Diversity and seasonality of ectoparasite burden on two species of Madagascar fruit bat, Eidolon dupreanum and Rousettus madagascariensis

Abstract

Background: Bats are important reservoir hosts for a variety of pathogens, some of which are transmitted by ectoparasite vectors including mites, fleas, lice, ticks, and bat flies (families Nycteribiidae and Streblidae). All these ectoparasite taxa are known to parasitize two endemic fruit bats of Madagascar, Eidolon dupreanum and Rousettus madagascariensis. We aimed to describe the diversity of ectoparasite infestation for both bat species through morphological observation and DNA barcoding and elucidate ecological and climatic correlates of seasonal nycteribiid parasitism of these hosts.

Methods: Eidolon dupreanum and R. madagascariensis fruit bats were live-captured in northern and central-eastern Madagascar periodically from 2013 to 2020. Ectoparasites on all captured bats were counted and identified in the field and then collected into ethanol. Field identification of a subset of samples was confirmed via microscopy and DNA barcoding of the cytochrome C oxidase subunit 1 (COI) and 18S genes. The seasonal abundance of nycteribiid bat flies on both host bats was analyzed using generalized additive models, and the role of climate in driving this seasonality was assessed via cross-correlation analysis combined with generalized linear models. Phylogenetic trees were generated to compare COI and 18S sequences of Madagascar nycteribiid and streblid bat flies with available reference sequences from GenBank.

Results: Ectoparasites corresponding to four broad taxa (mites, ticks, fleas, and bat flies) were recovered from 628 of 873 E. dupreanum (71.9%) and 831 of 862 R. madagascariensis (96.4%). Eidolon dupreanum were most commonly parasitized by Cyclopodia dubia nycteribiids and R. madagascariensis by Eucampsipoda madagascariensis nycteribiids and Megastrebla wenzeli streblids. We observed significant seasonality in nycteribiid abundance on both bat hosts, which varied by bat sex and was positively correlated with lagged temperature, precipitation, and humidity variables. Barcoding sequences recovered for all three bat fly species grouped with previously reported sequences, confirming morphological species identification. Our study contributes the first DNA barcodes of any kind reported for M. wenzeli and the first 18S barcodes for C. dubia.

Conclusions: This study explores the diversity and abundance of ectoparasite burdens in two Malagasy fruit bat species, highlighting the importance of seasonal ecology and the influence of climate variables on parasitism, which correlates with resource availability.

Keywords: Eidolon dupreanum; Rousettus madagascariensis; Bat ectoparasite; Bat fly; DNA barcoding; Madagascar; Nycteribiidae; Pteropodidae; Streblidae.

Fig. 1

Alluvial plot showing bat host species (center) associations with broad ectoparasite clades (top) and genus-species classifications (bottom). Fleas and bat flies in order Diptera are colored blue, while mites and ticks in class Arachnida (respectively, superorder Acariformes and Parasitiformes) are colored green. Images taken under the microscope at 40 × magnification are shown below the names of corresponding species

Fig. 2

Seasonal variation in the abundance of Nycteribiidae bat flies counted on A, B Eidolon dupreanum and C, D Rousettus madagascariensis bats captured at roost sites in central-eastern Madagascar (respectively, Angavobe/Angavokely and Maromizaha caves). A and C Seasonal ectoparasite count predictions (red line) from best-fit GAMs for male and female bats, with 95% CI by standard error shaded in gray. Background points in black show raw data from 2013–2020. Background shading in pink indicates gestation period for each species from [63], and shading in blue indicates the nutritionally deficient dry season for the region. B and D Partial effect (y-axis) of bat host mass: forearm residual (x-axis) on bat fly count, respectively, for E. dupreanum and R. madagascariensis. Solid lines (gray = non-significant; blue = significant effects) show mean effects, with 95% CIs by standard error in translucent shading

Fig. 3

Mean monthly nycteribiid count per bat for A Cyclopodia dubia parasitizing Eidolon dupreanum, from 2013–2020 for Angavokely roost (gray lines and points; left y-axis), compared with monthly averages for different climate variables in the region (horizontal panels; red lines and points; right y-axis): humidity (% relative humidity), total precipitation (mm), and temperature (°C). B Eucampsipoda madagascariensis parasitizing Rousettus madagascariensis, mirroring structure from A, for Maromizaha roost; 95% CIs by standard error are shown for both ectoparasite and climate data

Fig. 4

Association of climate and demography with abundance of nyteribiid bat flies for A, B Cyclopodia dubia on Eidolon dupreanum and C, D Eucampsipoda madagascariensis on Rousettus madagascariensis. A, C Top five GLMs using optimally lagged climate variables to predict bat fly abundance, ranked by δAICc. Rows represent individual models and columns represent predictor variables. B, D Incidence rate ratios of linear predictors from top-fit models indicated in A, C. Significant positive correlates are colored red, significant negative correlates blue, and insignificant correlates gray; 95% CIs by standard error are shown as horizontal error bars

Fig. 5

Maximum likelihood phylogeny of COI ectoparasite sequences from untrimmed alignment (RAxML-NG, GTR + I + G4) [83]. Bootstrap support values computed using Felsenstein’s method [84] are shown as shaded circles on each node, corresponding to legend. Sequences are collapsed into single species or genus clades for visualization; see Additional file 1: Fig. S7 for full phylogeny with individual sequences labeled. Tip shapes are colored by genera. Tip labels for the three Madagascar clades (Eucampsipoda madagascariensis, Cyclopodia dubia, Megastrebla wenzeli) are highlighted in yellow. Tree is rooted in Drosophila melanogaster, accession number NC_001709. Branch lengths are scaled by nucleotide substitutions per site, corresponding to scalebar