Introduction
The increasing demand for aquaculture products requires increasing both the productivity and sustainability of fish farming systems (Thomas et al., 2021). Polyculture has shown potential to do so because it can (i) use food waste better and (ii) increase nutrient recycling in the system, which can (iii) optimize resource-use efficiency and (iv) reduce environmental impacts (Aubin et al., 2021). However, nearly all polyculture systems are built empirically through trial-and-error, which is time consuming and does not allow a range of solutions to be explored. In contrast, studies that attempt to design optimized polyculture systems are rare, and even fewer aim to improve the system’s productivity and sustainability simultaneously. Thus, the objective of this study was to develop a method to design more efficient and sustainable polyculture systems. To this end, we developed and modeled scenarios of species combinations using a food-web modeling tool and then selected the best scenarios as a function of performance indicators.
Materials and methods
The complementarity and compatibility of the species combined in this study had previously been assessed in experimental ponds in Le Rheu, France (SEPURE project). The ponds were 0.1 ha in area and 1 m deep. The species differed in feeding strategy: common carp Cyprinus carpio (omnivore), pike-perch Sander lucioperca (carnivore), roach Rutilus rutilus (zoo/phyto/planktivore and detritivore) and tench Tinca tinca (benthivore). We developed 10 scenarios of polyculture (each with 10 trophic groups) with no external food supply that differed in each species’ percentage of total fish biomass at stocking. To assess the performance of these scenarios, we used Ecopath, a software package for modeling aquatic food webs (Christensen & Pauly, 2004). Biological parameters of these fish (e.g. specific growth rate, mortality rate, reproduction rate) and the productivity of macroinvertebrate, zooplankton, phytoplankton and detritus were extracted from the study of Aubin et al. (2021), which was conducted at the same study site. The individual weights of common carp, pike-perch, roach and tench used in the scenario were 10 g, 70 g, 150 g and 450 g respectively. The total stocking biomass was set to 87.5 kg/ha. Depending on the scenario, the percentage of fish species in the total biomass varied from 10-50% for carp, 5-10% for pike-perch, 22.5-48% for roach and 15-43% for tench. The duration of the rearing cycle was set to 270 days, which is sufficient for monitoring a temperate polyculture system. Each modeled scenario was assessed using indicators of agro-ecological performance that represented productivity, total system throughput (i.e. the amount of biomass flowing through the system over the rearing period), fish diversity, efficiency and recycling (Finn’s cycling index).
Results
Over the rearing cycle, the scenarios had a total system throughput of ca. 701±6 g/m2. The ecotrophic efficiency of the zooplankton, phytoplankton and detritus, which supported food resources in these unfed systems, depended on the fish species composition, varying from 0.52-0.61, 0.23-0.24 and 0.43-0.48, respectively. The species composition also influenced the percentage of biomass fluxes recycled in the system. The percentage of common carp in the scenarios was strongly and negatively correlated with Finn’s cycling index (r = _0.99), unlike the percentage of other species such as roach (r = 0.78) and tench (r = 0.87), for which the correlation was positive. As the density of common carp decreased from 50% to 10%, the percentage of biomass fluxes recycled in the system increased from 17.8% to 18.1%.
Discussion
Ecopath food-web modeling seemed sufficient to assess scenarios of species composition. However, it has some limitations due to its use of linear equations (Christensen & Pauly, 2004), which may underestimate or overestimate the correlation between variables. In addition, the integration of environmental variables in the model is limited (Plagányi & Butterworth, 2004), which may limit its ability to explain certain parameters. Variations among performance indicators of the scenarios, especially ecotrophic efficiency and recycling, indicated relations among the trophic and functional roles of species in the system. When species have different feeding strategies, they do not compete with each other for food and share effects of their functional roles in the system (Wang & Lu, 2016). Differences in overall productivity among scenarios were small since the growth performance of fish and productivity of zooplankton, phytoplankton and detritus were extracted from another study (Aubin et al., 2021). However, the ecotrophic efficiency differed among scenarios, which allowed us to identify the most efficient scenario. The scenario with the best performance indicators had the lowest percentage of common carp (i.e. 10%), which was much lower than that in most polyculture experiments (i.e. usually more than 50%; Lin et al., 2022). Since these results come from a modeling exercise, experiments are needed to validate the productivity and sustainability of the recommended scenario.
Conclusions
Ecopath’s trophic network modeling allowed us to analyze multiple scenarios of species composition in a polyculture system. The scenario with the lowest percentage of carp (10%) appeared to be the most sustainable, based on ecotrophic efficiency and Finn’s cycling index.
Acknowledgements
This study was part of the FEAMP projects SEPURE (https://www6.inrae.fr/sepureand EisaCam.
References
Aubin, J., Baizeau, V., Jaeger, C., Roucaute, M., Gamito, S., 2021. Modeling trophic webs in freshwater fishpond systems using Ecopath: towards better polyculture management. Aquaculture Environment Interactions. 13, 311-322.
Christensen, V., Pauly, D., 2004. Placing fisheries in their ecosystem context, an introduction. Elsevier, pp. 103-107
Lin, S., Milardi, M., Gao, Y., Wong, M.H., 2022. Sustainable management of non‐native grass carp as a protein source, weed‐control agent and sport fish. Aquacult. Res. 53, 5809-5824.
Plagányi, É.E., Butterworth, D.S., 2004. A critical look at the potential of Ecopath with Ecosim to assist in practical fisheries management. Afr. J. Mar. Sci. 26, 261-287.
Wang, M., Lu, M., 2016. Tilapia polyculture: a global review. Aquacult. Res. 47, 2363-2374.