Fine‐tuning for the tropics: application of eDNA technology for invasive fish detection in tropical freshwater ecosystems

Abstract

Invasive species pose a major threat to aquatic ecosystems. Their impact can be particularly severe in tropical regions, like those in northern Australia, where >20 invasive fish species are recorded. In temperate regions, environmental DNA (eDNA) technology is gaining momentum as a tool to detect aquatic pests, but the technology's effectiveness has not been fully explored in tropical systems with their unique climatic challenges (i.e. high turbidity, temperatures and ultraviolet light). In this study, we modified conventional eDNA protocols for use in tropical environments using the invasive fish, Mozambique tilapia (Oreochromis mossambicus) as a detection model. We evaluated the effects of high water temperatures and fish density on the detection of tilapia eDNA, using filters with larger pores to facilitate filtration. Large-pore filters (20 μm) were effective in filtering turbid waters and retaining sufficient eDNA, whilst achieving filtration times of 2-3 min per 2-L sample. High water temperatures, often experienced in the tropics (23, 29, 35 °C), did not affect eDNA degradation rates, although high temperatures (35 °C) did significantly increase fish eDNA shedding rates. We established a minimum detection limit for tilapia (1 fish/0.4 megalitres/after 4 days) and found that low water flow (3.17 L/s) into ponds with high fish density (>16 fish/0.4 megalitres) did not affect eDNA detection. These results demonstrate that eDNA technology can be effectively used in tropical ecosystems to detect invasive fish species.

Comparison of capture and storage methods for aqueous macrobial eDNA using an optimized extraction protocol: advantage of enclosed filter

“…This can be overcome by pre‐filtering (Robson et al . ) and/or increasing the number of filter replicates. Future research is needed to identify optimal procedures for highly productive and/or turbid waters.…”

Methods to maximise recovery of environmental DNA from water samples

“…For example, we have experienced filtration times of up to 3 hours for 2L samples from highly turbid waterways (>200 nephelometric turbidity unit (NTU)), despite using multiple 1.2 μm glass fibre filter papers. Previous work by Robson, Noble [35] suggested increasing filter pore size (10 or 20 μm) or using pre-filters to decrease filtration time particularly when using water with high sediment load or algae. This could mean though that higher water volumes must be filtered in order to get the same quantity of eDNA if one is to use smaller pore sized filters.…”

“…Many eDNA studies use pore sizes ranging from 0.45μm– 3 μm [9, 10, 32–34]. For highly turbid water however, such as in the tropics, even 3 μm filter papers are easily blocked, necessitating the use of larger pore size filters or pre-filtration to significantly minimize filtration time [35]. Glass fibre (GF), cellulose nitrate (CN), polycarbonate (PC), nylon, polyethersulfone (PES) and cellulose acetate (CA) filters have been used in eDNA studies [36].…”

Development and application of environmental DNA surveillance for the threatened crucian carp (Carassius carassius)

“…At the sample level, more individuals (reflected by CPUE) should increase eDNA concentration and improve detection. Temperature can increase physical, metabolic, or behavioural activity of organisms resulting in more eDNA release, breakdown, and degradation (Buxton, Groombridge, Zakaria, & Griffiths, ; Bylemans et al., ; Lacoursière‐Roussel, Rosabal, & Bernatchez, ; Pilliod et al., ; Robson et al., ; Strickler et al., ; Takahara et al., ). Links established between eDNA and pH support greater detectability, concentration, and persistence of eDNA in more alkaline waters (Barnes et al., ; Goldberg et al., ; Strickler et al., ).…”

“…Environmental variables play a substantial role in abundance/biomass estimation by influencing the ecology of eDNA (Barnes et al., ). Variables examined have included temperature, pH, salinity, conductivity, anoxia, sediment type, and UV light (Barnes et al., ; Buxton, Groombridge, & Griffiths, ; Buxton, Groombridge, Zakaria, & Griffiths, ; Goldberg, Strickler, & Fremier, ; Keskin, ; Pilliod, Goldberg, Arkle, & Waits, ; Robson et al., ; Stoeckle et al., ; Strickler, Fremier, & Goldberg, ; Takahara et al., ; Weltz et al., ). However, these variables are not always measured and only a handful of studies have assessed their effects in ponds (Buxton, Groombridge, & Griffiths, ; Buxton, Groombridge, Zakaria, & Griffiths, ; Goldberg et al., ; Takahara et al., ).…”