Introduction
Recent evidence suggests that increasing agriculture to meet the growing global food demand could simultaneously increase human infectious diseases (Rohr et al., 2019), necessitating studies that identify means of sustainably improving food production without increasing disease. One such human disease that is positively associated with agricultural expansion is schistosomiasis, a disease caused by parasites released from freshwater snails, with an estimated 779 million people at risk globally (Vos et al., 2016). Over 90% of global cases (~200 million cases) occur in sub-Saharan Africa and involve Biomphalaria and Bulinus snail species as intermediate hosts (Gryseels, Polman, Clerinx, & Kestens, 2006; Steinmann, Keiser, Bos, Tanner, & Utzinger, 2006). Schistosomiasis has been positively linked to expanding snail habitat and loss of native snail predators (Huang & Manderson, 1992; Savaya Alkalay et al., 2014; Steinmann et al., 2006) because each infected snail host can release thousands of cercariae, the human-infectious life stage, daily. Parasite production by snails can be influenced by local aquatic factors (Haggerty, Bakhoum, Civitello, et al., 2020), and thus, human infection rates respond rapidly to changes in local ecological factors (Huang & Manderson, 1992), emphasizing the importance of freshwater ecology in infection dynamics.
Agrochemical pollution has the potential to influence ecological factors that directly or indirectly influence schistosomiasis transmission. Both fertilizers and herbicides can increase periphyton resources required for snail growth and cercarial production (Halstead et al., 2014), and have been linked to increased trematode infection intensities in frogs in agricultural wetlands (Rohr et al., 2008). Herbicides kill phytoplankton that absorb light from the water column, which can indirectly increase growth of periphyton, attached algae that is a major resource for snails (Halstead et al., 2018). Additionally, fertilizers and herbicides can increase and decrease, respectively, submerged vegetation that acts as primary habitat for snails and may influence snail predation. Invertebrate predators that consume snails and help regulate snail populations can be killed by some insecticides even at environmentally relevant concentrations, leading to top-down insecticide effects that increase snails (Halstead et al., 2014). Snail intermediate hosts are difficult to control (Gryseels et al., 2006; Steinmann et al., 2006), and understanding the strength of various agrochemicals on snails might be key to understanding recent increases in human schistosomiasis coincident with agricultural expansion and preventing any future surges (Hoover et al., 2020).
Agrochemical use is predicted to increase 2-5 fold over the next few decades (Foley et al., 2011; Tilman, Balzer, Hill, & Befort, 2011; Tilman et al., 2001), yet laboratory studies of pesticides using species in isolation could miss the net effects of pesticides in natural communities (Clements & Rohr, 2009; Rohr & Crumrine, 2005; Rohr, Kerby, & Sih, 2006). Submerged aquatic vegetation that acts as snail habitat might decrease the toxicity of certain insecticides by changing pH (Brogan & Relyea, 2014), or influence predation by providing refugia (Davis, Purrenhage, & Boone, 2012). Thus, because agrochemicals in natural environments are differently subject to the above influences, simulating natural communities using mesocosm experiments is needed to improve predictions of the direct and indirect effects of agrochemicals on freshwater communities, such as those containing snail hosts (Rohr, Salice, & Nisbet, 2016).
We created experimental natural freshwater pond communities to determine the predictability of top-down effects of insecticides and bottom-up effects of fertilizer or herbicides on the biomass of two genera of snail hosts that can transmit schistosomiasis (Biomphalaria glabrata and Bulinus truncatus ). More specifically, we examined the top-down effects of six insecticides belonging to two insecticide classes (the organophosphates chlorpyrifos, malathion, and terbufos, and the pyrethroids esfenvalerate, λ-cyhalothrin, and permethrin), and the bottom-up effects of both fertilizer and six herbicides belonging to two herbicide classes (the chloroacetanilides acetochlor, alachlor, and metachlor, and the triazines atrazine, propazine, and simazine) in the presence of snail predators. We then performed a follow-up experiment using all six herbicides but without snail predators to determine if herbicide bottom-up effects on snails are only evident in the absence of strong top-down effects. In the follow-up experiment, we also examined if total parasite production by snails was predicted by their total biomass, and, thus, whether snail biomass changes associated with different agrochemical exposures could be related to human exposure to parasites.