The extent to which organisms from freshwater ponds and lakes will be affected by increasing pCO2 remains poorly understood. Some studies suggest that primary consumers will be indirectly affected since increased pCO2 can alter the nutritional quality of the primary producers on which they feed [16, 19]. However, potential direct effects of exposure to pCO2 on life history, independent of changes in their food, remain unclear. Here, we exposed three model species representing three main functional groups of primary consumers – a water flea, a seed shrimp and a rotifer – to both elevated and extreme pCO2 levels. We excluded indirect effects of pCO2 and found that both an elevated and an extreme pCO2 level resulted in strong life history responses in all three tested species. Since responses were taxon specific, pCO2 driven shifts in the composition of freshwater zooplankton communities are likely.
Survival in all three study species was only reduced under extreme pCO2 (~83,000 ppm) conditions which resulted in a pH of 6.33 (± 0.08 s. d.). After only three days, mortalities varied between 100% in the water flea, 23% in the seed shrimp and 31% in the rotifer. Current literature shows that weak acidification (pH 5.5-7.0) in freshwater [5], associated with high pCO2, can result in direct mortality in fish [30], snails [31], bivalves [32] and frog larvae [33]. However, mortality induced by pCO2 has not been reported in freshwater zooplankton. Here, we show that this is possible but only under extreme concentrations that are currently highly uncommon (Table S1, Appendix 1) and unlikely to become common in natural ecosystems. Problems with calcification under low pH [26, 34] could explain the observed mortality in both the water flea and the seed shrimp since they have a calcified carapace, but this is speculative.
Weiss and colleagues (2018) [6] concluded that pCO2 and not pH was responsible for the decreased expression of morphological anti-predator defenses in two Daphnia species, since this response was not observed when the same pH reduction was achieved using another acid. High levels of CO2 in the water have been shown to induce hypercapnia in crustaceans which leads to metabolic depression and decreased respiration [35, 36]. These physiological changes may impact life history traits and can potentially result in mortality. For instance, higher mortality and delayed growth after exposure to elevated pCO2 in larvae of a marine copepod was attributed to increased energetic costs to maintain metabolic homeostasis [37].
Although we did not find an effect of ~25,500 ppm pCO2 on mortality in any of the studied model species, several other key life history traits were strongly affected. For instance, growth in both the water flea and the seed shrimp was reduced under elevated pCO2. Such an effect was not detected in a previous study during which Daphnia longicephala and D. pulex were exposed to 16,000 ppm of CO2 [6]. Most likely, exposure time in that study was too short (i.e. 7 days vs. 24 days in our experiment) to induce or detect any life history responses. Reduced growth can result from a slower metabolism and increased allocation of energy towards metabolic homeostasis and defense mechanisms against acidosis [36, 37]. If elevated pCO2 can lead to a reduction in body size, such effects might be exacerbated by higher temperatures which have been linked to smaller body sizes in cold blooded animals [38], but this remains to be demonstrated. In the field, smaller body sizes might also be stimulated by indirect effects of pCO2 on phytoplankton. When a closely related water flea, D. pulicaria, was fed algae that had been previously cultured under elevated pCO2, water fleas also showed reduced growth [19]. However, this explanation is not valid for our study given that we fed dead algae grown under normal pCO2 conditions. As such, our study provides evidence for direct effects of elevated pCO2 on growth, which have not been demonstrated earlier in freshwater zooplankton.
Increased pCO2 also resulted in developmental delays. In the seed shrimp, there was an initial decrease in somatic growth of 34%, which was compensated after one week. This type of compensatory growth has been observed in other freshwater invertebrates in response to stressors [39]. However, the pCO2 exposed seed shrimp showed a delay in maturation, resulting in an 89% decrease in lifetime fecundity measured as the total no. of neonates. This indicates that short term acclimation to higher pCO2 comes with fitness costs in terms of lower per capita reproduction. In the rotifer, we found a trend of delayed population growth under elevated pCO2 levels, but it was not significant. Although such developmental delays may seem subtle, they could still impact the demography of natural populations. In short lived temporary pond systems, for instance, delayed maturation can prevent successful recruitment when populations fail to reproduce before the pond dries out [40]. In permanent systems, delays in population growth can disturb successional plankton dynamics [41].
