Genomic signatures of bottleneck and founder effects in dingoes

Dingoes arrived in Australia during the mid-Holocene and are the top-order terrestrial predator on the continent. Although dingoes subsequently spread across the continent, the initial founding population(s) could have been small. We investigated this hypothesis by sequencing the whole genomes of three dingoes and also obtaining the genome data from nine additional dingoes and 56 canines, including wolves, village dogs and breed dogs, and examined the signatures of bottlenecks and founder effects. We found that the nucleotide diversity of dingoes was low, 36% less than highly inbred breed dogs and 3.3 times lower than wolves. The number of runs of homozygosity (RoH) segments in dingoes was 1.6-4.7 times higher than in other canines. While examining deleterious mutational load, we observed that dingoes carried elevated ratios of nonsynonymous-to-synonymous diversities, significantly higher numbers of homozygous deleterious Single Nucleotide Variants (SNVs), and increased numbers of loss of function SNVs, compared to breed dogs, village dogs, and wolves. Our findings can be explained by bottlenecks and founder effects during the establishment of dingoes in mainland Australia. These findings highlight the need for conservation-based management of dingoes and the need for wildlife managers to be cognisant of these findings when considering the use of lethal control measures across the landscape.

Diet of the Dingo in Subtropical Australian Forests: Are Small, Threatened Macropods at Risk?

Carnivores fulfil important ecological roles in natural systems yet can also jeopardise the persistence of threatened species. Understanding their diet is, therefore, essential for managing populations of carnivores, as well as those of their prey. This study was designed to better understand the diet of an Australian apex predator, the dingo, and determine whether it poses a threat to at-risk small macropods in two floristically different yet geographically close reserves in subtropical Australia. Based on an analysis of 512 scats, dingo diets comprised 34 different prey taxa, of which 50% were common between reserves. Our findings add support to the paradigm that dingoes are opportunistic and generalist predators that prey primarily on abundant mammalian fauna. Their diets in the Border Ranges were dominated by possum species (frequency of occurrence (FOC) = 92.5%), while their diets in Richmond Range were characterised by a high prevalence of pademelon species (FOC = 46.9%). Medium-sized mammals were the most important dietary items in both reserves and across all seasons. The dietary frequency of medium-sized mammals was generally related to their availability (indexed by camera trapping); however, the avoidance of some species with high availability indicates that prey accessibility may also be important in dictating their dietary choices. Other prey categories were supplementary to diets and varied in importance according to seasonal changes in their availability. The diets included two threatened macropods, the red-legged pademelon and black-striped wallaby. Our availability estimates, together with earlier dietary studies spanning 30 years, suggest that the red-legged pademelon is resilient to the observed predation. The black-striped wallaby occurred in only two dingo scats collected from Richmond Range and was not detected by cameras so the threat to this species could not be determined. Two locally abundant but highly threatened species (the koala and long-nosed potoroo) were not detected in the dingoes' diets, suggesting dingoes do not at present pose a threat to these populations. Our study highlights the importance of site-based assessments, population monitoring and including data on prey availability in dietary investigations.

Quantification of flight times of aerial treatments targeting invasive species: the interplay of helicopter or drone with bait-delivery systems, flight speed and bait form

BACKGROUND

Aerial treatments for invasive species management are now common, but we are unaware of any work published in the scientific literature quantifying how the interplay of numerous factors affects flight time and therefore operational costs. Here, we analyse aerial treatment data collected from two ant-eradication programmes, quantifying how the relationships between flight time and area are influenced by numerous aircraft/delivery system/bait/flight speed combinations.

RESULTS

For bait dispersal by helicopters, and when swath widths are equivalent, side-mounted Isolair was significantly more efficient than the simultaneous use of two underslung buckets, and use of two buckets was slightly but not significantly more efficient than one bucket. In this scenario, delivery by Isolair was, on average, 39.8% and 31.5% more efficient than the use of one or two buckets, respectively. However, when the swath width used with Isolair was halved to 10 m and flight speed was increased slightly, flight time was significantly greater compared to the other configurations. For bait dispersed by drone, flights conducted using an upgraded Flight Management System (FMS) and greater flight speed but smaller swath width were significantly more efficient than flights using the older FMS and lower flight speed. Over 10 and 50 ha the helicopter was 2.87 and 4.82 times more time efficient than the drone.

CONCLUSION

We encourage practitioners to publish data from their aerial treatments, and to try new methods, to accelerate improvements in efficiency and reduce the costs of aerial treatments. © 2023 The Authors. Pest Management Science published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry.

Are canid pest ejectors an effective control tool for wild dogs in an arid rangeland environment?

