Beyond the fish-Daphnia paradigm: testing the potential for pygmy backswimmers (Neoplea striola) to cause trophic cascades in subtropical ponds

Introduction

An extensive body of literature has demonstrated the importance of food web effects, including indirect top-down food web effects, or trophic cascades, with much of this work focusing on freshwater ecosystems from an early time (Ripple et al., 2016). Many of the studies on trophic cascades in freshwater pelagic ecosystems have focused on large-bodied cladocerans, especially Daphnia, as the herbivorous prey linking predators to autotrophs (e.g., Carpenter et al., 2001; Walsh, Lathrop & Vander Zanden, 2017). The paradigm that developed from this research is that trophic cascades are common in lakes and ponds, and operate via a fish-Daphnia-phytoplankton food chain (Pujoni et al., 2016). However, a relative paucity of such studies were performed in tropical and subtropical lowlands, where Daphnia is rare (Dumont, 1994). This subset of studies suggests that the fish-Daphnia-phytoplankton cascade model does not apply well in warm climates, due to the lack of large herbivorous zooplankton and other factors including increased omnivory by fish and increased abundance of invertebrate predators (Rejas et al., 2005; Pujoni et al., 2016; Lacerot et al., 2021). As a result, the effects of predator removal and addition seem to be more variable in (sub)tropical lakes and ponds, and general predictions remain elusive (Rejas et al., 2005; Pujoni et al., 2016; Mayer, 2020).

Another bias in freshwater trophic cascade research, including both temperate and tropical studies, is a focus on the effects of fish rather than invertebrate predators. This bias is notable because fish tend to be absent from small and temporary water bodies whereas invertebrate predators are present in essentially all freshwater bodies. Certain invertebrate taxa such as Bythotrephes, Chaoborus, and Notonectidae have received attention, with studies showing they can have important effects on both traits and biomass of lower trophic levels (Carpenter et al., 1992; Carpenter et al., 2001; Walsh, Lathrop & Vander Zanden, 2017; Ingram & Burns, 2018; Horppila et al., 2019). However, other studies have found that invertebrate predators have weak to nonexistent top-down effects, and the food web effects of many common and widespread invertebrate predators remain unknown (e.g., Mermillod-Blondin et al., 2020). Characterizing the effects of these understudied predaceous invertebrate species is important for understanding their role in ecosystems and for conservation planning, as predators and freshwater species are at a higher extinction risk than species in lower trophic levels or other habitats (Purvis et al., 2000; Strayer & Dudgeon, 2010).

Members of the family Pleidae are small but common heteropteran insects closely related to the family Notonectidae which includes Notonecta (backswimmers). Pleids have been much less studied than their larger cousins the notonectids, especially Notonecta and Buenoa. In this paper we specifically study the pleid Neoplea striola (hereafter, Neoplea) as our manipulated zooplanktivorous predator. Neoplea is a widespread inhabitant of lentic freshwater in Central and North America, and is known to tolerate low oxygen conditions (Gittelman, 1975). It is a small insect, with adults measuring 1.5 mm in length. Neoplea is an active hunter which uses sight, vibrations, and possibly chemicals to sense its prey (Papacek, 2001). They have been shown to attack and consume small zooplankton such as rotifers, mesozooplankton such as Daphnia, and even prey as large as dipteran larvae (Gittelman, 1974; Hampton & Gilbert, 2001; Papacek, 2001). However, the effects of Neoplea on pond communities are still not fully understood, including whether they can cause trophic cascades.

