Temperature does not influence functional response of amphipods consuming different trematode prey.

Direct consumption on free-living cercariae stages of trematodes by non-host organisms interferes with trematode transmission and leads to reduced infections in the next suitable hosts. Consumer functional responses provide a useful tool to examine relationships between consumption rates and ecologically relevant prey densities, whilst also accounting for abiotic factors that likely influence consumption rates. We investigated how temperature influences the consumer functional response of the amphipod Gammarus lacustris towards the cercariae of three freshwater trematodes (Diplostomum, Apatemon and Trichobilharzia). Amphipods displayed different functional responses towards the parasites, with Type II responses for Diplostomum and Type I responses for Apatemon prey. Temperature did not alter the consumption rate of the amphipod predator. Trichobilharzia was likely consumed at similar proportions as Diplostomum; however, this could not be fully evaluated due to low replication. Whilst Type II responses of invertebrate predators are common to various invertebrate prey types, this is the first time a non-filter feeding predator has been shown to exhibit Type I response towards cercarial prey. The prey-specific consumption patterns of amphipods were related to cercarial distribution in the water column rather than to the size of cercariae or temperature influence. The substantial energy flow into food webs by non-host consumer organisms highlights the importance of understanding the mechanisms that modulate functional responses and direct predation in the context of parasitic organisms.

Introduction

Trematodes are highly abundant parasitic organisms in aquatic ecosystems, with the biomass of their free-living life stages often equivalent to the most abundant insect orders (Preston et al. 2013). The immense mass of trematodes is largely attributed to the asexual production and daily release of thousands of free-living infective cercariae (larvae) from their first intermediate molluscan hosts (Preston et al. 2013; Soldánová et al. 2016). Whilst successful transmission of cercariae is essential to complete the trematode’s life cycle, direct consumption of cercariae by non-host organisms can lead to reduced transmission to their next suitable host (Johnson et al. 2010). The loss of cercariae by consumption from different organisms has been reported (e.g. rotifers, molluscs, insect larvae, fishes; Orlofske et al. 2015; Welsh et al. 2017), but its importance appears to be largely overlooked despite the significant contribution parasites may have to energy flow through food webs when being consumed by non-host organisms (Thieltges et al. 2008; Johnson et al. 2010). Cercarial consumption may also be context dependent, influenced by environmental conditions such as temperature or habitat complexity, in addition to the impact of the size and behaviour of the parasite and the predator (e.g. Orlofske et al. 2015; Selbach et al. 2019).

Temperature, in particular, is recognised as a major driver of both predator-prey and parasite transmission dynamics as it simultaneously influences both consumption rates (e.g. Goedknegt et al. 2015; Laverty et al. 2017) and the production of free-living parasite stages (Poulin 2006). In order to understand how temperature may affect parasite transmission through its influence on parasite consumption by non-hosts, it is essential to examine potential predator-prey interactions. Consumer functional responses (hereafter FR) provide a useful tool to examine relationships between consumption rates and ecologically relevant prey densities (e.g. Holling 1959; Jeschke et al. 2004), and thus increasing our understanding of the impact predators have on prey populations (e.g. Bovy et al. 2015; Paterson et al. 2015). Here, we address the potential consumption of trematode cercariae by the omnivorous amphipod Gammarus lacustris G.O. Sars, 1863 (Malacostraca, Gammaridae), one of the most common macroinvertebrates in European subarctic lakes. This amphipod co-occurs in the nearshore benthic habitat with the lymnaeid snail Radix balthica L. (Gastropoda, Pulmonata), which release dense aggregations of larval trematodes, including Diplostomum Poirier, 1886 (Diplostomidae), Apatemon Szidat, 1928 (Strigeidae) and Trichobilharzia Skrjabin and Zakharow, 1920 (Schistosomatidae) (Soldánová et al. 2017). We utilised a FR approach to investigate whether the consumption of these three trematode prey taxa by G. lacustris was altered by temperature variation.

