Coinfection rates of avian blood parasites increase with latitude in parapatric host species

Abstract

Animals are frequently coinfected with multiple parasites concurrently, and advances in our sampling of these complex intra-host parasite communities have revealed important ecological impacts on their hosts. However, the spatial distributions and environmental determinants of parasite coinfection remain infrequently studied. Here, we investigated the drivers of haemosporidian blood parasite coinfection in the Bicknell's thrush (Catharus bicknelli) and grey-cheeked thrush (Catharus minimus), parapatric sister species that occur across a broad latitudinal range in northern North America. Using 298 samples from across the distributions of these species, we found high overall infection (86%) and coinfection (41%) rates within host populations. Coinfection rates within populations were highly variable across sampling sites, ranging from 7 to 75%. Latitude was a more important predictor of coinfection frequency than host species identity, with coinfections becoming more abundant at higher latitudes. The 2 host species exhibited similar parasite faunas, and an analysis of the co-occurrence patterns among haemosporidians showed that host species identity was largely not a factor in structuring which parasites were found within coinfections. To our knowledge, this is the first study to illustrate a reverse latitudinal gradient in coinfection frequency in a eukaryotic parasite system. Further work is necessary to determine whether vector ecology or some other factor is the primary proximate driver of this pattern.

Keywords: Catharus; Haemosporida; Leucocytozoon; coinfection; latitudinal diversity gradient.

Graphical abstract

Fig. 1.

(A) Map of breeding ranges of host species: grey-cheeked thrush (blue) and Bicknell's thrush (purple). Shape files were provided by Birdlife International and Handbook of Birds of the World (2021). (B) Maps showing the localities of sampling of both host species and the number of infections identified in individual birds (ranging from 0 to 4 infections).

Fig. 2.

Variability in prevalence of haemosporidians with latitude of (A) Leucocytozoon, (B) Plasmodium and (C) Haemoproteus. Linear regressions are fitted for each host species (same colours as Fig. 1) and for both host species combined (black-dotted line). Points indicate the infection status of individual hosts of each species.

Fig. 3.

Variability in Leucocytozoon coinfection probability with latitude. Linear regressions are fitted for each host species (same colours as Fig. 1) and for both host species combined (black-dotted line). Points indicate the infection status of individual hosts of each species.

Fig. 4.

Coefficient outputs of the CRF model produced by MRFcov. Each cell of the matrix contains the number of co-occurrences of 2 haemosporidian lineages that we found across all host species, with the colour of the cell shown based on the estimated association coefficient with darker colours indicating a more negative association coefficient.