Here we report the results of a field mesocosm experiment in which we manipulated densities of Neoplea and zooplankton to test the effects of both of these trophic groups on plankton composition and biomass, including direct effects on the next trophic level and indirect (trophic cascade) effects of Neoplea on phytoplankton. Our plankton communities were composed of freshwater plankton collected locally in central Texas, with no Daphnia or other large-bodied cladocerans present. Due to the fast escape response of copepods, we expected Neoplea would primarily affect other zooplankton, especially cladocerans and ostracods (O’Brien, 1979). We further predicted that zooplankton would reduce the biovolume of total phytoplankton, as phytoplankton is the primary food source for the zooplankton in this system, and because this aligns with most freshwater pelagic trophic cascade research including some of the limited (sub)tropical literature (Menezes, Attayde & Rivera Vasconcelos, 2010; Mayer, 2020). Following logically from these first two predictions, we predicted that Neoplea would cause a trophic cascade, i.e., indirectly increase total phytoplankton biovolume via its suppression of herbivorous zooplankton. We expected a low density of Neoplea to have weaker effects, leading to plankton densities intermediate between no Neoplea and a higher density of Neoplea. However, we found no evidence for food web effects of Neoplea and only weak evidence for effects of zooplankton on phytoplankton in the experiment. This study helps to fill a gap in knowledge about the top-down effects of invertebrate predators in freshwater bodies with warm climates.

Materials & Methods

Results

On average Neoplea survived at a rate of 74%, with no significant difference in survival between the two densities (d.f. = 7.98, t =  − 0.518, p = 0.618). No evidence was found of Neoplea reproduction during the experiment.

By the sampling dates, the zooplankton community had become dominated by copepods (Table S2). However, cladocerans, Spirostomum, and rotifers all individually composed at least 15% of the zooplankton mass on average in one or more treatments (Fig. S1). A small amount of zooplankton, mostly copepods, became established in the control tanks receiving no zooplankton (Fig. S1, Fig. 1). The zooplankton additions were successful in sustaining a significant increase in biomass of all zooplankton groups (Tables 1, 2, Fig. 1). However, Neoplea additions had no significant effects on the biomass of any zooplankton group (Fig. 1, Table 1). Mean copepod mass was higher and the mean proportion of rotifers was lower the more Neoplea were added (Fig. S1), but neither copepod nor rotifer mass was significantly affected by Neoplea (Fig. 1, Table 1).

The phytoplankton community became dominated by ovoid single-celled green algae, with pennate diatoms contributing the next most biovolume in tanks with no zooplankton added and Oocystis contributing the next most biovolume in tanks with 80 Neoplea added (Fig. S2, Table S1). While the average total phytoplankton biovolume with zooplankton was less than half of the average without zooplankton, there was no significant effect of zooplankton addition on total phytoplankton biovolume due to large variation within treatments (Fig. S2, Fig. 2A, Tables 3A, 4A). The summed biovolume of the largest morphospecies, including the larger ovoid chlorophytes, pennate diatoms, and Oocystis, was 77.9% lower in the presence of zooplankton, although the difference was only marginally significant (Fig. 2B, Tables 3B, 4B). On the other hand, summed biovolume of the smaller morphospecies (small ovoid chlorophytes, Chlorella, Selenastrum, and photosynthetic picoplankton) was not significantly impacted by zooplankton addition (Fig. 2C, Tables 3C, 4C). Similarly, Neoplea addition had no significant effects on the biovolume of any phytoplankton grouping (Fig. 2, Table 3).

Discussion

Our results show that Neoplea had no demonstrable effect on the plankton in this experiment. Furthermore, the copepod-dominated zooplankton only caused a marginally significant reduction of larger phytoplankton taxa. Any effect of Neoplea on phytoplankton composition or biomass would have likely been mediated by an effect on zooplankton composition or biomass, as Neoplea are carnivorous. Therefore it is unsurprising there was no indirect effect of Neoplea on phytoplankton, or trophic cascade, considering there were only weak or nonexistent direct effects between the three trophic levels.