Materials and methods

Study organisms and maintenance

All study organisms were obtained from Takvatn, a subarctic lake in northern Norway (69° 07′ N, 19° 05′ E), during August and October 2017 and 2018. Gammarus lacustris amphipods (n = 340 males; length [pereon + pleosome] mean ± SD: 11.9 ± 1.7 mm) were collected from the lake substrate (4–10-m depth) by dredge net with a rectangular metal frame (20 × 35-cm opening, 1-mm mesh), whereas R. balthica snails were hand-picked from the littoral zone (< 1-m depth). Amphipod sex was confirmed under the microscope. Study organisms were acclimatised for 3–10 days in a climatic chamber, with light:dark (20:4, 10:14 h) and temperature (13, 6 °C) conditions reflecting measures taken in August and October, respectively (data loggers Onset HOBO UA-002-64 Pendant 64K, MA, USA). Amphipods were maintained in 6-L aquaria with aerated lake water (replaced every 72 h) and provided ad libitum with natural prey (e.g., copepods, cladocerans), whereas snails were individually maintained in 40-mL beakers with lake water (replaced daily) and provided with lettuce (Lactuca sativa L.).

Snails were screened for trematode infections by placing each beaker under a light source for 24 h to stimulate cercarial emergence, with taxa identified in vivo (Soldánová et al. 2017). Three trematode genera that do not infect amphipods were selected for the experimental assays, i.e. Trichobilharzia, Apatemon and Diplostomum. Trichobilharzia franki haplotype “peregra” (hereafter Trichobilharzia) has a two-host life cycle with birds as definitive hosts, whereas Apatemon and Diplostomum have a three-host life cycle, with fish and piscivorous birds as second intermediate and definitive hosts, respectively (Soldánová et al. 2017). The three trematodes display intermittent swimming periods, with Diplostomum and Trichobilharzia tending to accumulate close to the water surface in relation to the light intensity and Apatemon continuously swimming with only short pauses. However, Apatemon exhibits nocturnal emergence, whereas Diplostomum and Trichobilharzia exhibit diurnal emergence (Bell 1996; Karvonen et al. 2004; Haas et al. 2008; Soldánová et al. 2016). Cercarial size also differed between these trematodes (Apatemon ~ 530 μm, Diplostomum ~ 690 μm, Trichobilharzia ~ 940 μm; total length, n = 33, 31 and 23 live cercariae specimens, respectively).

To obtain cercarial prey for the experiment, snails were individually incubated in 40 mL of filtered lake water under a light source (Trichobilharzia, Diplostomum) or covered with aluminium foil to mimic night-time (Apatemon). Host-induced variability in cercariae (e.g. maturity or size) was accounted for by preparing a pool of freshly emitted cercariae from multiple infected snails for each trematode genus. Thereafter, cercariae were individually pipetted into a plastic beaker with a small volume of filtered lake water (10 mL) in densities required for the FR experiment. All cercariae were the same age (< 4 h), alive and active when offered to the predator.

Experimental design

A randomised, fully factorial design was used to test the effect of temperatures relevant to the studied system (13, 6 °C, average water temperatures in August and October, respectively) on cercarial prey consumption. Each amphipod was supplied with one of eight prey densities (2, 4, 6, 10, 20, 50, 70 or 100 cercariae) from one trematode genus. Each treatment combination was replicated six times with the exception of Trichobilharzia, which was replicated twice due to the low availability of Trichobilharzia-infected snails.

Prior to each assay, amphipods were held without food for 48 h to standardise satiation levels preceding exposure to cercarial prey. Each amphipod was randomly assigned to an individual experimental arena (diameter 55 mm) containing 60 mL of filtered lake water, which was placed on a black surface to mimic the amphipod’s natural habitat. Amphipods were provided with a given density of cercariae prey from the prepared beakers, which were rinsed twice with filtered lake water. Consumed cercariae were not replaced during the assay (non-replacement design), with amphipods removed after 30 min. Unconsumed cercariae were counted by filtering the water from each arena through a 12-μm-pore filter membrane (Whatman, Nuclepore Track-Etch Membrane, Sigma-Aldrich), which was then fixed and stained in a 70% ethanol and carmine solution prior to counting cercariae under a stereomicroscope. Control replicates (n = 6 Apatemon and Diplostomum, n = 2 Trichobilharzia) in the absence of amphipods were run for each prey density to assess the potential loss of cercariae due to the filtration procedure (0–13 Apatemon, 0–21 Diplostomum, 0–22 Trichobilharzia cercariae lost, Supplementary Table S1), with the number of prey consumed by each amphipod adjusted by the mean loss of prey specific to each trematode genus and cercariae density. Prey survival (in terms of vital/mobile cercariae) in the absence of predators was 100%.

