A coupled hydrodynamic-ecological model of management options for restoration of Lake Ohinewai

Koi carp (Cyprinus carpio) is regarded as one of the ecologically most destructive invasive freshwater species capable of degradation of shallow aquatic ecosystems. This degradation is largely due to their benthic foraging activity which resuspends sediment and increases nitrogen (N) and phosphorus (P) concentrations, leading to algal blooms. In addition, carp have the potential to excrete large quantities of nutrients into the water column. This report describes a modelling study of a shallow Waikato lake subject to invasive fish removal to investigate how carp contribute to the overall nutrient cycling at an ecosystem scale. The study site was Lake Ohinewai, a relatively small shallow, polymictic, hypertrophic riverine lake located in the northern Waikato region. Evidence of water quality degradation has been documented since the 1980's, before the introduction of carp. Lake Ohinewai has undergone an intensive programme of invasive fish removal and a permanent adult koi carp barrier was installed on the outflow drain of Lake Ohinewai to prevent passage into the lake, but allow carp to exit. We applied DYRESM-CAEDYM (DYCD), a 1-D water quality model that has been developed at the Centre for Water Research, University of Western Australia. Daily inflows into Lake Ohinewai were derived from INCA, a process-based model. Carp-driven sediment resuspension was estimated using results from experimental ponds stocked with varying densities of carp. Carp-suspended sediment was then used to estimate carp-translocated porewater due to benthic feeding. Koi carp excretion was estimated using an allometric scaling model. Nine catchment and riparian management scenarios were tested with no fish (NF) and with current invasive koi carp removal (F). These scenarios included: simulation of current biomass; pre-removal fish biomass; reduction of external load by 50%; an all native forest catchment; climate change; DOC administered land sub-catchment retirement; enhanced lake riparian margins; stream riparian zones; and creation of a wetland. For all scenarios, specified changes in nutrient concentrations were applied to internal loads (phosphate and ammonium release rates) from the sediment. Simulations demonstrated that, at a density of 374 kg ha⁻¹ and assuming they are feeding on the benthos, koi carp can contribute 21% of total phosphorous load and 10% of total nitrogen load via nutrient translocation. The major proportion of this load was due to koi carp excretion. While koi carp introduction in Lake Ohinewai has contributed to its current hypertrophic state based on this modelling exercise, evidence of water quality degradation was originally documented prior to koi carp introduction, and has most likely been occurring since the catchment land use was originally modified from its natural state. The simulations suggest that the removal of koi carp decreased Trophic Level Index (TLI4) from 6.45 to 6.28, indicating that koi carp removal alone is not sufficient for a significant restoration of lake water quality. In order to return the system to a stable clear water state, water clarity must improve to a level to permit significant macrophyte re-establishment. However, restoration simulations showed that at a lower TLI4 the removal of koi carp is critical to lake restoration. Simulations estimated that the removal of koi carp decreased TSS concentrations by approximately two-thirds in most of the scenarios. However, in the all-native catchment scenario, removal of koi carp was predicted to increase Secchi depth by c. 2.5 m, and reduce TSS by c. 13 mg l⁻¹. The scenario simulations within the present study showed that integrated catchment management involving use of stream riparian zones and constructed wetlands would be required in order to restore lake water quality significantly. Using these mitigation techniques a TLI4 target of 5.27 would likely be achieved. This mesotrophic TLI4 may enable the reestablishment of macrophytes