Marshall Jonathan (Orcid ID: 0000-0002-7177-4543) Negus Peter (Orcid ID: 0000-0003-2680-2573) Ecological impacts of invasive carp in Australian dryland rivers Jonathan C. Marshall1,2 Joanna J. Blessing1 Sara E. Clifford1 Kate M. Hodges1,3 Peter M. Negus1,4 Alisha L. Steward1,2 1Department of Environment and Science, Queensland Government, Brisbane, Australia 2Australian Rivers Institute, Griffith University, Nathan, Queensland, Australia 3Institute for Applied Ecology, University of Canberra, Canberra, ACT, Australia 4School of Earth and Environmental Sciences, University of Queensland, Brisbane, Queensland, Australia Correspondence: Jonathan C. Marshall, Department of Environment and Science, Queensland Government, GPO Box 2454, Brisbane, Queensland 4001, Australia Abstract 1. Invasive carp are widely reported to harm ecosystems. In Australia, carp are a serious pest, and consequently investigations of biocontrol options are under way. 2. Best practice biocontrol requires cost/risk:benefit evaluation. To assist this, the impacts of carp on aquatic ecosystems have been summarized. 3. To aid the evaluation of benefits, general predictions were tested by comparing dryland river ecosystems with and without carp, and ecosystem responses to a gradient in local carp density. This is the author manuscript accepted for publication and has undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as doi: 10.1002/aqc.3206 This article is protected by copyright. All rights reserved. 4. Expectations were that in the presence of carp, and with increasing density, there would be increasing turbidity, decreasing densities of macrophytes and macroinvertebrates and associated changes in assemblage composition, resulting in decreasing native fish density. 5. Not all expected responses were found, indicating that the general understanding of carp impact requires modification for dryland rivers. Notably, carp did not increase turbidity or reduce macroinvertebrate density or composition, probably because of key attributes of dryland rivers. In contrast, there were large impacts to native fish biomass, not from the mechanisms expected, but from food resource monopolization by carp. Macrophyte occurrence was reduced, but macrophytes are naturally rare in these rivers. It is likely that the extirpation of an endangered river snail resulted from carp predation. 6. Impacts on native fish may be reversible by carp control, but reversal of impacts on the snail may require carp elimination and snail reintroduction. Modelling is necessary to predict the probability of beneficial versus undesirable outcomes from carp control, and complementary measures to control other stressors may be needed. 7. Benefits of carp control on dryland river ecosystems are fewer than generally predicted. This reinforces the point that ecological understanding cannot always be transferred between diverse settings, and highlights the need to understand system characteristics relevant to causal impact pathways when applying generic carp impact models to specific settings. This has global relevance to future carp control efforts. Key words alien fish, biocontrol, macroinvertebrates, macrophytes, molluscivory, Murray-Darling Basin, National Carp Control Plan, native fish biomass, Notopala sublineata, turbidity 1. Introduction From their origin in Eurasia, carp (Cyprinus carpio) have been introduced to waterways in many parts of the world. This has often been deliberate, dating back at least to the era of the Roman Empire, as they are an important food and aquaculture species and are highly valued for recreational fishing (Vilizzi, 2012). Carp have also been widely reported to cause ecological harm where they have been introduced (Vilizzi, Tarkan, & Copp, 2015). This has generated tension between conservation 2 This article is protected by copyright. All rights reserved. interests, who lobby for carp control for ecological benefit (Weber & Brown, 2009), and fishing interests, who advocate maintaining or boosting carp stocks (Carpworld Magazine, 2019). Contrasting attitudes towards carp in the United Kingdom (UK) exemplifies such tensions. Carp were introduced to the UK several hundred years ago, where they are recognized to cause environmental problems such as elevated nutrient concentrations, algal biomass and turbidity, and reduced abundances of native fish (Jackson, Quist, Downing, & Larscheid, 2010). These are impacts that should be managed under the requirements of the European Water Framework Directive (Council of the European Communities, 2000) in order to achieve ‘good ecological status’, yet many UK lakes are still heavily stocked with carp and managed for recreational carp fishing (Britton, Gozlan, & Copp, 2011; Hewlett, Snow, & Britton, 2009). Likewise, conservation efforts to increase populations of Eurasian otter (Lutra lutra) in the UK have created conflict, with carp fisherman believing otters feed on stocked carp and calling for culls of otters in important carp fishing regions (Almeida et al., 2012). In Australia, carp are widespread and abundant in the waterways, impoundments, wetlands and lakes of the Murray-Darling Basin and in some other river systems (Koehn, 2004). First introduced in the mid-1800s, they only became invasive following the escape of the Boolarra carp strain from captivity in the 1960s. They have since spread widely following extensive river flooding in exceptionally wet years during the 1970s, to reach the current major extent of their distribution by the latter few decades of the twentieth century. Today, carp dominate many aquatic systems in the Murray-Darling Basin, constituting up to nearly 80% of the total fish biomass (Koehn, 2004). Although, as elsewhere, carp were introduced to Australian waters as a potential new human food source (Koehn, Brumley, & Gehrke, 2000), they are not generally accepted here as a desirable food or as a worthy angling species. Rather, they are broadly derided as an invasive pest requiring control or elimination (Koehn et al., 2000). Native fishes in the Murray-Darling Basin have been subjected to numerous stressors since European settlement, including river regulation and water diversion, longitudinal and lateral fragmentation (Baumgartner, Zampatti, Jones, Stuart, & Mallen-Cooper, 2014), habitat degradation, and the introduction of alien species (Wilson, 2005). Many native species have undergone reductions in range and population size throughout the Basin, with the Yarra pygmy perch (Nannoperca obscura) recently identified as possibly the first fish extinction for the catchment (ABC News, 2019). Basin- 3 This article is protected by copyright. All rights reserved. scale restoration measures have been implemented or are planned, among other things, to restore native fish populations. These measures include environmental flow restoration by water recovery (Murray Darling Basin Authority, 2012) and measures to provide enhanced fish passage past dams and weirs (Barrett & Mallen-Cooper, 2006). Carp are considered to be a major remaining threat to native fish in the region (Koehn et al., 2000). This perception led the Australian Government to develop a National Carp Control Plan (NCCP) to reduce carp abundance in order to help restore the health of Australian waterways and aquatic biodiversity, with a focus on native fish recovery (NCCP, 2018). The principal approach currently under consideration by the NCCP is biocontrol using cyprinid herpesvirus 3 (CyHV-3). Although various authors have raised doubts and uncertainties concerning the likely efficacy, safety and risks of unintended perverse outcomes from potential biocontrol of Australian carp with CyHV-3 (Becker, Ward, & Hick, 2018; Kopf et al., 2017; Lighten & Van Oosterhout, 2017; Marshall, Davison, Kopf, Boutier, & Vanderplasschen, 2018; Paton & McGinness, 2018; Thresher, Allman, & Stremick- Thompson, 2018), it remains important to this planning process to quantify the ecological benefits that successful carp control could achieve. Such information provides the denominator in a cost/risk: benefit evaluation at the core of best-practice biocontrol evaluation (Kopf et al., 2017) and is thus fundamental to the NCCP and its eventual recommendations to government. The benefits of successful carp biocontrol are considered here in two stages. First, aspects of the ecological harm carp cause at present are quantified. Second, the reversibility of this harm is considered if carp abundances were to be successfully reduced. This is addressed in the setting of the rivers of the northern part of Australia’s Murray-Darling Basin, that are often described as dryland rivers because much of their length flows through arid and semi-arid landscapes and they experience large-scale nett evaporative water losses (Balcombe et al., 2010). Such rivers offer several practical advantages for this study. They are characterized by highly intermittent flow regimes, with short duration flow pulses interspersed with long periods of no discharge, during which obligate aquatic organisms are restricted to remnant, persistent, in-channel waterholes. Isolation of the aquatic system into waterholes as discrete habitat patches during dry phases means that local impacts of carp are likely to predominate over regional impacts (Balcombe et al., 2006), making this an ideal environment for testing ecological responses to variable local