Impacts of an invasive snail (Tarebia granifera) on nutrient cycling in tropical streams: the role of riparian deforestation in Trinidad, West Indies

Abstract

Non-native species and habitat degradation are two major catalysts of environmental change and often occur simultaneously. In freshwater systems, degradation of adjacent terrestrial vegetation may facilitate introduced species by altering resource availability. Here we examine how the presence of intact riparian cover influences the impact of an invasive herbivorous snail, Tarebia granifera, on nitrogen (N) cycling in aquatic systems on the island of Trinidad. We quantified snail biomass, growth, and N excretion in locations where riparian vegetation was present or removed to determine how snail demographics and excretion were related to the condition of the riparian zone. In three Neotropical streams, we measured snail biomass and N excretion in open and closed canopy habitats to generate estimates of mass- and area-specific N excretion rates. Snail biomass was 2 to 8 times greater and areal N excretion rates ranged from 3 to 9 times greater in open canopy habitats. Snails foraging in open canopy habitat also had access to more abundant food resources and exhibited greater growth and mass-specific N excretion rates. Estimates of ecosystem N demand indicated that snail N excretion in fully closed, partially closed, and open canopy habitats supplied 2%, 11%, and 16% of integrated ecosystem N demand, respectively. We conclude that human-mediated riparian canopy loss can generate hotspots of snail biomass, growth, and N excretion along tropical stream networks, altering the impacts of an invasive snail on the biogeochemical cycling of N.

Figure 1. Tarebia granifera.

The quilted melania snail has invaded freshwater habitats throughout much of the Neotropics. Photo credit: S. B. Snider.

Figure 2. Quantity and quality of snail food resources by canopy cover state.

Mean (±1 SE) chlorophyll a (A), ash-free dry mass (AFDM) (B), and molar C:N ratios of epilithon (C). RAM  =  Ramdeen Stream, ARI  =  Aripo River, YAR  =  Yarra River. Gray and white bars represent data collected in closed and open canopy sites, respectively (“closed” is ≥75% and “open” is ≤25% canopy cover). Asterisks above bars represent significant differences (p<0.05) among canopy types within streams.

Figure 3. Mean T. granifera biomass by size class.

Top row of panels are 2007 data and bottom row are 2008 data. Size classes were based on 5 mg AFDM increments. RAM  =  Ramdeen Stream, ARI  =  Aripo River, YAR  =  Yarra River. Gray and white bars represent data collected in closed and open canopy sites, respectively. Canopy type had a significant impact on T. granifera areal biomass (F = 22.06, p<0.0001). Note log scale. No data were collected in YAR in 2007.

Figure 4. Influence of canopy state on nitrogen excretion by T. granifera in three streams.

Panels from top represent Ramdeen Stream (RAM), Aripo River (ARI), and Yarra River (YAR). Open and closed circles represent individual snails collected in open and closed canopy habitats, respectively. Solid lines represent trends in closed canopy habitats, and broken lines represent trends in closed canopy habitats. All data were log-transformed.

Figure 5. Canopy state and snail growth rates.

Effect sizes of canopy state on snail growth rates as measured in a reciprocal transplant experiment. Error bars represent 95% confidence intervals and effect sizes were measured as response ratios (Hedges et al. 1997). “Open to closed canopy” represents snails collected from open canopy habitat and moved to closed canopy habitat where growth rates were measured after 10 days, and “closed to open canopy” is vice versa.

Figure 6. Influence of canopy state on areal snail excretion rates in three streams.

Mean (±1 SE) areal N excretion by T. granifera in 2008. RAM  =  Ramdeen Stream, ARI  =  Aripo River, YAR  =  Yarra River. Closed and open bars represent data collected in closed and open canopy sites, respectively. Canopy effect was significant across all streams (p<0.0001).