Unearthing a new approach to gravel extraction: Can it restore habitat for freshwater fish and improve their population health?

Despite their dynamic and resilient nature, river systems are degrading worldwide, with gravel extraction a key commercial activity that significantly threatens riverbed integrity and natural functioning. However, extraction methods aiming to combine commercial aggregate harvesting and ecosystem restoration are being trialled. To compare the environmental effects of an experimental gravel and overburden skimming technique, intended to be restorative, with traditional pit mining methods for gravel extraction, we evaluated aspects of the population health of a generalist fish species as biological indicators. We examined the diet contents, parasite load, and growth rates of the upland bully (Gobiomorphus breviceps) to evaluate ecosystem integrity. These bioindicators were compared amongst three treatments - pit-mined ponds, experimental gravel and overburden skimming river restoration sites, and unmodified river reaches - on a section of the Aparima River, Southland, New Zealand. There were clear differences between the river treatments and the pit mining ponds, with fish sampled in ponds exhibiting greater parasite loads and faster growth. The more suitable conditions for growth and parasites in pit-mined ponds may be attributable to a range of factors including higher temperatures, elevated nutrient levels, and lack of flow. Sites harvested via experimental gravel and overburden skimming (restoration) and unmodified river reaches (control) did not show any difference, leaving the efficacy of the experimental gravel and overburden skimming technique method uncertain as a restoration technique for native fish conservation, but showing that extraction via this method did not have a negative effect on fish population health based on the variables assessed. This contrasts with pit mining, which, as indicated by fish population health, was more environmentally disruptive. Further development of integrative gravel extraction and river restoration methodology could potentially enhance the ecosystem functioning and native biodiversity of river ecosystems.
Mean environmental variables among treatments across sampling months - Environmental variables included are depth, nutrients (total phosphorus and total nitrogen), velocity and discharge. All variables were sampled monthly in February, March, and April, apart from POND depth which was sampled in February and April only.
The growth of upland bullies in their second year (1+) was estimated to test the effect of treatment on fish growth. The annual growth increments (annuli) of 40 sagittal otoliths from 40 upland bullies (four per site, all captured in March 2023) were independently counted by two observers. Nine upland bullies (three from each treatment) had not reached their second year of growth, so were removed from the statistical analysis. An ANOVA was used to test the influence of treatment on growth.
The preserved condition of thirty upland bullies was poor and deemed unsuitable for parasite examination. Thus, 250 upland bullies were inspected for parasite infection. The parasite load was based on Apatemon jamiesoni, the most abundant parasite of upland bully (Gobiomorphus breviceps) among the three treatments. The parasite load of each individual fish was log-transformed to reduce skewness. The mean parasite load of fish from each site was calculated from their log-transformed values and used for repeated measures ANOVA analysis.
The digestive tract of all 280 upland bullies was dissected to determine macroinvertebrate prey diversity. Twenty-five digestive tracts were empty (twenty of these were in POND fish, likely due to some fish not feeding while in traps), and thus the gut contents of 255 upland bullies were considered for analysis. The Shannon diversity index (Shannon, 1948), calculated using the ‘vegan’ package (Oksanen et al., 2022), was used to determine the diet diversity of each individual fish. The mean diet diversity of fish from each site was calculated and used for repeated measures ANOVA analysis.
EPT taxa in upland bully (Gobiomorphus breviceps) diet among three treatments. The percentage abundance by number (%N) of EPT taxa in each treatment was determined by calculating the %N of each EPT taxon for each treatment and month ((number of each EPT taxon per treatment per month / total number of taxa per treatment per month) × 100). These values were averaged across sampling months, followed by summing all taxa per EPT order.
The three treatments were: sites where an experimental gravel and overburden skimming technique that aims to restore degraded river sections is being trialled (RES), pit mining ponds, which are homogenous in structure and located on the riverbed, but separated from the active channel (POND), and unmodified, degraded river reaches that were used as “controls’’ (UNRES).