Discussion
Since the green world hypothesis proposed by Hairston et al. (1960), a number of studies have examined effects of primary production and predation (Carpenter et al. 1985; Power 1992; Shurin et al. 2002; Hambright 1994; Hanley & La Pierre 2015; Ward et al. 2015;), and those of anti-herbivory defense or edibility of producers (Coley et al. 1985; Wolfe et al.1997; Agrawal & Fishbein 2006; Poelman et al. 2008; Mooney et al. 2010) on herbivore abundance relative to producer abundance (theH/P ratio). Although elemental stoichiometry of primary producers has often been considered as a factor that determines herbivore biomass (Ceberian 1999; Sterner & Elser 2002; Ceberian et al. 2009), few studies have experimentally examined whether nutrient contents can regulate herbivore relative to producer abundance at a community level (e.g., Urabe et al. 2002). Moreover, to the best of our knowledge, no studies have determined and compared the relative importance of these factors in natural communities, presumably because no theoretical framework has been developed for quantitatively examining their effects in a comparable way. This is the first study to examine simultaneously the effects of those four factors on the H/P mass ratio in a single natural community. By fitting our observed data to a modified Lotka-Volterra-based model, we have shown that in addition to primary production and predation, edibility and stoichiometry of primary producers plays pivotal roles in regulating the H/P mass ratio.
Our model, based on Lotka-Volterra equations, is derived from the equilibrium state of a system. In this study, both phytoplankton and zooplankton biomass dynamically changed in all the treatments throughout the experiment, indicating that no community in this study reached equilibrium. However, theoretically, temporal means of herbivore and producer abundances for at least one oscillation cycle should coincide with the equilibrium abundances in the Lotka-Volterra model (Haberman 1977). Theoretical and experimental studies have also shown that one oscillation cycle occurred within < 50 days in zooplankton-phytoplankton dynamics (McCauley & Murdoch, 1990). Thus, the present experimental run would be sufficiently long so that the temporal mean values among samples collected biweekly were close to equilibrium values in zooplankton-phytoplankton dynamics.
Among the factors we examined, fish abundance had the greatest effect on changes in the H/P mass ratio. This result might be caused by the fact that, in comparison with planktonic organisms, fish abundance was temporally more stable due to their longer life span. In all sections (treatments) of pond 217, where planktivorous fishes were abundant, the density of large cladocerans in the zooplankton community was low. The result was in accordance with the well-known fact that planktivorous fish prey selectively on larger zooplankton species (Carpenter et al. 1985; Hambright 1994; Lampert &Sommer 2007). Other than the direct effect of predation, several studies suggest that fish can indirectly affect zooplankton biomass by stimulating primary productivity through nutrient recycling (Williamson et al. 2018). However, in this study, the specific production rate (μ ) was not related to fish abundance, suggesting that the net impact of fish abundance on the H/P mass ratio was largely attributable to a direct top-down force on zooplankton biomass rather than indirect bottom-up forcing through nutrient cycling.
A number of studies have argued that the H/P mass ratio is regulated by the efficacy of the producer’s anti-predator defense (Coley et al. 1985; Poelman et al. 2008; Moony et al. 2010). In this study, occurrence of algae that might have contained some toxicities for zooplankton such as cyanobacteria (Lampert & Sommer 2007; Smith & Schindler 2009) was limited. Therefore, we focused physical defense of phytoplankton. Since most herbivore plankton cannot efficiently graze phytoplankton species with a cellular or colony size larger than 30 µm (Lampert and Sommer 2007), enlargement of cellular or colony size can be viewed as a defense trait against herbivory (Pančić & Kiørboe 2018). Therefore, we examined fraction of edible phytoplankton (αedi ) on H/P mass ratio. No significant relationship was detected between these in simple regression. However, if we considered only the treatments in pond 218 where fish abundance was limited, the H/P ratio tended to increase with this fraction. Indeed, αedi was significantly related with the mass ratio when other factors, such as fish abundance, were simultaneously considered in the multiple regression analysis. These results indicate that edibility or a defense trait such as enlargement of cellular or colony sizes indeed play a role in regulating H/Pmass ratio, even in aquatic communities.
Other than the defense trait, nutritional values or nutrient contents of producers have often been proposed as crucial factors determining the abundance of herbivores relative to that of producers (Cebrian 1999; Srerner & Elser 2002; Ceberian 2009; Konno, 2016). In this study cyanobacteria biomass was less than 10%, suggesting that deficiency of polyunsaturated fatty acids were not prime factors affecting quality of phytoplankton food for zooplankton in this study (Urabe et al. 2018). Cebrian (1999) argued that a lower H/P mass ratio in terrestrial communities compared with aquatic communities is attributable to lower nitrogen and phosphorus contents relative to carbon in terrestrial producers. In this study, we focused on phosphorus as the main nutrient since phosphorus limitation of herbivore growth at an individual level has been repeatedly pointed out. Indeed, this study successfully showed that seston carbon to phosphorus ratio was a significant factor affecting H/P mass ratio across communities with different taxonomic compositions of phytoplankton and zooplankton. The result supports a theory of ecological stoichiometry which states that the relative abundance of herbivores to producers changes depending on the stoichiometric mismatch between them (Sterner & Elser, 2002).