Data for "Declines in freshwater mussel density, size and productivity in the River Thames over the past half century"

Abstract 1. A pioneering, quantitative study published in the Journal of Animal Ecology in 1966 on freshwater mussel populations in the River Thames, UK, continues to be cited extensively as evidence of the major contribution that mussels make to benthic biomass and ecosystem functioning in global river ecosystems. 2. Ecological alteration, as well as declines in freshwater mussel populations elsewhere, suggest that changes to mussel populations in the River Thames are likely to have occurred over the half century since this study. 3. We resurveyed the site reported in Negus (1966) and quantified the changes in mussel population density, species composition, growth patterns and productivity. 4. We found large declines in population density for all unionid species. The duck mussel Anodonta anatina decreased to 1.1% of 1964 density. The painter’s mussel Unio pictorum fell to 3.2% of 1964 density. The swollen river mussel Unio tumidus showed statistically non-significant declines. In contrast to 1964, in 2020 we found no living specimens of the depressed river mussel Pseudanodonta complanata (classified as Vulnerable by the IUCN Red List) but found new records of the invasive, non-native zebra mussel Dreissena polymorpha and Asian clam Corbicula fluminea. Additionally, we found strong decreases in size-at-age for all species, which now grow to 65-90% of maximum lengths in 1964. As a result of reduced density and size, estimated annual biomass production fell to 7.5% of 1964 levels. 5. Since mussels can be important to ecosystem functioning, providing key regulating and provisioning services, the declines we found imply substantial degradation of freshwater ecosystem services in the River Thames, one of the UK’s largest rivers. Our study also highlights the importance to conservationists and ecologists of updating and validating assumptions and data about wild populations, which in the present era of anthropogenic ecosystem alteration are undergoing significant and rapid changes. Regular population surveys of key species are essential to maintain an accurate picture of ecosystem health and to guide management. Data collection Sampling was conducted on 22 September 2020 in the River Thames adjacent to Wokingham Waterside Centre (51°27'35.7"N 0°56'34.3"W). The Thames is a 346km river with a densely populated catchment covering the southeast of England, comprising both tidal and non-tidal stretches and with 45 navigation locks and associated weirs. The study area was located in the non-tidal stretch, approximately 152km downstream of the source, and directly downstream of the city of Reading. Mean flow for the study area is 37.9 m3s-1 (data from the UK National River Flow Archive). By consulting maps published in Mann (1965) and Negus (1966) we were able to resurvey the identical locality to that surveyed in 1964, a 250m stretch of river varying in width from 50 to 65m. Data collection Sampling was designed to replicate as closely as possible the methods employed by Negus (1966). Consultation of the original paper was supplemented by in-person discussions with the original author. We sampled across four depth zones: 0-1m (n=32), 1-2m (n=32), 2-3m (n=14) and 3-4m (n=15), for a total of 93 samples, compared with a total of 24 samples conducted by Negus. We took equal numbers of replicates from each side of the river for each depth zone and allocating sampling effort proportionally to the different microhabitats present. Sampling in the 0-1m and 1-2m depth zones was conducted using randomly-placed 1m2 quadrats, with all live unionids within the quadrat area collected and transported to the lab for measurement. Sampling in the 2-3m and 3-4m zones was conducted by dredging (dredge width 45cm, mesh size 15mm) from a boat along replicate 20m-long upriver transects. Since dredge transects covered a greater area than the quadrats in shallower zones, we conducted fewer replicates for these deeper zones. The dredges used in 1964 and the present study were similar in design and conformed to the National Rivers Authority (1996) and Environment Agency (present) specifications, including a rectangular frame and angled blade to enable sampling from sediments including gravel, silt and mud. The dredges were also operated in a similar way, with samples collected by towing from a motor boat. The total area dredged was 580m2, compared with a total dredged sample area of 23.22m2 in 1964. All live mussels and all unionid shells were identified to species and recorded. For live unionids, we measured the length (longest anterior-to-posterior axis), height (dorsal-ventral axis) and width (left-right axis) using digital callipers. We additionally measured the length of each shell annulus (along its longest anterior-to-posterior axis). These are distinctive dark bands on the shell formed during periods of temporary growth cessation and have been confirmed to be annual (Rypel et al., 2008), including for populations in the Thames (Negus, 1966). They can therefore be used as a reliable measure of a mussel’s yearly growth (Aldridge, 1999). We dissected a subset of 50 mussels, distributed across species, sampling depths and sizes, to measure shell wet mass and total wet mass separately in order to calculate an estimate of biomass production, following the method reported by Negus. To limit the extent of destructive sampling, we regressed wet mass on length and used this to interpolate total and shell wet mass for the remaining individuals. We report these equations for future reference in Table S1. Data for mussel populations in 1964 were obtained from Negus (Negus, 1966) and extracted from graphs using the software DataThief III (Tummers, 2006). Data used for comparison were those reported from ‘1964, unheated’ surveys. Additional surveys reported from 1963 and from heated effluents near the now-closed Earley Power Station offered less complete and less comparable data and were excluded. Water quality monitoring data were obtained from the Environment Agency for the River Thames at Caversham Weir monitoring point (sampling point ID: TH-PTHR0080), approximately 2km upstream of our sampling location. Data from 2000-present are publicly available (Environment Agency, 2021) and data for 1972 – 1999 were obtained via a Freedom of Information request. This study did not require ethical approval and no licences were required for the collection of mussels.