A review of nonlethal and lethal control tools for managing the damage of invasive birds to human assets and economic activities

Invasive birds cause damage to economies, natural resources, and human safety across the globe. In the United States, rock doves ( Columba livia ), Eurasian collared doves ( Streptopelia decaocto ), rose-ringed parakeets ( Psittacula krameri ), monk parakeets ( Myiopsitta monachus ), common mynas ( Acridotheres tristis ), European starlings ( Sturnus vulgaris ), and house sparrows ( Passer domesticus ) are among the invasive and often harmful small-bodied birds inhabiting periurban habitats. The destructive nature of these species warrants a review of methods to reduce or eradicate populations along with methods to reduce damage when population eradication cannot be achieved. We reviewed damage management literature from these species’ native and introduced ranges. Additionally, we used the behavior and ecology of these species to inform tool recommendations and potential efficacy under various damage scenarios, while being sensitive to cultural preferences and location of implementation (residential, commercial, and agricultural). Although this review focuses on invasive birds in the United States, it is applicable to other pest species across the globe. Our review highlights areas where research is needed to validate promising damage management methods (lethal control, fertility control, habitat modification, exclusionary methods, frightening devices, and chemical repellents). Where birds are invasive, integrated pest management techniques should focus on eradication or population reduction (toxicants, shooting, and trapping) to keep populations at levels where nonlethal tools can reduce damage. We acknowledge the efficacy of an eradication campaign depends on biological, environmental, and economic factors, along with social license for lethal removal. We recommend integrated pest management strategies including lethal and nonlethal tools specific to the damage problem. Sustained efforts to reduce invasive populations should be used along with integrated deterrent strategies for short-term damage relief.


Introduction
Invasive species are introduced, nonnative species that spread rapidly and are hazardous to native ecosystems, agricultural systems, and human health and safety (Iannone et al. 2021;Mack et al. 2000;Paini et al. 2016). Invasive birds have unique impacts (Martin-Albarracin et al. 2015) with three species included on the list of the 100 of the world's worst invasive species (Lowe et al. 2004). The success of avian invaders is due to a generalist diet, high fecundity, tolerance of humans, dispersal capabilities, and prevalence in the pet trade (O'Connor 1986;Sol et al. 2017;Evans et al. 2018;Gippet and Bertelsmeier 2021). The rock dove (Columba livia), Eurasian collared dove (Streptopelia decaocto), rose-ringed parakeet (Psittacula krameri), monk parakeet (Myiopsitta monachus), common myna (Acridotheres tristis), European starling (Sturnus vulgaris), and house sparrow (Passer domesticus) are examples of such avian invaders (Invasive Species Compendium 2012; Dyer et al. 2017). Although not a comprehensive list, it encompasses smaller-bodied birds that inhabit urban and periurban habitats, thus species that simultaneously impact agricultural and urban economies along with natural resources ). Most of these species are widespread and ubiquitous across the globe, while also encompassing the suite of harmful nonnative birds in the United States (Downs and Hart 2020;Pruett-Jones 2021). Most of these species can form large flocks while foraging and roosting, reaching numbers that increase negative interactions with the public (Shirley and Kark 2009).
Birds that roost in urban settings and forage in agricultural environments are of particular concern (Coombs et al. 1981;Yap et al. 2002;Hetmański et al. 2011;Avery and Lockwood 2017;Avery and Shiels 2018;Linz et al. 2018). Invasive birds are reservoirs and vectors of human, wildlife, and livestock diseases (Pimentel et al. 2000) and can impact native wildlife though disease transmission, resource competition, aggression, and predation (Kumschick and Nentwig 2010;Baker et al. 2014;Martin-Albarracin et al. 2015;Hernández-Brito et al. 2018). Birds flocking near airports are a hazard to human safety through airplane strikes, with invasive birds ranking high on strike risk (rock dove = 4 th , European starling = 6 th , common myna = 43 rd , house sparrow = 57 th ; DeVault et al. 2018). Urban nighttime roosts result in noise complaints and unsanitary conditions, which despoil buildings and increase the risk of disease transmission (Spennemann et al. 2017;Linz et al. 2018;Shiels and Kalodimos 2019;Mori et al. 2020). Risk of foodborne illness also increases when flocks interface with livestock facilities or contact food for human consumption (Carlson et al. 2011a, b;Mori et al. 2018;Chandler et al. 2020). Birds can become pests in agroecosystems due to consumption of crops and feeding behaviors that increase the severity of crop damage, such as dropping food and discarding partially eaten food (Toor and Ramzan 1974;Ali et al. 1981;Tillman et al. 2000;Sebastián-González et al. 2019). Some birds completely consume fruits, while others cause partial damage which increases spoilage and reduces marketability (Carlson et al. 2013). Invasive birds also consume stored grains and unripe fruit, extending the temporal damage window (Chakravarthy 2004;Bhargava and Kumawat 2010;Woods et al. 2022).
The destructive nature of invasive birds warrants methods for damage reduction in addition to population suppression . Olsen (1998) emphasizes a focus on damage reduction, not just reducing pest numbers, and outlines steps of successful management plans, including 1) setting clear objectives (e.g., reducing damage to an acceptable level), 2) identifying management options, 3) selecting control tools, and 4) establishing criteria for monitoring efficacy. Previous work has focused on eradication of specific species of invasive birds (Pruett-Jones et al. 2007;Johnson and Donaldson-Fortier 2009;Bednarczuk et al. 2010;Phillips et al. 2012;Bunbury et al. 2019;Feare et al. 2021a) or the global trends and impacts of invasive birds (Menchetti and Mori 2014;Pitt et al. 2017;Shivambu et al. 2020;Downs and Hart 2020), but few have included a comprehensive evaluation of nonlethal damage management methods and tools for situations where eradication is not possible or delayed (Braysher 2017;Conover 2002;Tracey et al. 2006;Linz et al. 2018). Thus, our objective was to complete a comprehensive review of nonlethal techniques to reduce damage while also highlighting management methods for lethally controlling populations of invasive birds commonly found to cause damage in the United States. Although important, we do not focus on preventing the introduction of nonnative species through international trade nor implications of such regulations. We reference avian behavior and ecology to inform tool recommendations and efficacy in various damage scenarios (i.e., urban, periurban, and rural). If methods lack field implementation or testing on invasive species, we reference studies on other birds to gauge potential efficacy. We identify candidate tools for evaluation and provide guidelines for actions that can be taken to protect resources. We also briefly address the importance of human dimensions in strategic plans to reduce disagreements among people concerning strategies and methods to decrease bird damage.