All amphipods were maintained for 48 h post assay in individual arenas under the same conditions. Amphipods which died or moulted prior to, during or post assay were replaced with another randomly selected amphipod, as moulting may influence the feeding behaviour of Gammarus (Bovy et al. 2015). Amphipods were then euthanised by immersion in carbonated water.

Statistical analysis

All statistical analyses were completed in R (R Core Team 2017, version 3.6.1). The presence of outliers and/or differences in prey consumption rates resulting from the experimental timeframe was determined by visual inspection of raw data plots, with one outlier detected in the Diplostomum dataset at the 10-prey density (Supplementary Fig. S1a).

To determine which FR model type (I, II or III) best described the relationship between cercariae density and the number of cercariae consumed, logistic regressions were fitted to each amphipod-trematode combination using frair::frair_test (version 0.5.100; Pritchard et al. 2017). Type I FRs assume a linear relationship between cercariae density and the number of cercariae consumed. Type II FRs (decreasing proportional consumption with increasing prey density) are indicated by a significant negative first-order term, whereas Type III FRs are indicated by significant positive first-order term, then a significant negative second-order term (initial increase then decrease in proportional consumption; Juliano 2001; Pritchard et al. 2017). Whilst parameters a (capture/attack rate), h (handling time, i.e. time spent subjugating, ingesting and digesting each prey item) and T (experimental time) describe a Type II response, Type I response show negligible h, being thus characterised by a and T only. The fixed parameter T was set to 0.5/24 (30 min), and after starting estimates had been provided, FR models were fitted using maximum likelihood estimation (MLE) by using frair::frair_fit (Pritchard et al. 2017). Due to the low number of replicates in Trichobilharzia prey, FR analyses could not be run for this prey and results are presented as a plot in Supplementary Fig. S1c. A single model was fitted for each trematode-amphipod dataset, including data at 6 and 13 °C, to test for differences in cercariae consumption between temperatures. Amphipods displayed different FR types towards each prey genus, and therefore, model parameters between these types could not be compared. A Type II FR was detected with Diplostomum, whereas in the analyses involving Apatemon, frair_test did not show any evidence of Type II or III FRs. Raw data comparing consumed prey or proportion of consumed prey across densities were plotted to check the fitting of the Apatemon data to a Type I FR. To visualise the uncertainty around the fitted FRs, bootstrapping (n = 1500) was used to construct 95% confidence intervals.

Results and discussion

Amphipods showed different consumption rates when preying on Diplostomum and Apatemon, with higher proportion of cercariae consumed at low prey densities (< 10 prey densities) and a slight decrease in consumption at 70- and 100-Diplostomum prey density. However, the proportion of consumed cercariae is similar between temperatures within the same prey genera. Raw data suggests that amphipods consumed Trichobilharzia cercariae at similar proportions to Diplostomum; however, this could not be fully evaluated for Trichobilharzia due to low replication. Raw data and percentage of consumed cercariae at each amphipod-prey combination are summarised in Supplementary Table S2.

In the current study, amphipods, namely G. lacustris, for the first time have been shown to efficiently consume trematode cercariae, potentially constituting an additional foraging resource for this abundant and widespread freshwater crustacean. The amphipods displayed Type II FRs towards Diplostomum at both temperatures (logistic regressions; Fig. 1a; Supplementary Table S3), which are commonly observed in gammarids and other invertebrates consuming various invertebrate prey types (e.g. Jeschke et al. 2004; Bovy et al. 2015; Wasserman et al. 2016; Iltis et al. 2018). In contrast, the consumption of Apatemon prey at both temperatures could best be described by a Type I FR, as the proportion of prey consumed was not modified by increasing prey density (Fig. 1b; Supplementary Fig. S2). Type I FRs assume a linear relationship between prey density and consumption and are usually restricted to filter feeding species (Jeschke et al. 2004). Predators showing Type II and Type III FRs typically exhibit long handling time h of their prey, whilst those displaying a Type I FR show negligible handling time h and fulfil the satiation condition. This means unless their gut is completely filled, consumers should search for food at a maximal rate with maximal effort, thus likely spending longer time foraging (Jeschke et al. 2004). Even if non-filter feeders would neither fulfil the handling nor the satiation condition described for Type I FRs, our data show for the first time that a non-filter feeding predator may also exhibit a Type I FR towards specific cercarial prey. Future experiments should include higher Apatemon prey densities to determine when the saturation point is reached.