Abstract Context Wild dogs are a significant pest species of livestock production and native wildlife in Australia. A suite of control tools is used to mitigate predation impacts. Baiting with sodium fluoroacetate is the most commonly used control tool in Australia; however, its effectiveness can be reduced by interference by non-target species, and in some contexts by microbial degradation of the toxin. Canid pest ejectors (CPEs) are a mechanical device with an attractant ‘lure head’ designed to eject a lethal toxin into the mouth of canids pulling on the lure head. A range of lure heads can be used to attract canids to pull, and trigger CPEs. Aims We aimed to determine whether uptake of CPEs by wild dogs in an arid rangeland environment could cause a decline in a wild dog population. We also aimed to determine whether there are particular lure heads that increase the rate of CPEs being triggered by wild dogs. Methods We deployed one hundred CPEs over four sessions of control across three properties in the southern rangelands of Western Australia from 2018 to 2020. Each session consisted of 2 months of CPE deployment with two different lure heads, totalling eight lure head types over the entire study. All CPEs were monitored using camera traps. Key results Wild dog density varied over the study period. In all four control sessions, a decrease in wild dog density was recorded (–46%, –5%, –13%, –38%). Wild dog activity events on camera and their interest in CPEs differed between sessions and lures (i.e. higher with scent-based lures). Non-target species did not interfere with CPEs significantly, despite a higher number of activity events by non-target species than wild dogs. Conclusions CPEs caused a reduction of 5–46% of wild dog density when deployed in the southern rangelands of Western Australia. Non-target interference was minimal when using CPEs for wild dog control. Implications Use of scent-based lures on felt lure heads is recommended for successful use of CPEs for wild dog control in arid rangeland environments. Future on-ground wild dog control should include CPEs as a complementary tool for the reduction of wild dog density.

The Dingo Barrier Fence: Presenting the case to decommission the world's longest environmental barrier in the United Nations Decade on Ecosystem Restoration 2021-2030

The longest environmental barrier in the world is Australia's 5614 km Dingo Barrier Fence. The structure was completed in the 1950s, designed to facilitate the eradication of the country's apex predator and cultural keystone species the dingo (Canis dingo) from sheep (Ovis aries) grazing areas to the south-east of the continent. The fence and its support systems now present an immense obstacle to ecological restoration in Australia's arid zone, preventing traditional management practices, and are hazardous to all terrestrial wildlife in the immediate vicinity. The barrier presents a worst-case scenario for animal-generated seed dispersal patterns over the wider region and limits genetic transfer. Plummeting biodiversity inside the fence line and increasing pressures of climate change have left this region highly vulnerable to ecological collapse. Concurrently, sheep numbers have contracted over 75% in the arid zone since 1991, due to market forces and climate change, while demand for ethically produced goods such as predator-friendly meat production and organic produce is increasing. Decommissioning the Dingo Barrier Fence, moving the stock protection zone south and diversifying land use would not impact significantly on the current livestock production. It offers a sound economic alternative for the region, with the potential for regeneration of 82 million hectares of land, a scale encouraged for inclusion in the global initiative the United Nations Decade for Ecosystem Restoration (2021-2030). This would restore connectivity across the region, including vital access to the waters of the Murray Darling Basin. This would provide mitigation for the effects of climate change, new markets in organic and sustainable industries, and support ecological and cultural renewal.

Terrestrial mesopredators did not increase after top-predator removal in a large-scale experimental test of mesopredator release theory

Removal or loss of top-predators has been predicted to cause cascading negative effects for ecosystems, including mesopredator release. However, reliable evidence for these processes in terrestrial systems has been mixed and equivocal due, in large part, to the systemic and continued use of low-inference study designs to investigate this issue. Even previous large-scale manipulative experiments of strong inferential value have been limited by experimental design features (i.e. failure to prevent migration between treatments) that constrain possible inferences about the presence or absence of mesopredator release effects. Here, we build on these previous strong-inference experiments and report the outcomes of additional large-scale manipulative experiments to eradicate Australian dingoes from two fenced areas where dingo migration was restricted and where theory would predict an increase in extant European red foxes, feral cats and goannas. We demonstrate the removal and suppression of dingoes to undetectable levels over 4–5 years with no corresponding increases in mesopredator relative abundances, which remained low and stable throughout the experiment at both sites. We further demonstrate widespread absence of negative relationships between predators, indicating that the mechanism underpinning predicted mesopredator releases was not present. Our results are consistent with all previous large-scale manipulative experiments and long-term mensurative studies which collectively demonstrate that (1) dingoes do not suppress red foxes, feral cats or goannas at the population level, (2) repeated, temporary suppression of dingoes in open systems does not create mesopredator release effects, and (3) removal and sustained suppression of dingoes to undetectable levels in closed systems does not create mesopredator release effects either. Our experiments add to similar reports from North America, Asia, Europe and southern Africa which indicate that not only is there a widespread absence of reliable evidence for these processes, but there is also a large and continually growing body of experimental evidence of absence for these processes in many terrestrial systems. We conclude that although sympatric predators may interact negatively with each other on smaller spatiotemporal scales, that these negative interactions do not always scale-up to the population level, nor are they always strong enough to create mesopredator suppression or release effects.