Population Control
Criteria for attempting eradication include: 1) removal rate exceeds replacement, 2) immigration can be prevented, 3) all reproductive animals are accessible for removal, 4) detection is possible at low densities, 5) a favorable cost:benefit ratio, and 6) a suitable sociopolitical environment (Bomford and O'Brien 1995;Olsen 1998). Frequently, all these conditions cannot be met and therefore species eradication is often impractical. However, population suppression may be feasible for reducing damage if well-funded, sustained, and broad-scale control plans are established (Feare et al. 2021a) within the bounds of cost-benefit ratios (Bomford and O'Brien 1995). Additionally, cooperative approaches coordinated at broad landscape scales are better than individual attempts (Olsen 1998). Innovative studies have focused on identifying the subset of birds responsible for the damage to cull (Khatri-Chhetri et al. 2020), identifying the best habitat or locations to perform lethal control (D'Amico et al. 2013;Klug and Homan 2020), or population connectivity to inform if local control will be effective (Woolnough et al. 2006;Rollins et al. 2009;Jacob et al. 2015). Rash, poorly planned, and poorly executed culling could hamper an effective shooting campaign (Grarock et al. 2014;Bunbury et al. 2019). Safe, discrete methods to lethally take invasive birds are available while they are foraging, loafing, nesting, roosting, and along flight paths (Conroy and Senar 2009;Avery and Feare 2020). In a lethal campaign, birds may change behavior to avoid risky areas after flock mates have been removed (Invasive Species Compendium 2012; Anderson et al. 2022a). Thus, swift action is needed to remove the most birds prior to behavioral changes (e.g., change in roosting locations) and ongoing monitoring programs are necessary to pinpoint new locations. Established populations of invasive birds are likely impossible to eradicate at the continental scale given population suppression has been shown to be limited for large populations of native species at broad landscape scales (Linz et al. 2015). Furthermore, most successful eradications of invasive species are limited to islands with small populations or scenarios on the mainland where small population are contained (Olsen 1998).
Numerous examples of eradication campaigns exist, which highlight scenarios and methods that bolster success. In the Seychelles, eradication of 545 rose-ringed parakeets on the 157-km 2 island of Mahé cost approximately US$1 million and took five years (Bunbury et al. 2019), lending evidence to the expense of complete eradication (Menchetti et al. 2016). Saavedra and Medina (2020) also successfully eradicated 175 rose-ringed parakeets from La Palma, Canary Islands. Eradication of 1,477 rock doves from the Galápagos took seven years but the cost was relatively less due to the species' association with humans (Phillips et al. 2012). Additional successful eradications include 750-1,000 common mynas removed from the Seychelles with trapping and shooting (Canning 2011;Feare et al. 2017Feare et al. , 2021b and 310 wild turkeys (Meleagris gallopavo) removed from Santa Cruz Island over six years (Morrison et al. 2016). Eradication of 13 house crows (Corvus splendens) from Socotra Island, Yemen, took 15 days of professional marksmanship compared to many years of a bounty program removing > 550 eggs and chicks (Suliman et al. 2010). In an eradication attempt on Robinson Cruise Island, elusive house sparrows remained, and reinvasion is likely due to lack of biosecurity measures (Hagen et al. 2019). Attempts to suppress large, established invasive populations have been unsuccessful or ended prior to complete eradication (e.g., house sparrows from Mauritius; Bednarczuk et al. 2010;Avery and Feare 2020). Populations of mitred conures (Psittacara mitratus) on Maui Island, Hawaii, USA, have been reduced to lows of 20 birds mainly though shooting techniques and extensive public outreach (Radford and Penniman 2014). Sol and Senar (1995) found that removal of urban rock doves in their native range resulted in rapid immigration of individuals from areas where no control was exerted despite limited home ranges. Controlling European starlings over continental scales has proven difficult (Woolnough et al. 2005(Woolnough et al. , 2006Campbell et al. 2016), but genetic evaluation has pinpointed where local control is the most effective (Rollins et al. 2009) or how local control can prevent range expansion (Rollins et al. 2011). Thus, factors including avian population size, geography, movement behavior, habitat and terrain, and association with humans all influence success along with the timeline and cost of the culling campaigns.

Shooting (Shotguns and Rifles)
Firearms are regularly used to dispatch invasive birds in culling campaigns (Millett et al. 2004;Suliman et al. 2010;Canning 2011;Phillips et al. 2012; Table 1, Figure 1, see Supplementary material Table S1, Appendix 1). The proper selection of firearms increases culling efficacy, while being sensitive to public perception. For coordinated shooting campaigns, a CO 2 air rifle can be used when birds are perched at foraging, loafing, nesting, and communal roosting sites. The quiet accuracy of an air rifle can reduce public attention and flushing of conspecific birds (Blanvillain et al. 2020). A pellet caliber of .22 is preferred for lethality, but smaller .177 pellets may be a better safety option for small birds at close range. A silenced conventional .22 rifle along with high quality subsonic ammunition can be as quiet as an air rifle and allow more efficiency at various distances with reduced need for a rangefinder (Per-Arne Åhlén, Swedish University of Agricultural Sciences, personal communication, 2021). The rifle could be accurately sighted for shooting individual birds via a green laser or a red spotlight to not disturb adjacent birds . Alternatively, night vision rifle scopes can be used to reduce alarm by birds and attention of onlookers. While studies on mammals use forward looking infrared (FLIR) to detect the body heat of animals in vegetated areas, Christiansen et al. (2014) found that body size and low contrasts in temperature limit the application of thermal sensors in small endotherms (e.g., birds). Extreme care should be taken in identifying line of sight, target, and backdrop to avoid property damage, injury, and ricochet. A 12-gauge shotgun is ideal for culling birds in flight along regular flight lines or upon arrival at foraging areas but should be limited at loafing and roosting sites to prevent abandonment or lessen behavioral shifts in site use. Alternatives to lead pellets or shot should be prioritized to avoid environmental contamination. Table 1. Current and potential management options for damage situations involving invasive birds. We include citations if the method has been tested in the native or introduced ranges of the following birds considered invasive in the United States: Rock doves (Columba livia; RODO), Eurasian collared doves (Streptopelia decaocto; EUCD), rose-ringed parakeets (Psittacula krameri; RRPA), monk parakeets (Myiopsitta monachus; MOPA), common mynas (Acridotheres tristis; COMY), European starlings (Sturnus vulgaris; EUST), and house sparrows (Passer domesticus; HOSP). We include the number of studies pertaining to each invasive species for a tool; § = study type of field (F), lab (L), or modeling (M); ‡ = response variable for field studies that report reduction in bird abundance (A), reduction in damage (D), or bird behavior (B) including movement or habitat use (references listed in Supplementary Material). All tools and methods vary in efficacy depending on pest species, landscape, and deployment strategy (e.g., reduce habituation by switching, combining, and moving both lethal and non-lethal devices. All methods and tools should be operated in accordance with local, regional, and national regulations, including depredation permits, firearm laws, and pesticide regulations. Lethal removal can occur on foraging grounds including row-crop agriculture, livestock facilities, backyard gardens, urban parks, fruit farms, and natural areas (Shiels et al. 2018). Understanding species-specific feeding activity and breeding behavior in space and time will identify the best season and time of day for deploying control tools (Morrison et al. 2016). To reduce crop damage, an air rifle or sniper rifle may be advantageous when birds are perched (Shafi et al. 1986). Flock size can range from a few to thousands of birds; thus, shotguns are preferred in row-crops or when flocks are first approaching the area to be protected. Removing the first birds to approach a foraging area (i.e., sentinels) may effectively stop the flock if the aim is to reduce damage rather than increase take (e.g., roseringed parakeets; W. Bukoski, USDA-WS, personal observation, 2021; J. Young, Kani Wildlife Control, LLC, personal communication, 2021). Lethal control at foraging sites can be performed year-round with specific areas targeted as preferred foods become available (Shiels et al. 2018).