Similar as in the seed shrimp, the total reproductive output of the water flea was also reduced by an elevated pCO2 of ~25,000 ppm. However, this was not the result of delayed maturation since water fleas were already mature at the start of the experiment. A lower number of offspring under an elevated pCO2 of 7,000 ppm was also recently observed in D. pulex, by Pötter and coworkers (2021) [21], however only in presence of predator cues. Here, we show that elevated pCO2 reduces the total number of offspring in different clonal lineages of D. magna exposed for 24 days. This is in accordance with findings of Parra and colleagues (2016) [42] where a single D. magna clone was exposed to a CO2-induced acidification from pH 8.7 to 7.0. However, remarkably this study did not report actual pCO2 values which limits conclusions and prevents direct comparison with our results.
While water fleas were most sensitive in this study, it should be noted that individuals of D. magna were mature at the start of the exposure while the other species were juvenile. However, it is reasonable to assume that juvenile water fleas are even more sensitive than adults since previous studies have shown that early developmental stages are typically more sensitive to environmental stress such as elevated pCO2 [37, 43, 44].
If the responses in these three model species turn out to be representative of the broader taxonomical groups they represent, direct effects of pCO2 might contribute to functional shifts in freshwater communities. For instance, if larger water flea species are indeed systematically more sensitive to elevated pCO2 – as suggested by higher mortality in this study - control of algal blooms by these filter feeders may be compromised. This, in turn, could lead to smaller water flea species or rotifers increasing in importance as pelagic filter feeders. However, this remains to be confirmed by exposing communities of competitors to different pCO2 treatments.
The experiment performed here ran across a longer and more ecologically relevant time-scale than some earlier experiments with zooplankton (e.g. [6]) and therefore provides a more realistic test of the likely responses of individuals. For instance, individuals may be able to acclimate to an environment that frequently experiences elevated pCO2. Such acclimation mechanisms could include an improved capacity of CO2 buffering, transport and exchange as observed in deep-sea fish and invertebrates adapted to fluctuating pCO2 environments [45]. We found a few indications for acclimation in the accelerated growth of the seed shrimp and the compensation of an initial reduction in population growth in the rotifer. As acclimation via phenotypic plasticity could be important in the field, we must be careful not to overestimate effects of pCO2 based on exposure experiments. Also, genetic diversity and associated differential sensitivity of specific genotypes may impact the response of natural populations. While the tested rotifer and seed shrimp were homozygous laboratory populations, several different genetic water flea genotypes were used. Some genotypes were more sensitive and died under ~25,500 ppm pCO2, some survived but at the cost of a reduced body size while others were relatively unaffected. This suggests adaptive potential to cope with an environment characterized by elevated pCO2 and weak acidification, at least in the water flea, but likely in all tested species.
We used natural poorly buffered pond water, of which the physicochemical properties cannot be adequately reconstructed in the lab starting from demineralized water. While this has merits with regard to realism, it also implies that the studied responses may not be representative for the full range of natural pond and lake conditions. In well buffered systems, pCO2 effects will be weaker. A logical next step would be to investigate if effects of elevated pCO2 can be effectively mitigated via the buffering capacity of the water and to test whether a similar pH drop generated by other proton donors than carbonic acid would lead to different life history responses.
Overall, this study serves as a proof of principle that pCO2 can have direct effects on representatives of different functional groups of primary consumers in freshwater. These effects can be substantial at concentrations that have already been measured in the field and could become more common for many freshwater environments in the near future. How common such conditions may become is still unknown. The reason being that pCO2 in freshwater results from a combination of physiochemical conditions such as atmospheric CO2, temperature and different biological processes such as photosynthesis and respiration e.g. via decomposition of allochthonous organic matter [14]. As such there is a need for quantitative models to assess how changes in atmospheric CO2 will be reflected in pCO2 in different freshwater systems. Nevertheless, we show that the life history of freshwater zooplankton is sensitive to the physicochemical properties of CO2. While this insight is valuable, in natural environments direct and indirect pCO2 effects (e.g. via modulation of food and pH changes) operate simultaneously, complicating the ultimate response. At the very least the current observations confirm that predictive models for the performance of aquatic organisms under different climates should not simply focus on indirect effects of CO2 but also integrate direct effects.