Selectivity of invasive species suppression efforts influences control efficacy

Highly fecund invaders and size-selective suppression efforts often limit the effectiveness of invasive species control programs, as compensatory processes can allow suppressed populations to recover. While population models have long explored how demographic characteristics impact management feasibility, there is a growing need to evaluate how the selectivity of suppression efforts might impact the long-term feasibility of control. We use a simulation framework that integrates age-based selectivity to evaluate the effect of increasing the range of ages selected for during harvest-based invasive species control. We applied this approach to common carp in Utah Lake, the location of one of the world’s largest freshwater vertebrate species control programs, to assess how selectivity impacts the level of control effort necessary to achieve management targets. Model simulations suggest that increasing the range of ages effectively targeted by removal gears has the potential to significantly reduce the amount of effort required to reach control targets. We found that increasing selectivity on younger, but mature, age classes allowed the control program to maintain the carp population below the 75% biomass reduction target with only 2.5 times the maximum historic effort level, while further increasing juvenile selectivity conferred minimal benefit. Furthermore, we evaluated historic levels of harvest effort against both previously set management targets and theoretical sustainable harvest targets (MSY). The historic level of suppression effort was less than that required to produce MSY regardless of the selectivity scenario explored, suggesting the control program would be harvesting at a sustainable rate even if it increases the range of ages effectively targeted by removal gears. Controlling highly fecund invasive species becomes much more feasible if managers can identify an approach that targets all adult age classes. Explicitly considering sustainable harvest metrics provides a framework for evaluating a harvest control program’s ability to overcome density-dependent processes and achieve management objectives.

Methods
Site description

Utah Lake is a large (~380 km2), shallow (average depth 3.2 m) lake located in Utah County, Utah, USA. In the early 2000s, non-native carp accounted for over 90% of the lake’s fish biomass (SWCA, 2002) and have contributed to reduced water quality, altered food web dynamics (King et al., 2024), and pose a threat to the endemic and federally threatened June sucker (Chasmistes liorus; USFWS, 2021). Since preliminary reports deemed complete carp eradication infeasible due to the lake’s size and connectivity, managers began carp removal efforts in 2009 with a target to reduce carp biomass by 75% (SWCA, 2002; 2005; Walsworth et al., 2020). Initial reports suggested the harvest program could suppress the population’s reproductive potential, achieve the control target in five years, and decrease harvest once the program achieved the target (SWCA, 2005; 2006).

Data collection

Since 2012, standardized annual surveys using the same commercial gear applied during removal efforts, a 184 m long, 3 m deep commercial beach seine with 1½-inch square mesh (hereafter “large mesh” seine) have been conducted by Utah State University, the Utah Division of Wildlife Resources, and Loy Fisheries. Beginning in 2020, a subset of hauls was conducted using a “small mesh” seine with ¾-inch mesh. While commercial fishers typically conducted hauls in targeted locations, each year researcher-led crews collected between 27 and 43 (depending on site accessibility) standardized seine samples from a consistent set of spatially stratified sites around the lake perimeter (see Landom et al., 2014). Field data collection was conducted under the auspices of Utah State University IACUC 10229 and Utah Division of Wildlife Resources collection permit 4COLL10240.

For each seine sample in 2020 and 2021, we recorded the total number of carp captured and randomly selected a subsample (up to 30 individuals) for total length and weight measurements. The age composition of carp caught was calculated with a probabilistic approach based on fitting a Von Bertalanffy growth curve to length-at-age data (ages estimated from the dorsal spines of a subset of carp) and length-frequency data (see Walsworth et al., 2020; Walsworth, Wallace, et al., 2023 for a full description of methods). The data presented in UTL_CarpSeine_DensityAge_byHaul is the resulting density-by-age for each seine haul from the 2020 and 2021 surveys, respectively. 

The UTLCarp_Natage_Medians data are the median model outputs of abundance from the carp population model developed for Utah Lake. We used estimates of carp abundance-at-age from a statistical catch-at-age model which used standardized seine sampling catches, commercial fisheries effort, commercial removal data, and lake level measurements (see Walsworth et al., 2020 for a full model description, Walsworth & Landom, 2021 for an updated catchability component, and Walsworth, Wallace, et al., 2023 for updated abundance estimates). The model follows a standard age-structured framework (Caswell, 2001) with the addition of a Ricker stock-recruitment model (Ricker, 1954) and incorporates parameters for fishing effort, age-based gear selectivity, and lake-level effects on both catchability and recruitment.