There are several potential reasons why Neoplea did not reduce the biomass of any zooplankton groups in our experiment. It is possible that Neoplea consumed chironomid larvae, which were common in the mesocosms but not effectively sampled due to their benthic nature and our sampling methods which targeted plankton. Neoplea has been observed to consume chironomid larvae (Papacek, 2001). Indeed, the Neoplea were most commonly observed clinging to the sides and bottom of the mesocosms, putting them in close proximity to the chironomids. In nature, Neoplea is normally found clinging to submerged vegetation (Gittelman, 1974). However, it was impossible to observe the Neoplea throughout most of the experiment due to the high density of phytoplankton. This high turbidity also may have made it difficult for the Neoplea to hunt effectively, as they can detect prey visually and are most commonly found in clear waters, though they can also detect prey by tactile and possibly chemosensory methods (Gittelman, 1974; Papacek, 2001). When the Neoplea did swim through the water column, they would have mostly encountered copepods, the largest and dominant group of zooplankton. However, even in clear water copepods are relatively resistant to predation due to their fast escape response. Among copepods, diaptomids have an especially fast escape response, and the diaptomid Arctodiaptomus dominated our mesocosms (O’Brien, 1979). Therefore it may not be surprising that the Neoplea were unsuccessful in suppressing the zooplankton in the experiment. While other more easily captured prey were also present, it may have been difficult for the Neoplea to encounter these rare prey among all the copepods in a turbid environment.

The food web ecology of Neoplea and pleids generally will remain unclear without further research. More behavioral research on these diminutive predators is needed to better understand their hunting habits, such as where in the habitat they make most of their captures, their relative dependence on different senses for prey detection, and their relative preference for various prey. Experimental work is needed to better quantify the ecology of pleids, and these experiments will likely benefit from better replicating their preferred habitat of clear, still water with submerged vegetation. A higher population density may be necessary for their food web effects to be apparent when they are the sole predator.

The weak effects of zooplankton on phytoplankton in the experiment may be explained by the dominance of copepods in the zooplankton community. Unlike Daphnia, copepods do not generally impose strong top-down control on community phytoplankton biomass, due in part to their selective grazing on larger phytoplankton (Sommer & Sommer, 2006). Indeed, the copepod-dominated zooplankton only reduced larger phytoplankton, and even this effect was marginal, fitting the general understanding of copepods’ top-down effects on phytoplankton in ecosystems with copepod-dominated herbivore communities such as open oceans (Sommer & Sommer, 2006). Such weak top-down control of crustacean zooplankton on phytoplankton may be more pervasive at low latitudes, as Daphnia is largely absent from tropical and subtropical lowland freshwaters and copepods are instead more likely to dominate (Dumont, 1994; Havens & Beaver, 2011).

Between the weak effect of copepods on phytoplankton biomass and the ability of copepods to evade predation much more easily than Daphnia, it appears likely that trophic cascades mediated by zooplankton are less common in warm, low-latitude lakes and ponds than in their colder counterparts (Rejas et al., 2005). If this is the case, then differences in plankton biomass among these warm-climate lakes and ponds may stem largely from bottom-up effects. For example, phytoplankton biomass may be largely controlled by nutrient loading and zooplankton biomass may be generally controlled by access to edible phytoplankton. Furthermore, the strategy of biomanipulation –or top predator addition to control algal blooms in eutrophic lakes –may have a low chance of success in warm climates, at least with invertebrate predators. Instead, the often more difficult method of reducing nutrient loading may be necessary to control algal blooms in fishless warm-climate ponds.

Conclusions

We predicted that Neoplea would suppress non-copepod zooplankton, and that zooplankton would suppress phytoplankton, resulting in a trophic cascade where Neoplea indirectly increased phytoplankton biomass. Instead, Neoplea had no significant effect on plankton biomass or composition in this field mesocosm experiment. The zooplankton, dominated by copepods and lacking Daphnia as is typical of lowland tropical and subtropical lakes and ponds, only weakly reduced larger phytoplankton. While our data cannot definitively explain these weak effects, they could have resulted from Neoplea consuming benthic prey which was not effectively sampled, or from the dominance of copepods which are adept at evading capture and which selectively feed on larger phytoplankton. This study suggests that lentic ecosystems dominated by Neoplea and copepods may be characterized by weak top-down control. This represents another example of a weak or non-existent trophic cascade in lowland tropical or subtropical lentic freshwater, which appears to be more likely in these systems than in colder lakes and ponds. However, more research will be needed to achieve a clearer understanding of the ecological impacts of pleids. A better understanding of the ecology of these and other understudied freshwater invertebrate predators will be important for conservation planning, as predators and freshwater species are more threatened with extinction than lower trophic levels or terrestrial or marine species.

Supplemental Information