Measuring, evaluating and improving the effectiveness of invasive predator control programs: Feral cat baiting as a case study

Reducing the impacts of invasive predators is a key objective for conservation managers, livestock producers and human health agencies globally. The efficacy of invasive predator control programs, however, is highly variable. To improve control efficacy, managers require a fundamental understanding of the factors that contribute to the success or failure of a control program. Using a predator baiting program as a case study, we measured the efficacy of baiting as a control tool to significantly reduce feral cat (Felis catus) populations. We used camera traps and cat-borne GPS collars to monitor changes in feral cat populations at a baited site and an unbaited site, using a Before-After, Control-Impact (BACI) design. We also identified five key elements required for a successful baiting program (bait encounter rate, availability, attractiveness, palatability and lethality) and simultaneously measured these to identify areas for potential improvement. Baiting was ineffective at reducing feral cat populations; collared cat mortality was only 11% (1/9), with camera traps revealing negligible reductions in the number of cat detection events (9%), naïve occupancy (15%), and no significant change in the relative abundance of feral cats (F_1,54 = 0.8641, P = 0.357). Several factors contributed to the poor control efficacy. Bait encounter rates were low, with cats active along tracks (where baits were laid) < 4% of the time. Cats encountered only 14% (7/50) of monitored baits, but none were eaten. Initially, baits appeared attractive to cats; however meat ants and desiccation rapidly decreased bait palatability. Bait availability to cats declined rapidly, with 36% of monitored baits (18/50) removed by non-target species within the first 48 h. The mortality of one collared cat and chemical assays confirmed that, on average, each bait contained sufficient 1080 to kill a large (>5 kg) feral cat. Our findings suggest that altering bait deployment patterns, increasing bait densities and improving bait palatability could potentially improve the efficacy of baiting programs to reduce feral cat populations. Our study provides a framework to measure and evaluate the key elements that contribute to efficacy of pest control programs, and to identify opportunities for improving outcomes of future control programs.

On the right track: placement of camera traps on roads improves detection of predators and shows non-target impacts of feral cat baiting

Abstract ContextTo understand the ecological consequences of predator management, reliable and accurate methods are needed to survey and detect predators and the species with which they interact. Recently, poison baits have been developed specifically for lethal and broad-scale control of feral cats in Australia. However, the potential non-target effects of these baits on other predators, including native apex predators (dingoes), and, in turn, cascading effects on lower trophic levels (large herbivores), are poorly understood. AimsWe examined the effect that variation in camera trapping-survey design has on detecting dingoes, feral cats and macropodids, and how different habitat types affect species occurrences. We then examined how a feral cat poison baiting event influences the occupancy of these sympatric species. MethodsWe deployed 80 remotely triggered camera traps over the 2410-km2 Matuwa Indigenous Protected Area, in the semiarid rangelands of Western Australia, and used single-season site-occupancy models to calculate detection probabilities and occupancy for our target species before and after baiting. Key resultsCameras placed on roads were ~60 times more likely to detect dingoes and feral cats than were off-road cameras, whereas audio lures designed to attract feral cats had only a slight positive effect on detection for all target species. Habitat was a significant factor affecting the occupancy of dingoes and macropodids, but not feral cats, with both species being positively associated with open woodlands. Poison baiting to control feral cats did not significantly reduce their occupancy but did so for dingoes, whereas macropodid occupancy increased following baiting and reduced dingo occupancy. ConclusionsCamera traps on roads greatly increase the detection probabilities for predators, whereas audio lures appear to add little or no value to increasing detection for any of the species we targeted. Poison baiting of an invasive mesopredator appeared to negatively affect a non-target, native apex predator, and, in turn, may have resulted in increased activity of large herbivores. ImplicationsManagement and monitoring of predators must pay careful attention to survey design, and lethal control of invasive mesopredators should be approached cautiously so as to avoid potential unintended negative ecological consequences (apex-predator suppression and herbivore release).