Discussion
We experimentally demonstrated that relative ecosystem size can mediate the impact of resource flows on the function of autotrophic-heterotrophic meta-ecosystems. Positive effects of resource flow exchanges on meta-ecosystem total biomass were observed when the larger patch was autotrophic, while the effects were negative when the larger patch was heterotrophic. These findings were explained by the coupling via resource flows between autotrophic and heterotrophic patches increasing the local biomass density of autotrophic patches while decreasing it in heterotrophic patches. Overall, our findings suggest that resource flows between autotrophic and heterotrophic ecosystems of different sizes can affect ecosystem functions at larger scales and that the trophic balance of the larger ecosystem (net autotrophic vs. heterotrophic) can determine the direction of the meta-scale effect.
The negative impact of resource flows on heterotrophic ecosystem biomass contrasts with previous results. We expected the carbon-enriched resources from autotrophs to have a bottom-up positive effect on heterotrophic protist biomass via a boost of the bacterial populations they prey on (Gounand et al., 2017). We see at least two mutually non-exclusive explanations for this counterintuitive result: stoichiometric mismatch or oxygen limitation. A stoichiometric mismatch would occur if imported resources dilute the limiting nutrients compared to local resources (Pichon et al., 2023). This would mean that our heterotrophic ecosystems were not carbon-limited, unlike in Gounand et al. (2017). A difference with this previous experiment is the higher diversity of our heterotrophic protist community (up to 10 protist species compared to one or two), which could have caused more efficient recycling (Delong & Gibert, 2019) and made the heterotrophic patch less responsive to carbon-rich resources coming from the autotrophic patch in the equally-dominated meta-ecosystems. Alternatively, resource flows from autotrophic patches may have caused hypoxia through increased bacterial growth. This might have negatively affected protist biomass in small heterotrophic patches due to high relative carbon import and in large ones due to pre-existing oxygen limitation from a low area-volume ratio. Similar hypoxia from resource flows is widespread in nature, for instance, when massive amounts of hippo faeces and urine enter confined areas in streams (Dutton et al., 2018) or when fertilisers leach into lakes or estuaries (e.g., Rabalais & Turner, 2019; Steffen et al., 2012).
In contrast, the positive effect of resources coming from heterotrophic ecosystems on the local production of autotrophic ecosystems has consistently been observed in experiments (e.g., Gounand et al., 2017; Harvey et al., 2016), as well as in empirical studies (Montagano et al., 2019). This positive effect likely comes from relaxing the nutrient limitation in autotroph species, which in our control settings might just emerge from the differences in the stoichiometry of the abiotic resources between ecosystem types resulting from carbon fixation and direct use of inorganic nutrients in autotrophic ecosystems. In this context, heterotrophic ecosystems export resources that are much richer in nutrients relative to carbon, which should boost autotrophic ecosystem production.
Our study suggests that resource flows between ecosystems of different sizes can exacerbate the differences in production between meta-ecosystems that were already present in their unconnected counterparts. Autotrophic patches in our experiment sustained higher biomass density than heterotrophic patches, leading to a decrease in meta-ecosystem total biomass from autotrophic-dominated to equally-dominated, and heterotrophic-dominated unconnected meta-ecosystems (without resource flows connecting ecosystems). The total biomass of a meta-ecosystem was, therefore, influenced by how much of the size of the meta-ecosystem was allocated to the two types of patches. Adding resource flows between patches allowed autotrophic-dominated meta-ecosystems to become even more productive and heterotrophic-dominated meta-ecosystems to become even less productive. This suggests that disparities in function among ecosystems within a landscape may arise not only from their intrinsic properties but also from differences in their size, which influence resource flows and their effects. Further, our findings suggest that ongoing changes to resource flows connecting ecosystems around the world (e.g., Elser & Bennett, 2011; Peller & Altermatt, 2024; Smith et al., 2025) can significantly impact ecosystem function across scales.
The results of this experiment also show that relative ecosystem size influences the effects of resource flows and that it matters if the heterotrophic or autotrophic patch is dominating in size, even when resource flow quantities are equal and bi-directional. Unidirectional resource flow effects on the recipient ecosystem have been shown in field surveys and experiments to depend on the size of the ecosystem they enter. For example, marine subsidies (e.g., seaweed detritus and seabird fish scraps) increase secondary production the most in smaller islands (Polis & Hurd, 1995, 1996), nitrogen brought by salmon gets incorporated into plants and invertebrates the most in smaller river watersheds (Hocking & Reimchen, 2009), and ocean seaweed provided to the shores of islands increases lizards densities only on small islands (Wright et al., 2020). Many ecosystems, however, are reciprocally coupled, with resources flowing in both directions between ecosystems (Gounand et al., 2020; Gounand, Little, et al., 2018). A recent protist experiment manipulating the relative size of connected patches showed that the connection through flows of resources could decrease the total biomass of the meta-ecosystem (Giacomuzzo et al., 2024). While this work explored the interaction between resource flows and ecosystem size, it did not address the common case where the connected ecosystems differ in type or trophic status. Our findings support the results of Giacomuzzo et al., (2024), but significantly extend our knowledge by showing that the trophic status of the ecosystem dominating the meta-ecosystem’s size, can reverse the impact of resource flows on meta-ecosystem function. Global change is altering the size of ecosystems, and our findings suggest the potential for resource flows to mediate the effects on ecosystem function.
An important factor to consider when interpreting these results in relation to natural systems is the role of ecotones or the interfaces between connected ecosystems. In our microcosm experiment, ecosystems are relatively well-mixed and thus, resource flows can influence all areas of the ecosystem. However, in many natural systems, the extent of the ecotone, rather than the total size of the ecosystem, determines the extent of coupling between ecosystems (Muehlbauer et al., 2014). For instance, not all parts of a forest receive resource flows from an adjacent river; rather, the sections at the interface (ecotone) are predominantly involved (Harvey et al., 2023; Muehlbauer et al., 2014). Consequently, it can be that in some landscapes the size of ecosystems affected by resource exchanges is determined by the size of their ecotones, while in others it is more affected by the spatial structure/layout. Such finding is also consistent with continental-scale observations that landscape diversity, that is, the diversity of different ecosystems, per se can result in positive ecosystem function effects (Mayor et al., 2025; Oehri et al., 2020), which has been directly postulated through resource couplings and other cross-ecosystem dynamics. In the context of our study, this suggests that the influence of size may not be solely determined by the total area of autotrophic and heterotrophic patches, but by the spatial extent of their ecotone, where resource exchanges are most pronounced. Notably, ecotones are recognized as relatively dynamic areas that naturally change in size across time (Smith & Goetz, 2021). Therefore, our results can potentially explain why resource flows have different effects at different times (Anderson et al., 2008; Mulholland et al., 2006; Valett et al., 2008).
In conclusion, our experiment provides experimental evidence that the relative size of coupled autotrophic and heterotrophic ecosystems can indirectly affect ecosystem function at local and meta-ecosystem scales. Autotrophic and heterotrophic ecosystems are ubiquitously connected by spatial flows of resources, yet these ecosystems commonly differ in size—a feature that has generally been overlooked in studies of spatial flows. Our findings underscore the importance of considering how ecosystem size and type, as well as resource flow dynamics, interact to drive ecosystem function across scales. By acknowledging the role of ecosystem size and resource flows, we can better address the complexities of ecosystem function in the context of ongoing global change.