Fig. 1

The effects of temperature (6 and 13 °C) on the consumer functional response of Gammarus lacustris amphipods towards a Diplostomum (Type II) and b Apatemon (Type I) cercariae prey. Lines indicate mean functional response, and coloured bars are 95% confidence intervals (blue, 6 °C; pink, 13 °C). Attack rate a and handling time h are shown by cercarial prey and temperature (T)

Differences in FR type depending on the cercarial prey genera consumed by G. lacustris suggest that prey-related traits could be affecting their predation risk (e.g. size or behaviour, Orlofske et al. 2015; Selbach et al. 2019). Jeschke et al. (2004) showed that in addition to the low handling time and satiation conditions typically associated with Type I FRs, invertebrate consumers must also be relatively large compared to their food. The size difference between Diplostomum and Apatemon is roughly 160 μm and should not have a major effect on the handling time displayed by amphipods. However, even if the relative predator/prey size of G. lacustris - Apatemon would set the limit to justify a Type I response according to the size condition shown by Jeschke et al. (2004), other factors linked to Apatemon’s cercarial behaviour or distribution in the water column might enable G. lacustris to continuously feed on them. Gammarus lacustris remains on or close to the bottom substrate during the day, swimming throughout the water column at night, mostly feeding on zooplankton (Wilhelm and Schindler 1999). Whilst both Apatemon and Diplostomum perform similar swimming behaviour, Diplostomum tends to occupy a position in the upper water column (Haas et al. 2008), and Apatemon likely swims faster (Santos et al. 2007; Selbach and Poulin 2018) and rests for shorter periods (Bell 1996) closer to intermediate positions of the water column (pers. obs.). The continuous movement and shorter distance to Apatemon prey may likely facilitate a constant consumption by benthic dwelling G. lacustris visual predators.

The overlapping 95% confidence intervals indicated that temperature did not alter FRs of amphipods towards any of the trematode prey (Fig. 1), even though FR attack rate a differed significantly between temperatures for amphipods preying on Diplostomum (Supplementary Table S4). Our results thus indicate a low impact of temperature on the consumption of cercarial prey, similar to previously reported for other invertebrate predator-prey systems (e.g. Paterson et al. 2015; Wasserman et al. 2016), but contrasting what was observed in gammarids by Laverty et al. (2017). Laverty et al. (2017) showed an increased prey consumption at increased temperatures, but their higher temperature range (11 and 20 °C vs current study 6 and 13 °C) may have triggered this change in consumption rate. Furthermore, the larger prey sizes used in their experiment (0.4–0.8 cm) may have altered the temperature impact on the predator’s FR.

Cercarial emergence patterns, as well as activity and survival rates, can be species-specific and/or temperature-dependent (Poulin 2006). Therefore, the absence of the ambient temperature influences on cercariae consumption may demonstrate the ability for amphipods to exploit the different emergence and active swimming periods intrinsic to Diplostomum and Apatemon throughout the year (Bell 1996; Karvonen et al. 2004). Nevertheless, the most profound impact could be expected on the Diplostomum population, as Type II FRs are considered to have larger implications for prey populations, potentially resulting in local prey extinction because high proportions of prey are consumed at low densities (Hassell 1978).

For the first time, a Type I FR has been revealed for a non-filter feeder species. The likely constant consumption of Apatemon by amphipods represents a continuous energetic source for this non-host organism, even when different prey densities are available. Therefore, the increased consumption rate and thus energy profit per unit time that amphipods gain when preying on intermediate and high abundances of Apatemon, in contrast to the lower predation on Diplostomum, might likely impact the energy transfer through the food web, making Apatemon a more profitable prey for amphipods. Further FR experiments on large-bodied cercariae such as Trichobilharzia might help to understand the role of cercarial size for FR types. Investigating the mechanisms that mediate changes in FRs and direct predation has important implications for trematode population and disease dynamics, as well as for ecosystem energetics.

(PDF 387 kb)