Effects of watercress farming on fish populations

Abstract

Watercress (Nasturtium officianale) is a salad crop grown commercially in watercress beds irrigated
with water from chalk aquifers. The effluent from irrigation, and in some instances, salad washing
processes, are discharged into adjacent chalk streams. There is concern that macroinvertebrate
assemblages downstream of discharges reflect organic pollution, which has been attributed to
siltation and more recently the release of phenethyl isothiocyanate (PEITC). An antiherbivore
metabolite produced by watercress in response to physical damage, PEITC has the potential to be
released into chalk streams via two pathways: (i) via irrigation water emanating from watercress
beds following disturbances such as harvesting and other crop-damaging activities; (ii) via the
discharge of salad wash effluent from the rinsing of watercress, a process carried out on small
number farms. While PEITC toxicity to macroinvertebrates is well-studied, the impact that PEITC
may be having on fish populations has received little attention. Watercress farms discharge into
chalk stream headwaters where fish embryos are incubated, so there is a potential for this sensitive
early life stage to be directly exposed to PEITC.

To determine the impact watercress farm discharges are having on habitat, macroinvertebrate and
fish, three watercress farms each were surveyed biannually. Each farm varied in its utilisation of
salad washing, with the aim of the study to investigate whether any changes in physicochemistry,
habitat, macroinvertebrate prey availability from watercress bed irrigation and salad wash effluent
rendered sites suboptimal to support fish populations. Organic pollution stress was assessed
through macroinvertebrate assemblages using the Walley Hawkes Paisley Trigg (WHPT) biotic
index, which provides a score based on the relative abundances of pollution sensitive and pollution
tolerant taxa. Sites receiving salad wash effluent had significantly lower WHPT scores than expected
and higher total macroinvertebrate abundance, while sites receiving watercress bed irrigation
water only scored higher than predicted. Fish species other than brown trout (Salmo trutta) were
found at higher densities below discharges. However, S. trutta found at lower densities were in
better condition, possibly due to decreased intraspecific competition and greater
macroinvertebrate prey abundances. In contrast to sites that received just watercress bed irrigation
discharge, sites below salad wash effluent had densities of young-of-year S. trutta that were lower
than expected, suggesting that PEITC release from salad washing may reduce embryo survival.

To assess embryotoxicity of PEITC, a series of laboratory trials exposed S. trutta, common carp
(Cyprinus carpio) and zebrafish (Danio rerio) embryos to 0.01, 0.1 and 1µg/L PEITC during embryonic development. In all three species, exposure to 1µg/L resulted in complete mortality between 1-3 dosing days, while embryos exposed to 0.1 µg/L PEITC suffered higher mortality rates, significantly delayed hatching, higher incidence of spinal deformations and significantly altered behaviour compared to controls. These levels of exposure were orders of magnitude below estimates of PEITC discharges from salad washing, suggesting that salad wash effluent may have been a factor in the low densities of young-of-year S. trutta downstream of salad wash effluent discharge.

Date of Award	Sept 2020
Original language	English
Awarding Institution	University of Brighton
Supervisor	Neil Crooks (Supervisor), Angelo Pernetta (Supervisor) & Christopher Joyce (Supervisor)