Many invasive birds have large, stationary nighttime roosts, which are accessible for population reduction. However, roosts are often located in urban areas where culling activities are scrutinized (Butler 2003;Khan 2003;Avery and Shiels 2018). Thus, the timing of culling activities not only depend on animal behavior but human activity. Communal roosting species are active from dawn to dusk, leaving and returning to roosts in a reliable manner (Mabb 1997;Khan 2002;Kotagama and Dunnet 2007;Luna et al. Figure 1. The number of studies (i.e., field, lab, and modeling) using A) lethal methods to control populations at nesting, foraging, loafing, and roosting sites, B) nonlethal methods to control damage at urban nesting, foraging, loafing, and roosting sites, and C) nonlethal methods to control damage at agricultural foraging sites, including those conducted in the native or introduced ranges of the following birds considered invasive in the United States: rock doves (Columba livia; RODO), Eurasian collared doves (Streptopelia decaocto; EUCD), rose-ringed parakeets (Psittacula krameri; RRPA), monk parakeets (Myiopsitta monachus; MOPA), common mynas (Acridotheres tristis; COMY), European starlings (Sturnus vulgaris; EUST), and house sparrows (Passer domesticus; HOSP). Above each column on the left-hand side is the number of studies that were conducted in the field (i.e., not laboratory or modeling studies; if a study used multiple tools it was counted for each tool). The right-hand side is the subset of field studies for each category (D-F) that included damage assessments in the results. 2017). Birds are less likely to be disturbed on dark nights, thus moonless nights may be preferred for culling. Birds may loaf before settling down, providing opportunities to remove high-ranking individuals and breeders as indicated by antagonistic interactions with other birds and habitat selection, such as use of the optimal perch substrate or height by dominant birds.
Population suppression should focus on sexually mature adults, otherwise breeding pairs will replace nonbreeding individuals removed by culling (Grarock et al. 2014). We recommend collecting sex and age data of birds removed to identify preferred locations of females, socially dominant birds, and optimal seasons to locate these individuals for targeted population reduction. Population suppression should occur prior to or during breeding to limit annual recruitment. We recommend locating nesting areas to destroy nests or reproductive adults, but cost-effectiveness of this practice is species-and site-specific and depends on if breeding pairs are colonial or preferred nesting habitat is accessible. Locating areas with abundant nesting habitat through habitat evaluation or observed mating behaviors increases the ability to remove breeders and provides locations for repeated management. Managers tasked with lethal removal should be aware of sex-specific behaviors to target females (e.g., mating behavior, timing of incubation, and male provisioning; Radford and Penniman 2014; Klug et al. 2019). If nests are inaccessible or hard to find, flight lines between foraging areas and nesting colonies may allow for removal of breeders (e.g., mitred conure nesting on cliffs; Radford and Penniman 2014).

Capture Devices (Traps and Nets)
Invasive birds have been trapped using various trap designs (Table 1, Figure 1, see Table S1, Appendix 1). Although labor-intensive, communitybased volunteer programs can increase spatial scale of trapping while reducing cost (Grarock et al. 2014;Linley et al. 2017;Blanvillain et al. 2020). Modified Australian crow traps have been used to capture rose-ringed parakeets, house sparrows, common mynas, and European starlings (Bashir 1979;Montplaisir et al. 2006; Copsey and Parkes 2013; Thiele 2020) but can be less successful in environments with abundant alternative food yearround (Gaudioso et al. 2012;Bunbury et al. 2019). Remotely triggered, spring-loaded traps and Potter walk-in traps can be deployed if regular feeding stations can be established (Phillips et al. 2012;Avery and Lindsay 2016). Placing traps on top of preferred food (e.g., corn at the milky stage) or using live decoy birds within traps may increase capture rates (Conover and Dolbeer 2007;Campbell et al. 2012a;Peck et al. 2014;Saavedra and Medina 2020). Decoy traps were identified as the most effective tools to remove common mynas (Canning 2011), but social cues can also inform conspecifics about the threat of traps (Diquelou and Griffin 2019) or in other cases increase capture rates (Copsey and Parkes 2013). The most effective capture methods are species-and site-specific, including the location, timing, and type of bait (Tidemann 2005;Saavedra 2010;Canning 2011;Linley et al. 2017;Feare et al. 2021a). Seasons with reduced alternative food and high energetic demands (e.g., migration) would be the most productive for trapping (Feare et al. 2017). Long-handled nets have been used for removing monk parakeets from nests (Avery and Lindsay 2016) and roseringed parakeets from low roosting branches (Gaudioso et al. 2012). Dip nets have been used to capture individual starlings (Spencer and De Grazio 1962), whereas floodlight traps capture larger numbers of birds at marsh roosts (Mitchell 1963). Avery and Shiels (2018) suggested elevated mist nets to capture birds upon arrival or departure from roost sites after flight lines are identified. Whereas Tidemann (2010) considered enclosing preferred roosting trees with nets to capture large numbers of mynas, and Strubbe and Matthysen (2011) placed mist nests around established feeders to capture rose-ringed parakeets. In some cases, trapping and hand capture of urban rock doves has been successful, but in other scenarios were ineffective due to trap neophobia (Phillips et al. 2012;Farfán et al. 2019). We classify such live-capture techniques as lethal removal because the only practical recourse for trapped birds is almost always euthanasia.

Avicides (Toxicants and Wetting Agents)
There is a long history of testing avicides in both lab and field conditions for pest species (Table 1, Figure 1, see Table S1, Appendix 1). The primary difficulty with toxicants is establishing a delivery system that avoids negative impacts on nontarget animals while attracting target birds, especially where alternative food is abundant (Linz and Bergman 1996;Avery and Shiels 2018). Although prototype devices for excluding nontarget birds have been tested on parakeets and are operational for rock doves (Tillman 2016;Senar et al. 2021;Anderson 2022b), public sentiment and the inability to lure target birds and deter nontargets to bait limits effectiveness. As a result, exclusion of nontarget animals and reduced environmental impacts are primarily achieved through label use restrictions limiting the type of use sites, timing of bait applications, and requiring the monitoring and cleanup of use sites by certified applicators.
