Figure 2: Relative contribution of prey carbon biomass to the total ingested biomass by Limnomysis benedeni plotted as a function of the relative contribution of each prey to total available carbon biomass. Prey types used are Cryptomonas sp. (dark green) andB. calyciflorus (orange). Solid dots represent samples while error bars indicate the 95% confidence intervals. The dashed 1:1 line represents where L. benedeni would be consuming a prey proportional to its availability, meaning random food uptake (neither preference, nor avoidance).
Discussion
Our results clearly showed that the diet preference of L. benedeni depends on the relative (C) biomass concentration of prey types, by shifting from microzooplankton to phytoplankton prey with increasing phytoplankton prey biomass. The patterns of consumption rates in response to relative prey biomass differed between the prey types suggesting slightly different functional responses for the microzooplankton and phytoplankton prey. While both resembled a sigmoidal pattern (Fig. 1A), the ingestion rate on B. calyciflorus followed a type III functional response more clearly, which usually indicates either prey switching or efforts to modify foraging behaviour in response to the prey density (Holling, 1959; Kiørboe et al., 2018; Thomas et al., 1996). In the case ofCryptomonas sp. after a slower increase at low densities, the pattern was closer to linear and did not show saturation within the tested biomass range during the experiment, overall resembling a type I functional response (Holling, 1959), suggesting that consumption of Cryptomonas sp. is primarily a function of density.
The decreasing consumption rates on rotifers with increasing algae concentration illustrated how the presence of alternative prey items in sufficient densities can lead to decreased predation pressure on another type of prey. If more than one alternative food source is present, predators can select based on prey-specific encounter rates to optimize foraging time and energy intake (Krebs et al., 1977). Specifically, the relative attack rate is a function of relative density and the switching can happen only if the likelihood of attacking the last eaten species is higher than attacking other species (Oaten & Murdoch, 1975). In our case, the cost-to-benefit ratio changes in favour of algae at higher algae densities. Indeed, we found that algae consumption increased in comparison to rotifer consumption for biomass equal to or higher than 1 mgC L-1. This is also supported by Ivlev’s index results which point towards the flexibility in the feeding behaviour of L. benedeni depending on algal density. The Ivlev’s indices indicate a preference for rotifers in the presence of low algae biomass and avoidance of algae and vice versa with high algal biomass.
Predators typically select their diet based on size, nutritional quality, and escape responses of the prey (Cotonnec et al., 2001; Frost, 1972; Viitasalo & Rautio, 1998). They would try to increase their fitness by acquiring prey with easier access and lower cost-to-benefit ratio (MacArthur & Pianka, 1966; Pyke et al., 1977; Stephens & Krebs, 1986). Earlier studies suggest that L. benedeni is not an optically-oriented predator but selects its food based on size (Fink et al., 2012). Here we provided L. benedeni with two types of food with very different sizes. Purely from a size-based perspective, feeding on the smaller prey is only beneficial when available in sufficient densities, which was indeed confirmed by our results, with L. benedenishowing a gradual shift to the smaller prey with its increasing quantity.
Prey nutritional quality can also affect the development of consumers and influence their dietary choices (Gulati & Demott, 1997; Meunier et al., 2016). For instance, algae generally have a higher carbon-to-nutrient ratio than animal prey (Elser et al., 2000; Sterner & Hessen, 1994). In the case of Cryptomonas SAG 26.80, a C:P ratio of ~150 and a C:N ratio of ~22 were reported during the exponential growth phase (i.e. when nutrients are not limiting growth; Vad et al., 2020). For B. calyciflorus , C:P ratios are relatively constant around ~92, while C:N are ~4 (Jensen & Verschoor, 2004), which is very close to the body stoichiometry of omnivorous crustaceans such as mysids (C:P~90 and C:N ~4) (Arbačiauskas et al., 2013). Therefore, based on a purely elemental stoichiometric point of view (Laspoumaderes et al., 2010), rotifers would be the preferred prey. The higher C to nutrient ratios of algae may explain the pattern we found in the case of total ingested carbon biomass, which instead of steadily increasing showed an accelerating increase at the highest Cryptomonas sp. biomass level. This may indicate that L. benedeni needs to increase overall food uptake when feeding on algae, to obtain sufficient amounts of nutrients.
Other factors such as essential fatty acids (Ahlgren et al., 1990; Brett & Müller-Navarra, 1997; Ramlee et al., 2021; Trommer et al., 2019), or vitamins (Fridolfsson et al., 2018, 2019; Hessen, 1992) can also influence food quality.Cryptomonas sp. is considered to be a high-quality algal food source based on the high cellular content of essential fatty acids (Von Elert & Stampfl, 2000; Weers & Gulati, 1997). Consequently, the results may differ if the autotrophic food quality is less cost-efficient for the predator than the heterotrophic prey. For instance, had a ‘less favourable’ algal food source, such as green algae, been used, the preference for rotifers might have been more pronounced. Nonetheless, a preference for green algae (Chlamydomonas sp. ) over zooplankton has been observed previously at an algal biomass of 0.3 mg POC L-1 as well (Fink et al., 2012).
Given that our study is based on a short-term experiment, there are some aspects in which its results might differ from patterns arising in natural communities. For instance, the container size in laboratory studies is a factor that may influence predator behaviour (Bergström & Englund, 2004; Toonen & Fu-Shiang, 1993). Therefore, a small-sized jar could have increased the competition among mysids thus forcing them to choose algae, a ‘costly prey’, to consume maximum food to maintain their energy budget. In addition, in this grazing experiment, we did not include predators of mysids, though in natural ecosystems optimal foraging strategy is based on a tradeoff between nutritional needs and simultaneously minimising the risk of predation as described in the unified foraging theories (e.g., Mangel & Clark, 1986). Experiments with another trophic level (e.g., small fish), carried out in larger mesocosms, could therefore provide more specific predictions on these tradeoffs in the future, including longer-term effects on ecosystem stability.
Despite these limitations, our results provide important implications for plankton community dynamics in natural ecosystems. Depending on the relative biomasses of phyto- and zooplankton, mysid shrimps may act as predators or competitors of zooplankton, being intraguild predators of the latter. By always suppressing the more abundant planktonic prey, they may reduce the amplitude of predator-prey oscillations, thereby contributing to ecosystem stability. However, there are certainly other possible scenarios as well, as illustrated by the effects of other invasive omnivorous Ponto-Caspian mysids. For example, Hemimysis anomala and Paramysis lacustris were both found to contribute to the alteration of trophic pathways in their invaded habitats, due to their strong predatory impact on zooplankton and benthic macroinvertebrate communities (Evans et al., 2018; Ketelaars et al., 1999; Rakauskas, 2019). For a better understanding of the potential effects of omnivorous mysids on trophic cascades and food web stability, studying their feeding mechanisms and diet selection is of high importance. Field observations coupled with laboratory and mesocosms studies could contribute to understanding the mechanisms underlying community and ecosystem-level effects of these widespread (and still spreading) invasive species.
Acknowledgements
This study was supported by the NKFIH-138215 project. We acknowledge the valuable suggestions given by Pavel Kratina and Peter Borza.
Author’s contributions
Conflict of Interest Statement
The authors declare no conflict of interest.