Mixing two species can have contrasting effects on the focal species
Based on the stress-gradient hypothesis (SGH, e.g. Callaway et al., 2002), we hypothesized that in a more stressful environment we would observe more facilitative interactions while in the medium without Zn pollution we would observe more competitive interactions. We did not find evidence for this hypothesis, since the levels of Zn pollution we used in our experiment were not high enough to impact duckweed growth rates. Instead, we found that species interactions were highly dependent on focal species identity and competitor identity. For example, the effect of the setting on growth rate was only consistent for one of the three species, L. gibba . However, contrary to expectations, it did not profit from growing in the presence of a second species that could have helped it to accumulate Zn. Instead, L. gibba grew the best alone. Previous research with L. gibba competing against a different duckweed species, the greater duckweed Spirodela polyrhiza , showed that L. gibba was highly competitive due to its gibbous fronds which could overgrow S. polyrhiza (Clatworthy & Harper, 1962). However, in our experiment there was no physical contact between the species and L. gibba only produced non-gibbous fronds. The outcome of species pairings could, therefore, be dependent on frond morphology, which in turn is dependent on the environmental conditions. Gibbosity in L. gibba depends on the environmental conditions, and may be restricted to optimal growth conditions, including very high nutrient availability (Vaughan & Baker, 1994). Under our conditions in the experiment, L. gibba may have had a disadvantage due to the lack of gibbous fronds and the inability to overgrow the competitor.
For the other two species, growth performance depended on the pairing, but there was no evidence of facilitation. Instead, we observed strong competition between L. minor and L. turionifera , with L. minor outcompeting L. turionifera . To the best of our knowledge, this is the first study investigating L. turionifera growth rates in the presence or absence of a competitor. Recently, it has been found that some populations previously thought to be L. minor were in fact L. turionifera (Senevirathna et al., 2021). It is thus possible that L. turionifera has a wider geographic distribution than previously thought and is, therefore, a good candidate for phytoremediation in many regions of the world.
For L. minor, we found that, even though it profited from growing with L. turionifera , the presence of L. gibba had a negative impact on its growth. Simultaneously, L. gibba was negatively influenced by the presence of L. minor . Thus, this pairing had a negative effect on both interacting species, which is surprising given that they frequently co-exist in nature. Due to their morphological similarity, L. gibba and L. minor have rarely been used in competition experiments. In instances where they did compete in experiments the analysis was not completed (Clatworthy & Harper, 1962; Wołek, 1972), or the two species were even grouped together (Peeters et al., 2016). As an exception, Rejmánková (1975) found that in gibbous form, L. gibba was always the stronger competitor, overgrowing L. minor . Thus, L. gibba may have had the disadvantage of non-gibbous fronds in our experiment.
Duckweeds have great potential for heavy metal removal in wastewaters (Abdel-Gawad et al., 2020) but many questions remain, in particular in terms of how polycultures of multiple species may improve efficiency. Our study is a first assessment of the interaction between competition and tolerance to Zn pollution for a subset of Lemnaceae species. A next step could be to test the accumulation of Zn not only for species in an isolated setting across different metal concentrations (Lahive, O’Callaghan, et al., 2011) but also in a mixed setting and, importantly, in a natural environment. In addition, due to their fast growth, it is possible that duckweed species will evolve in response to their competitor (Hart et al., 2019). For future phytoremediation efforts, species mixing could be interesting, but the effects of a second species need to be evaluated first as they cannot be readily predicted from the performance in individual cultures.