Other lethal pesticides registered for birds in the United States include chemical frightening agents and wetting agents. 4-aminopyridine ; also known as Avitrol®) is a frightening agent and causes erratic flight, distress calls, and death when ingested, which may cause entire flocks to disperse due to an antipredator response. 4-aminopyridine is registered by the USEPA to target European starlings, rock doves, and house sparrows for use in or on structures and other non-crop areas used as feeding, nesting, loafing, and roosting sites (Dolbeer and Linz 2016;USEPA 2022). Wetting agents destroy insulating properties of feathers and leave birds susceptible to hypothermia at temperatures < 41 °F (Byrd et al. 2009), thus are only effective in temperate regions where applications coincide with cold weather. Sodium lauryl sulfate (CAS No. 151-21-3) is classified as a minimum risk pesticide active ingredient by the USEPA and, therefore, its use as a wetting agent for starlings is exempt from federal registration under FIFRA section 25(b) (40 CFR 152.25(f); USDA-APHIS-WS 2012). However, spray applications at roosting sites negatively impact vegetation (Byrd et al. 2009).
Additional toxicants with potential future use in the United States include alpha-chloralose (CAS No. 15879-93-3), which is a sedative that can act as an acute toxicant at higher doses (Nelson 1994). Alpha-chloralose is registered for house mice in the United States but is not yet registered as an avicide. The US Federal Drug Administration previously authorized research on its use for immobilization and live capture of certain bird species under an Investigational New Animal Drug file (INAD 6602;O'Hare et al. 2007), but the INAD file was closed in 2019 and those uses are no longer authorized. Alpha-chloralose is currently used as a stupefying agent and avicide for controlling invasive bird populations in Australia and New Zealand. It has also been used in eradication campaigns but is considered labor-intensive and in some scenarios ineffective (Belant and Seamans 1999;Bednarczuk et al. 2010;Phillips et al. 2012). Sodium nitrite (CAS No. 7632-00-0) is an acute toxicant for European starlings and rock doves (Shapiro et al. 2017;Werner et al. 2021); thus, it may be a future direction in the development of new avicides.

Fertility Control (Contraceptives and Nest Destruction)
Reproductive inhibition is often considered when conventional control is not feasible or culling of charismatic animals is viewed unfavorably (Fagerstone et al. 2010; Table 1, Figure 1, see Table S1, Appendix 1). Nicarbazin ; also known as OvoControl®) is a multi-feed contraceptive that affects egg hatchability by altering yolk pH. It is non-toxic, reversible, and cleared from the body after 48 hours, but requires daily ingestion prior to and during egg laying (Avery 2014). Nicarbazin is the only contraceptive currently registered in the United States for birds and it can be used to target rock doves, European starlings, and common mynas (Avery et al. 2008a;USEPA 2022). Nicarbazin is effective in aviary settings for rock doves (Avery et al. 2008a) and eared doves (Olivera et al. 2021) but varied in its ability to reduce feral rock dove populations (Giunchi et al. 2007;Albonetti et al. 2015;Senar et al. 2021).
Another potential multi-feed contraceptive for use in invasive birds is 20,25-Diazacholesterol dihydrochloride (CAS No. 1249-84-9; also known as DiazaCon and Ornitrol), which reduces fertility by reducing blood cholesterol and cholesterol-dependent hormones to disrupt egg production. 20,25-Diazacholesterol dihydrochloride was previously registered for rock doves, but the registration was voluntarily cancelled by the registrant in 2003 and is not currently registered in any other products (USEPA 2022). Lambert et al. (2010) indicated that 10 days of dosing at 18 mg kg -1 reduces fertile eggs across the breeding season and causes rose-ringed parakeets to incubate infertile eggs for 3× the normal period, but it has not been tested in the wild. 20,25-Diazacholesterol dihydrochloride was found to reduce monk parakeet clutch size in an aviary and by 68.4% in a field setting (Yoder et al. 2007;Avery et al. 2008c). In house sparrows, 20,25-Diazacholesterol dihydrochloride reduces hatching success but is not registered for the species (Mitchell et al. 1979).
Although fertility control appears promising, managers need to establish bait stations for target species and limit nontarget exposure (Tillman 2016;Anderson et al. 2022b), requiring free-ranging target birds to reliably feed at these stations (Peck et al. 2014;Avery and Shiels 2018). Bait stations may work for small populations of urban birds (i.e., rock doves; Pellizzari 2017) but remain questionable where birds have dispersed into rural settings with abundant alternative food (Lambert et al. 2010). Surgical sterilization is a fertility control option used in longer-lived, large-bodied vertebrates. Endoscopic vasectomy of male feral rock doves resulted in reduced fertilized eggs in rock dove housing (Heiderich et al. 2015), but it is an impractical method to reduce wild populations. Barriers to chemical contraception include lack of products for permanent sterilization, large breeding populations, long lifespans (e.g., parakeets), risks to nontarget species, cost, and regulatory requirements.
Manipulating the nesting environment or destroying eggs reduces reproductive success (Ridgway et al. 2012; Table 1, Figure 1, see Table S1, Appendix 1). Corn oil (CAS No. 8001-30-7) is a minimum risk pesticide active ingredient under 40 CFR 152.25(f) that can be used for oiling and suffocating eggs in pest bird nests (USDA-APHIS-WS 2001b; Fagerstone et al. 2002). Oiling or addling eggs (i.e., shaking to destroy viability while leaving the shell intact) is often preferred over removing or smashing eggs, given that birds will continue to incubate, resulting in delayed renesting and continued occupancy of the nest site (Lambert et al. 2009;Fernandez-Duque et al. 2019). However, finding adequate nests to impact populations is often labor-intensive and logistically difficult, especially given the nest site characteristics of many invasive species (e.g., high cavities in tall trees or buildings).
Removal of preferred nesting habitat has been suggested to limit invasive birds, while considering the regional assemblage and requirements of native birds (e.g., removal of invasive trees to limit breeding cavities; Khan 1999;Yap et al. 2002;Gaudioso et al. 2012;Dodaro and Battisti 2014; Table  1, Figure 1, see Table S1, Appendix 1). Preferred nesting cavities could be filled to restrict breeding (i.e., artificial building cavities not conducive to nesting by native birds) or natural entrance holes modified to limit access to invasive birds while maintaining function for native species (Strubbe and Matthysen 2009;Charter et al. 2016;Le Roux et al. 2016). Nest boxes can also be used as traps to remove breeding birds or destroy eggs (Jacquin et al. 2010;Canning 2011;Tidemann et al. 2011;Campbell et al. 2012b). Feral rock doves responded to egg removal by renesting with implications for reduced female body condition due to the extra resources required for renesting, which may also increase a bird's susceptibility to disease (Jacquin et al. 2010). Nest removal was considered inefficient for reducing monk parakeets, even though their colonial nests are easy to locate and allow destruction of multiple clutches (Conroy and Senar 2009). Burgio et al. (2014) suggest excluding monk parakeets from electrical lines adjacent to utility poles to reduce nest building.

