Introduction
One of the features that most universally affects ecological systems is ecosystem (patch) size. Ever since the foundational work by MacArthur & Wilson (1967), it has been known that ecosystem size affects diversity, abundance, and extinction risks. This finding has broad implications from theoretical ecology (Luo et al., 2022) to conservation biology (Riva et al., 2024). A second feature that universally affects ecological systems is their spatial connection through the movement of non-living materials, creating spatial coupling or even spatial feedbacks (Gravel et al., 2010; Loreau et al., 2003; Peller et al., 2024), whose effects can depend on the type of ecosystems connected (Gounand et al., 2017; Osakpolor et al., 2023; Pichon et al., 2023), as ecosystem type can determine the type and quality of resources flowing (Gounand, Harvey, et al., 2018; Gounand, Little, et al., 2018; Pichon et al., 2023; Sitters et al., 2015). However, despite recognising the ecological significance of these two features, we do not know how the size of ecosystems of different types influences the effects of resource flows on their function.
Worldwide, different and heterogeneous ecosystems are connected through flows of resources (Gounand, Little, et al., 2018), forming meta-ecosystems (Gounand, Harvey, et al., 2018; Harvey et al., 2023; Leroux et al., 2017; Loreau et al., 2003; Osakpolor et al., 2023). For example, a meta-ecosystem could be composed of a stream and a forest (Harvey et al., 2023; Leroux et al., 2017; Osakpolor et al., 2023), where leaf litter falls from the forest into the stream (Cereghetti & Altermatt, 2023; Naiman & Link, 1997), fish carcasses are brought from the stream to the forest by bears fishing (Helfield & Naiman, 2002), and emergent aquatic insects end their life cycle in the forest (Scharnweber et al., 2014). In a meta-ecosystem, ecosystems can sustain their consumers mainly through their resources (autotrophic ecosystems) or through resources coming from other ecosystems (heterotrophic ecosystems). For example, a forest can be net autotrophic because it has enough plant production to sustain the growth and reproduction of terrestrial insects. In contrast, a stream can be net heterotrophic because it does not produce enough algae to sustain insects–such as stoneflies–that, therefore, rely on the input of leaves from the terrestrial ecosystem. The flow of resources between autotrophic and heterotrophic ecosystems is widespread in nature (Gounand, Little, et al., 2018). For example, deep benthic systems (heterotrophic) are coupled to pelagic systems (autotrophic) (Griffiths et al., 2017; Schindler & Scheuerell, 2002), small headwater streams (heterotrophic) are coupled to riparian forests (autotrophic) (Carter et al., 2024; Gounand, Little, et al., 2018), and coral reefs (heterotrophic) are coupled to pelagic ecosystems (Morais et al., 2021; Morais & Bellwood, 2019).
Resources flowing between autotrophic and heterotrophic ecosystems can positively or negatively affect the overall meta-ecosystem production (Allen et al., 2024; Gounand et al., 2017; Pichon et al., 2023). Resource flows are predicted to increase the production of the two ecosystems if the production of the two systems is limited by the compounds that the other provides (Pichon et al., 2023). By contrast, resource flows would likely decrease production if the ecosystems provided each other with unnecessary compounds, making limiting resources less dense and, therefore, less available (Pichon et al., 2023). For example, if a stream connected to a forest receives leaf litter rich in carbon, this flow of resources will increase stream production if the stream is carbon-limited and decrease it if it is nitrogen-limited (Pichon et al., 2023). Also, if a stream is connected to a terrestrial ecosystem from which it receives too many resources such as hippo faeces, these resources cannot be absorbed by algae as quickly as received, and subsequently provoke hypoxia, reducing fish function (Dutton et al., 2018). As the connection among autotrophic and heterotrophic ecosystems via resource flows is widespread in nature, understanding how their connection influences their production will deepen our understanding of a ubiquitous ecological process, ultimately enhancing our ability to manage landscapes effectively.
However, how the size of the autotrophic and heterotrophic ecosystems shape the effects of this spatial coupling has been overlooked. The size of an ecosystem can alter its permeability to resource flows and, therefore, how they affect its production. This interaction between ecosystem size and resource flows was shown, for example, by field studies on unidirectional flows of resources to islands, where smaller islands get proportionally more resources coming from the ocean than larger ones, increasing the density of spiders and insects in islands in Baja California (Mexico) (Polis & Hurd, 1995) and birds in islands in British Columbia (Canada) (Obrist et al., 2020). Furthermore, resources flowing bidirectionally between ecosystems of different sizes can change the overall biomass in these ecosystems, as shown by a protist experiment (Giacomuzzo et al., 2024). In light of the influence of ecosystem size on meta-ecosystem total biomass and the fact that ecosystems have different sizes in nature, we expect that including ecosystem size should be important in understanding how resource flows between autotrophic and heterotrophic ecosystems change the total biomass of meta-ecosystems.
Here, we studied how relative ecosystem size mediates the effects of spatial couplings on meta-ecosystem biomass using a protist experiment. Such microcosm experiments are a well-known approach to study general questions in ecology and allow a high level of control and a close conceptual analogy to mathematical models, yet still contain realistic ecological dynamics (Altermatt et al., 2015; Benton et al., 2007). We studied two-patch meta-ecosystems composed of one autotrophic and one heterotrophic patch with only non-living resources flowing between the two ecosystems (‘patch’ and ‘ecosystem’ are used as synonyms). We investigated the effects of resource flows in three meta-ecosystems with the same total volume but different relative patch sizes: one in which the autotrophic patch was larger, one in which the heterotrophic patch was larger, and one in which both patches were of the same size. We investigated the effects of resource flows by comparing the same three meta-ecosystems (varying in patch size of heterotrophic and autotrophic systems) and manipulating if patches were coupled or not (i.e., connected or unconnected by resource flows).