Spatiotemporal variability in transmission risk of human schistosomes and animal trematodes in a seasonally desiccating East African landscape

Abstract

Different populations of hosts and parasites experience distinct seasonality in environmental factors, depending on local-scale biotic and abiotic factors. This can lead to highly heterogenous disease outcomes across host ranges. Variable seasonality characterizes urogenital schistosomiasis, a neglected tropical disease caused by parasitic trematodes (Schistosoma haematobium). Their intermediate hosts are aquatic Bulinus snails that are highly adapted to extreme rainfall seasonality, undergoing dormancy for up to seven months yearly. While Bulinus snails have a remarkable capacity for rebounding following dormancy, parasite survival within snails is greatly diminished. We conducted a year-round investigation of seasonal snail-schistosome dynamics in 109 ponds of variable ephemerality in Tanzania. First, we found that ponds have two synchronized peaks of schistosome infection prevalence and cercariae release, though of lower magnitude in the fully desiccating ponds than non-desiccating ponds. Second, we evaluated total yearly prevalence across a gradient of an ephemerality, finding ponds with intermediate ephemerality to have the highest infection rates. We also investigated dynamics of non-schistosome trematodes, which lacked synonymity with schistosome patterns. We found peak schistosome transmission risk at intermediate pond ephemerality, thus the impacts of anticipated increases in landscape desiccation could result in increases or decreases in transmission risk with global change.

Keywords: Bulinus; Schistosoma haematobium; aestivation; ephemerality; schistosomiasis; seasonality.

Figure 1:

Map of localities of 109 ponds surveyed (red squares) across six districts of Northwestern Tanzania, with inset of locality of sites within the continent of Africa.

Figure 2:

Change in area of A) a desiccating pond in Misungwi district, Kisima cha Longo (reduction of pond area of 100%) and B) a non-desiccating pond in Busega district, Lambo la Wachina (reduction of pond area of 79.78%) within a survey year. Inset shows the shallow substrate depth at which an aestivating snail was found in a pond in Busega district, Lambo la Wachina (the pictured snail had died).

Figure 3:

Generalized Additive Models (GAMMs) representing circannual (A) rainfall patterns (mm per day) in Mwanza, Tanzania of data from the airport weather station, and variability in B) pond depth (m) and C) Bulinus snail abundance (number collected in 30-minute survey) in our 109 sites. The shaded region surrounding the line represents the best fit ± 1 standard error. Points represent monthly means and standard error. Depth and snail abundance were significantly non-linear and different between the two ephemerality categories.

Figure 4:

GAMMs representing circannual A) schistosome prevalence, B) schistosome intensity, C) non-schistosome prevalence, and F) non-schistosome intensity in desiccating and non-desiccating ponds. The shaded region surrounding the line represents the best fit ± 1 standard error. Points represent monthly means and standard error. All patterns were significantly non-linear and different between the two ephemerality categories.

Figure 5:

Overall yearly A) schistosome and B) non-schistosome trematode infection prevalence in across an ephemerality gradient (% decrease in pond area in the dry season). Schistosome prevalence peaks at intermediate ephemerality (~80% area decrease, Binomial GLM, p<0.01). Non-schistosome prevalence peaks at low ephemerality (~50% area decrease, Binomial GLM, p<0.01).