Deterrence Methods
Deterrence to reduce damage is appropriate when eradication of invasive species is not feasible (i.e., habitat modification, exclusion, frightening devices, and chemical repellents; Table 1, Figure 1, see Tables S2, S3, Appendix 1). In most cases, deterrence is short-lived and requires constant perseverance in moving and combining devices to create novel environments that birds find alarming (Avery and Werner 2017). A persistent deterrence campaign may be cost effective; however, such economic valuations are not always possible. Although nonlethal methods are required for the immediate protection of valued resources, such methods may shift bird damages to neighbors or other resources. Ultimately, population reduction is preferred, given fewer birds result in less damage (e.g., less fecal matter or less crop loss). Unless invasive populations can be eradicated, integrating nonlethal damage management with population reduction measures is necessary. Nonlethal methods are often given preference by the public out of concern for animal welfare (Crowley et al. 2019;Ribeiro et al. 2021). However, the use of nonlethal methods alone does not always result in cost-effective damage management when bird numbers are overwhelming, along with the decline in tool efficacy over time (e.g., habituation) and the limited extent of tool effectiveness over space (Linz et al.2015;Klug 2017).

Exclusion, Camouflage, and Repellent Methods
Complete (e.g., netting) or partial exclusion (e.g., overhead wires that obstruct landing) can protect crops and roost structures (Taber 2002; Table 1, Figure 1, see Tables S2, S3, Appendix 1). Farmers report that netting is effective but labor-intensive and expensive (Reddy and Gurumurthy 2003;Koopman and Pitt 2007). The spacing of overhead wires needs to be close enough to deter birds from passing through but wide enough to limit cost (Agüero et al. 1991). Small-bodied birds are agile and often not excluded by wires (Pochop et al. 1990). Nets have effectively kept European starlings out of livestock barns (Medhanie et al. 2015) and protected fruit crops from starlings, mynas, parakeets, and sparrows (Curtis et al. 1994;Abd El-Aal et al. 2006;Wang et al. 2020). Fewer studies exist that evaluate the use of nets or lines in urban or periurban areas (Andelt and Burnham 1993).
Covering fruiting bodies of crops with plastic containers or paper bags inhibits birds from detecting and selecting the crop or acts as a protectant to make crops inaccessible (Ruelle and Bruggers 1982;Conover 1987;Patel et al. 2002; Table 1, Figure 1, see Table S3, Appendix 1). Although effective, the practice is labor-intensive and cannot be implemented at a broad scale (Conover 1987). The availability of natural, alternative food is also be needed to reduce pressure on the wrapped plots (Dhindsa et al. 1992). Depending on environment, crop, and timing of management, the practice may also increase insects and mold (Dhindsa et al. 1992).
Physical devices to deter perching include sharp spikes, wire barriers, unstable systems of coils, and electrified cables by creating an uncomfortable or painful surface (Andelt and Burnham 1993;Haag-Wackernagel 2000;Seamans et al. 2007;Gorenzel and Figure 1, see Table S2, Appendix 1). Antiperch devices, such as spikes, have mainly been evaluated for urban rock doves (Harris et al. 2016) or other urban pests. Water can function to reduce visibility or cause birds to reflexively withdraw due to direct water pressure or wet feathers (Bishop et al. 2003). For example, a sprinkler activated by a motion detector can be set up to startle pest animals (Heidenreich 2007;McLellan and Walker 2021). Perch deterrents have reduced larger-bodied birds (i.e., rock doves) inside human structures and raptors on antennas ), but smaller birds can avoid these devices (Bishop et al. 2003).
Tactile repellent products containing polybutene (CAS No. 9003-29-6) are sticky pastes or gels that induce a negative reaction when touched and are registered as perch and roost repellents for rock doves, European starlings, and house sparrows (USEPA 2022; Table 1, Figure 1, see Table S2, Appendix 1). Gel repellents for feral rock doves have shown temporary, local efficacy, but concerns for animal welfare have been raised (Stock and Haag-Wackernagel 2014;Gagliardo et al. 2020). Under lab conditions, DeLiberto et al. (2020) evaluated three tactile, surface-application repellent formulations that reduced fecal accumulations under perches from European starlings under lab conditions (i.e., anthraquinone-based repellent Airepel® HC with castor oil; Airepel HC with castor oil excluding anthraquinone; and MS2 a novel inert formulation with a tacky, oily texture).
Anthraquinone (AQ; CAS No. 84-65-1) is a registered repellent for geese at non-residential turf use sites and for use as a corn seed treatment to repel starlings and other birds (USEPA 2022; Table 1, Figure 1, see Tables S2, S3, Appendix 1). AQ is a secondary repellent with a post-digestive antifeeding effect that is registered for use on starlings ; DeLiberto and Werner 2016) but may also have potential as an urban roost repellent (DeLiberto et al. 2020). AQ products are not available close to harvest for crops intended for the food stream but are available as a seed treatment at planting (Barzen and Ballinger Jr. 2018;DeLiberto et al. 2020). Anthraquinone has been tested as a feeding deterrent for monk parakeets, European starlings, rock doves, and house sparrows (Table 1, Figure 1, see Table S3, Appendix 1).
Methyl anthranilate (MA; CAS No. 134-20-3) is a registered repellent for rock doves, European starlings, house sparrows, and house finches and is sold as volatilizing paint-on liquids, blocks, and pouches, or as a fog or spray for perches and roosts (Table 1, Figure 1, see Tables S2, S3, Appendix 1). It can also be applied to a variety of crops and is aversive to birds by acting as an irritant on the trigeminal nerve (Mason et al. 1989;Avery et al. 1996;USEPA 2022). Although ineffective field application methods limit residues and thus efficacy (Kaiser et al. 2021). Application of repellents at crop emergence has been shown to be effective to reduce rock dove and house sparrow damage (Moran 2001;Esther et al. 2013). Near crop maturity, aerosolized treatment may be more effective than direct application to the crop (Stevens and Clark 1998), given aerosolized MA influenced avian flight lines at airports (Engeman et al. 2002). Monk parakeet behaviors indicated sensitivity to aerosolized MA but did not cause abandonment of established nests (Avery et al. 2006). Systems that deliver aerosolized MA are not recommended for public areas due to an adverse smell and the general application on fruit may influence how the crop tastes to humans.
