponds to the spiny forest and the green line (squares) corresponds to the gallery forest; (b) Species accumulation curves for the three different within the gallery forest areas: Anaramalangy (triangles), Ankoba (squares), and Malaza (circles). The x-axis represents cumulative sampling effort, with each point corresponding to an additional sampling site (or plot) surveyed. Vertical lines on each curve show the 95% confidence interval.

Fig 4. Species Importance Value Index (IVI_s) for the gallery and spiny forests.

The IVI_s is shown in (a) for the gallery forest and in (b) for the spiny forest.

Fig 5. Species Importance Value Index (IVI_s) the three forest sites within the gallery forest.

(a) Anaramalangy, (b) Ankoba, and (c) Malaza.

Fig 6. Distribution of adult and seedling tamarind densities in the gallery forest at Berenty Reserve.

(a) Density of adult tamarind trees in each diameter size class across all the gallery forest plots (b) Number of subplots (frequency) per seedling density category. Each category represents the number of tamarind seedlings per hectare within that range (e.g., ‘1 to 100’ indicates subplots where the seedling density was between 1 and 100 individuals per hectare). In this study, we defined adult trees as tamarinds with a DBH ≥ 10 cm and seedlings as plants with diameter at root collar ≤ 1 cm.

Fig 7. Lemur habitat quality in Berenty Reserve.

The habitat quality is represented as the total basal area of all food tree species consumed by the three species of diurnal lemurs in (a) the gallery and spiny forests, and (b) the three different gallery forest sites. Comparisons are done within lemur species. Vertical bars represent standard errors and different letters indicate significant differences between forests.