Effects of habitat edges on vegetation structure and the vulnerable golden-brown mouse lemur (Microcebus ravelobensis) in northwestern Madagascar

Abstract

Background: Edge effects can influence species composition and community structure as a result of changes in microenvironment and edaphic variables. We investigated effects of habitat edges on vegetation structure, abundance and body mass of one vulnerable Microcebus species in northwestern Madagascar. We trapped mouse lemurs along four 1000-m transects (total of 2424 trap nights) that ran perpendicular to the forest edge. We installed 16 pairs of 20 m² vegetation plots along each transect and measured nine vegetation parameters. To determine the responses of the vegetation and animals to an increasing distance to the edge, we tested the fit of four alternative mathematical functions (linear, power, logistic and unimodal) to the data and derived the depth of edge influence (DEI) for all parameters.

Results: Logistic and unimodal functions best explained edge responses of vegetation parameters, and the logistic function performed best for abundance and body mass of M. ravelobensis. The DEI varied between 50 m (no. of seedlings, no. of liana, dbh of large trees [dbh ≥ 10 cm]) and 460 m (tree height of large trees) for the vegetation parameters, whereas it was 340 m for M. ravelobensis abundance and 390 m for body mass, corresponding best to the DEI of small tree [dbh < 10 cm] density (360 m). Small trees were significantly taller and the density of seedlings was higher in the interior than in the edge habitat. However, there was no significant difference in M. ravelobensis abundance and body mass between interior and edge habitats, suggesting that M. ravelobensis did not show a strong edge response in the study region. Finally, regression analyses revealed three negative (species abundance and three vegetation parameters) and two positive relationships (body mass and two vegetation parameters), suggesting an impact of vegetation structure on M. ravelobensis which may be partly independent of edge effects.

Conclusions: A comparison of our results with previous findings reveals that edge effects are variable in space in a small nocturnal primate from Madagascar. Such an ecological plasticity could be extremely relevant for mitigating species responses to habitat loss and anthropogenic disturbances.

Keywords: Abundance; DEI; Depth of edge influence; Edge effects; Habitat choice; Habitat loss; Madagascar; Microcebus ravelobensis; Mouse lemur; Vegetation structure.

Fig. 1

Map of Mariarano Classified Forest (MCF) showing the study area and the location of all four study transects (C1, C2, C3 and C4) in relation to settlements and forest edge. The inset at the top right shows the location of the study site (star) within Madagascar (map drawn by: B. Andriatsitohaina)

Fig. 2

Illustration of the best function (logistic and unimodal) and averaged values of all nine vegetation parameters across the length of the transects (a–i) and their respective depth of edge influence DEI. Mean values (dots) are shown together with minimum and maximum values (whiskers). Lines represent the best fit of the distribution of all nine vegetation parameters across distance from edges

Fig. 3

Comparison between edge and interior habitats for the two significant vegetation parameters: height of small trees (a) and number of seedlings (b). Black squares: mean, violin plots range from minimum to maximum value with violin width illustrating data distribution

Fig. 4

Variation of abundance (a) and body mass (b) of M. ravelobensis from the edge into the interior of forest and calculation of their DEI, respectively. Mean values (dots) are plotted together with minimum and maximum values (whiskers). In addition, the first derivatives of the fitted curves (c, d: abundance and body mass) and the second derivatives (e and f: abundance and body mass) are plotted. The DEI (vertical arrows in (e) for the abundance and in (f) for the body mass) is determined as the distance between the two first inflection points in the curve of the second derivative

Fig. 5

Comparison of a abundance (males only), and b body mass (females only) of M. ravelobensis between edge and interior habitats in form of violin plots. Black squares show the mean, violin plots end with minimum and maximum values

Fig. 6

Relationship between three vegetation parameters (a–c: density of large, small trees and saplings, respectively) and the abundance of M. ravelobensis. Linear regression lines are shown for the three parameters

Fig. 7

Relationship between three vegetation parameters (a, b: dbh of small trees and density of seedlings, respectively) and body mass of M. ravelobensis. Linear regression lines are shown for the two parameters