Other chemicals have been tested to reduce feeding at planting through seed treatments or on turf (Table 1, Figure 1, see Tables S2, S3, Appendix 1). Methiocarb (CAS No. 2032-65-7) is still registered by the USEPA, but only for limited use as an aversive conditioning egg treatment (repellent) for crows and ravens (Corvidae) preying on the nests of protected species (Eisemann et al. 2011). Another example is thiram (CAS No. 137-26-8;Kennedy and Connery 2008), which is currently registered as a turf repellent for geese in the United States (USEPA 2022). Thiram products are registered as mammalian repellents and could be used for pest birds when applied in accordance with FIFRA Section 2(ee). Mannitol, an organic carbon derived from seaweed (ScudoSeed), and aluminum and ammonium sulfate (Eurodif) have been tested as seed treatments in Europe (Furlan et al. 2021), but these active ingredients are not registered for birds in the United States. Natural plant derivatives (e.g., mint, caffeine, and cinnamon) have been tested with some classified as minimum risk pesticide active ingredients and sold as repellents. However, variable efficacy in the lab and a lack of economic incentives result in few commercial products (Crocker et al. 1993;Gill et al. 1998;Hile et al. 2004;Day et al. 2012).

Auditory and Visual Frightening Devices
Frightening devices offer temporary protection of days to weeks but are not long-term solutions (Avery and Werner 2017; Table 1, Figure 1, see  Tables S2, S3, Appendix 1). Success is limited by avian fidelity to established feeding areas and habituation to nonrandom noise, as well as device shortcomings including extent of effectiveness, static location, and labor intensity (Gilsdorf et al. 2002). Early implementation and random presentation of a variety of sounds and visual deterrents used in combination or reinforced by a negative stimulus such as shooting is recommended . Most frightening devices have not been objectively tested in the field and when tested the appropriate replication and controls are not easily achieved (Bomford and O'Brien 1990). Definitive statements about tool effectiveness are not feasible given the unpredictable nature of wildlife damage that varies with pest species, protected resource, tool's mode of action, and surrounding landscape and habitat.
Bioacoustics include predator sounds (e.g., barking dogs, raptor calls, human noise) and avian distress and alarm calls (Gorenzel and Salmon 2008; Table 1, Figure 1, see Tables S2, S3, Appendix 1). When natural vocalizations are used as opposed to nonnatural auditory deterrents, habituation may be delayed because antipredator communication between birds is conditioned with actual predator threats. Ribot et al. (2011) tested broadcast alarm stimuli in orchards and found reduced activity of crimson rosellas (Platycercus elegans), a pest parrot species within its native range in Australia. Rose-ringed parakeets have been temporarily deterred from crops in both invasive (Hawaii; CJ Anderson, personal observation, 2021) and native ranges (India and Pakistan; Khan et al. 2011;Mahesh et al. 2017) using species-specific alarm calls and predator calls. Although distress calls can repel conspecifics and reduce crop damage (e.g., starlings; Conover and Perito 1981;Berge et al. 2007;Delwiche et al. 2007), distress calls may draw in some species (Conover 1994), a behavior that should be considered when attempting to deter pest birds but also when lethally removing invasive species through shooting or trapping. In addition, distress calls may repel nontarget passerine species, potentially impacting their behavior and space use. Despite documented success of bioacoustics, results are short-lived, and a continual rotation of tools is needed to prolong deterrence (Heidenreich 2007;Cook et al. 2008). Enos et al. (2021) found that compared to passerines, parrots and doves are less represented in the literature evaluating biologically salient frightening devices, suggesting continued testing on a variety of pest birds.
Land managers use scarecrows to mimic human predators in appearance and movement (Marsh et al. 1992; Table 1, Figure 1, see Tables S2, S3, Appendix 1). Inflatable "wavy men" have been reported as effective by landowners, but limited replication in studies revealed uncertain or minor effectiveness (Steensma et al. 2016;Lindell et al. 2018b). The addition of unpredictable loud sounds coupled with motion can increase effectiveness of scarecrows, but most birds habituate if deployed in established foraging grounds (Cummings et al. 1986). Wildlife species may begin to tolerate the appearance of human harassers when no painful stimulus is included (Griffin and Boyce 2009), thus reducing the efficacy of hazing (Grant et al. 2011). Alternatively, wildlife can become sensitized to actual threats, which allows for fear conditioning or the paring of painful and benign stimuli (e.g., modeling scarecrows after humans performing aversive actions; Blumstein 2016). Dead bird effigies (e.g., real feathers) have been used to disperse vultures and crows from roosts (Avery et al. 2002b(Avery et al. , 2008b, but monk parakeets did not respond to parakeet effigies at nest sites (Avery et al. 2002a).
Falconry, native predators, raptor models, humans, and protector dogs create an environment that prey birds perceive as risky (Table 1, Figure 1, see Tables S2, S3, Appendix 1). Hawk kites are suspended predator models that move in the wind, but efficacy is limited to directly below the model (Hothem and DeHaven 1982;Conover 1983Conover , 1984. Passive encouragement of natural predators (e.g., nest boxes and raptor perches) capitalizes on natural predator-prey systems (Kross et al. 2012;Harris et al. 2018;Lindell et al. 2018a;Peisley 2017). The response of monk parakeets to raptors was more pronounced when not associated with stork nests, which impacted nest locations (Hernández-Brito et al. 2020). Endangered falcons reintroduced to vineyards have resulted in reduced crop damage and pest birds, including European starlings (Kross et al. 2012). Although increased predation is often insufficient to reduce population growth (Bendjoudi et al. 2013), it is unknown if increased predator abundance reduces damage due to a "landscape of fear" (Laundré et al. 2001). Falconry has been used as a controlled predator method for starlings (Daugovish and Yamomoto 2006) and monk parakeets (Rodríguez and Tiscornia 2005), but the high cost and temporary responses of pest birds are limitations (Erickson et al. 1990). Rock doves have been successfully deterred from structures with a combination of falconry and shooting (Ryzhov and Mursejev 2010;Heck and Schwartze 2020). The most successful hazing strategies for Canada geese (Branta canadensis) was a combination of border collies (Canis familiaris) and remote-controlled boats, which allowed for hazing in areas inaccessible to dogs (Holevinski et al. 2007). The use of protector dogs to disperse invasive birds may be possible in row-crops and have successfully reduced other vertebrate pest damage (e.g., deer deterred from fruit and vegetable farm; VerCauteren et al. 2005). Limitations of dogs to deter birds are inherent at roosts and orchards due to the perching locations and flight abilities of birds. Uncrewed aircraft systems (i.e., drones) can be used as dynamic hazing devices that reduce human safety risks and operation costs compared to crewed aircraft, while overcoming mobility limitations of stationary devices to reach refugia used by birds (Klug 2017). The efficacy of drones depends on species-specific responses to drone type and flight dynamics and is influenced by flock size and landscape context (Egan 2018;Wandrie et al. 2019;Egan et al. 2020;White 2021). Drones have been tested to move rock doves from buildings (Schiano et al. 2021) and starlings and mynas from vineyards (Wang et al. 2019. Some species may mob drones (e.g., rose-ringed parakeets; Shiels and Kalodimos 2019), suggesting the possibility of moving birds toward other control tools (i.e., within shotgun range).
Pyrotechnics include products that produce flashes of light and whistles, whereas propane cannons produce a loud (e.g., 160 dB) directional blast (Bomford and O'Brien 1990). Loud noises have mainly been tested in rural areas and include studies on rose-ringed parakeets, starlings, and house sparrows (Table 1, Figure 1, see Tables S2, S3, Appendix 1). Advantages of combustion devices are initial affordability, portability, and inexpensive operation and maintenance. Disadvantages include fire hazards, habituation without reinforcement by negative stimuli such as lethal removal, the need to routinely move devices, reduced range in adverse weather, and inability to use in human-inhabited areas due to noise complaints . Auditory deterrents are limited in range with suggestions of one cannon per 2-3 acres (Avery and Werner 2017). A Vortex Ring Accelerator Deterrent (VRAD) propels exhaust through a vortex ring generator via combustion which then passes through an accelerator creating a high-velocity vortex ring of air movement that is propelled up to six miles at speeds up to 200 mph, frightening or dispersing birds with sound and a non-lethal physical concussion. The device has been considered for keeping waterfowl out of mine tailings (Opar 2017) but has not been tested. The sound intensity and physical concussion effect makes the VRAD an unlikely method for urban roosting sites or periurban agricultural sites.
Understanding the auditory physiology of pest birds and how it influences social interactions and antipredator behavior will inform the use of sound deterrents (Table 1, Figure 1, see Tables S2, S3, Appendix 1). The auditory sensitivity of most birds is between 2-5 kHz with an upper limit of < 10 kHz, thus most birds cannot hear ultrasonic sound (> 20 kHz ;Beason 2004;Jenni-Eiermann et al. 2014). Sonic devices used in abandoned buildings for feral rock doves did not reduce their populations (Haag-Wackernagel 2000) but did show temporary changes in behavior (Woronecki 1988). A "sonic net" produces sound that masks avian communication (2-10 kHz at 80 dB SPL) to effectively displace pest birds (i.e., European starlings; Mahjoub et al. 2015;Swaddle et al. 2015;Woods et al. 2022). When birds cannot communicate or hear predators, risk perception increases, resulting in abandonment of foraging grounds. The deterrent response is enhanced with real predatory threats and alternative food resources (Werrell et al. 2021). The sonic net is not feasible in urban areas due to the noise being audible to humans.
Novel objects can temporarily repel birds (e.g., tape, streamers, flags, balloons, mirrors, and lights; Table 1, Figure 1, see Tables S2, S3, Appendix 1). The reflectance, physical barrier, and sound of wind through the lines elicits a fear response that varies by species and environment (Bruggers et al. 1986;Dolbeer et al. 1986;Tobin et al. 1988;Conover and Dolbeer 1989;Firake et al. 2016). Large gaps between devices allow access, thus narrow spacing and routine maintenance influences efficacy, but increases cost (Bishop et al. 2003). Reflecting ribbons, plastic bags, and silver plates attached directly to plants have limited parakeet damage in sunflower (Basappa 2004;Shivashankar and Subramanya 2008) and dove damage in soybeans (Firake et al. 2016). Flags have been successfully used against birds that are considered pests in their native ranges, such as red-billed quelea (Quelea quelea) in rice, blackbirds (Icteridae) in corn, snow geese (Chen caerulescens) in winter wheat, gulls (Larus spp.) in loafing areas, and corvids in roost trees (Cardinell and Hayne 1945;Manikowski and Billiet 1984;Gorenzel and Salmon 1992;Mason et al. 1993;Belant and Ickes 1997). Eyespot balloons trigger a fear of being observed by predators. The influence is context-and species-specific and often short-lived (Greer and O'Connor 1994;Bishop et al. 2003;Fukuda et al. 2008) with mixed success at reducing bird activity at foraging and roosting sites (Shirota et al. 1983;Tipton et al. 1989;McLennan et al. 1995). Lasers along with rotating, strobe, and barricade lights are silent deterrent options for pest birds (Blackwell et al. 2002;Gorenzel and Salmon 2008), but light pollution and human safety should be considered. Lasers have been used to temporarily deter rose-ringed parakeets from roost trees in Hawaii ), rock doves from buildings (Matsyura 2018), and reduce blackbird and starling damage in sweet corn (Brown and Brown 2021). Monk parakeets are sensitive to red lasers (50 mm aperture, 650 nm, 50 mW [class3 IIIb]) and their use reduced birds at a nesting colony but did not reduce overall numbers in the area (Avery et al. 2002a). Automated laser models are available, which spatially and temporally confine laser beams to increase safety and reduce labor. Lasers may be an option for moving birds out of tall trees or buildings or encouraging movement closer to control tools (i.e., within firearm range or towards mist nets).

Modification of Crops and Surrounding Habitat
Modification of roosts and foraging habitat can reduce pest bird presence and damage (Table 1, Figure 1, see Tables S2, S3, Appendix 1). In North Dakota, cattail roosts were modified to disperse large blackbird flocks . Managers can also remove loafing sites surrounding crop fields to reduce habitat suitability  including trees for rose-ringed parakeets (Schäckermann et al. 2014) monk parakeets (Canavelli et al. 2014), European starlings (Lyon and Caccamise 1981), and house sparrows (Benras et al. 2019). Habitat surrounding airports can have tall as opposed to short grass to limit foraging by monk parakeets and European starlings (Marateo et al. 2015). Modification of urban roosting sites may not be possible due to building design or established landscaping (Klug and Homan 2020), but suggestions for rock doves include minimizing holes and using specific ledge widths and angles for new buildings to deter perching (Haag-Wackernagel and Geigenfeind 2008).
Crop damage varies due to field location or timing of maturity (Khan and Ahmad 1983), thus changing location or size of fields may reduce damage (Table 1, Figure 1, see Table S3, Appendix 1). Changing the sowing times, planting depth, or distance between seeds reduces damage at emergence among rose-ringed parakeets (Mahli 2000), house sparrows (Alizadeh 2009;Abd El-Gawad et al. 2010), and rock doves (Nasu and Matsuda 1976;Lawson 1979;Firake et al. 2016). Mukherjee et al. (2000) found crop damage was more severe at edges of sunflower fields, thus using larger fields or reducing the space between plots may limit preferred foraging spots where birds can maneuver and be vigilant to threats (Subramanya 1994). However, it is important to note that smaller plots allow better access for deployment of control tools . Small, diversified farms may be at greater risk because birds can meet their nutritional needs in one location as a different crop is continually ripening throughout the year. It is suggested to coordinate planting time with neighbors to eliminate early and late-maturing crops in the same locality . Advancing the harvest date reduces the damage window and yield loss from birds (Clark et al. 2020). In cereal crops, harvest date can be advanced two weeks by desiccating the crop without compromising quality or yield . Advancing harvest can also be practiced in some fruit crops.
Bird resistant crop varieties have been tested for rose-ringed parakeets (Ejaz-ul-Hassan et al. 1994), monk parakeets (Castro et al. 2022), European starlings (Dolbeer et al. 1986;Woronecki et al. 1988), and house sparrows (Tipton et al. 1970;Seiler and Rogers 1987;Alizadeh 2009;Khaleghizadeh 2011; Table 1, Figure 1, see Table S3, Appendix 1). In many cases crops bred to be bird resistant are not preferred by producers due to reduced crop yield or quality. At livestock operations, changing to enclosed barns or troughs (Feare and Swannack 1978;Medhanie et al. 2015) and altering feed type, size, shape, or placement may reduce consumption by European starlings (Feare and Wadsworth 1981;Glahn and Otis 1986;Depenbusch et al. 2011). "Trap crops" or "decoy crops" have been suggested to lure in pest birds to prevent damage on higher-value crops (Iqbal et al. 2001;Kubasiewicz et al. 2016), although the concept has not been directly tested on any of the included invasive species. Fields positioned closest to the roosts are ideal locations for decoy crops (Khan et al. 2006;Hagy et al. 2008), but in some situations decoy crops should be positioned close to the field to be protected. Birds feeding in the decoy crop should not be harassed. Understanding the feeding physiology and behavior of the pest species will inform crop selection for decoy plots (Kotten et al. 2022). Decoy crops are more feasible where tillable land is available and alternative food is enticing. Additionally, alternative food can be provided by delayed disking of grain fields or delayed removal of unharvested fruits . Alternatively, supplemental food in urban or periurban areas has been shown to support rose-ringed parakeets (Clergeau and Vergnes 2011;Borray-Escalante et al. 2020), monk parakeets (Hyman and Pruett-Jones 1995;Borray-Escalante et al. 2020), Eurasian collared doves (Coombs et al. 1981), rock doves (Senar et al. 2017;Soh et al. 2021), European starlings (Crick et al. 2002Galbraith et al. 2015;Klug and Homan 2020), common mynas (Galbraith et al. 2015;Soh et al. 2021), and house sparrows (Galbraith et al. 2015;Bernat-Ponce et al. 2018. Thus, removal of supplemental food, including human wastes, may reduce bird damage in cities by reducing invasive populations (Table 1, Figure 1, see Tables S2, S3, Appendix 1).

Human Dimensions
Public support of invasive bird management programs can be critical to their success; culling programs of invasive parakeets were halted in Britain (Crowley et al. 2019), Spain (M. Sabaté personal communication), and the United States (Eaton-Robb 2005; Pruett-Jones 2021) due to public backlash. Conversely, programs that have incorporated public education and participation as part of the management process have attributed these efforts to their success (Bunbury et al. 2019;Saavedra and Medina 2020).
Steps should also be taken to promptly locate and remove dead birds to avoid alarming the public (Klug and Homan 2020). Societal preferences for tools to decrease wildlife damage are often related to sociopsychological and demographic factors. In Argentina, attitudes about native monk parakeets, perception of damage, and knowledge of tool effectiveness were important in management preferences (i.e., lethal vs nonlethal alternatives; Canavelli et al. 2013). Although education programs work to inform the public about invasive species, sometimes attitudes do not change because of intervention (Braun et al. 2010). Thus, eradication programs targeted at charismatic species can face public opposition (Blackburn et al. 2010), especially in urban areas where gregarious birds are a novelty (Burger and Gochfeld 2009). The longer a species is present, the more difficult eradication campaigns may become as positive public sentiment increases (Papworth et al. 2009). Conversely, Mori et al. (2020) found that number of loud calls negatively influenced human tolerance of rose-ringed parakeets in Italy and not time since introduction. Belaire et al. (2015) also found that European starlings and house sparrow were known for negative qualities including loud calls and damage to personal property. In urban environments it was shown that public education programs to limit supplemental feeding reduced feral rock dove populations (Senar et al. 2017). Emphasis should be placed on campaigns informing the public about the harm caused by invasive birds, while being sensitive to animal rights groups and exploring positive collaborations when possible (Perry and Perry 2008). For example, Crowley et al. 2019 suggest reconfiguring management approaches to be more anticipatory, flexible, sensitive, and inclusive to minimize conflicts. Lindell (2020) concluded that clear guidelines about effectiveness and feasibility of implementation increase farmer adoption of sustainable management tools, while also indicating the importance of markets, government policies, and research priorities of commodity groups. We suggest further research on perceptions of nonlethal damage management tools for use in integrated pest management strategies from urban to rural settings, including individuals directly affected by the damage as well as other concerned stakeholders (e.g., Herrnstadt et al. 2016;Sausse et al. 2021).

Rock Doves
Of the studies designed to reduce rock dove populations, four used shooting, seven used trapping, eight used avicides, 13 used fertility control, and four used natural predators (Table 1, Figure 1, see Table S1, Appendix 1). The most common damage management tools investigated for rock doves included methods where feeding areas can be established that do not impact nontarget species (e.g., contraceptives; Table 1, Figure 1, see Tables S1, S2, S3, Appendix 1). Given the relationship of doves to predators, we identified four studies that used falconry or raptor sign to control bird numbers. Due to the behavior of rock doves inhabiting urban areas, we found numerous studies evaluating what building designs promoted or deterred birds, what tools worked as antiperch devices, and the impact of supplementary food (Table 1, Figure 1, see Table S2, Appendix 1). We did not find as many studies focused on agriculture (n = 13), but those identified evaluated avicides (n = 1), chemical repellents (n = 9), cultural practices (n = 3), or visual deterrents (n = 1) that may reduce crop damage (Table 1, Figure 1, see Table S3, Appendix 1). Only studies conducted in agriculture evaluated damage (cultural practices = 3, visual deterrent = 1, chemical repellent = 1), not just numbers of birds, indicating the need to understand if declines in urban damage (e.g., reduced fecal matter or public complains) occur after implementation of population reduction or nonlethal management.

Eurasian Collared-Doves
All the studies concerning invasive Eurasian collared doves focused on habitat use or their unprecedented range expansion in the United States (Table 1, see Table S2, Appendix 1). No studies have evaluated lethal techniques to reduce bird populations (Table 1, Figure 1, see Table S1, Appendix 1), likely due to the species not reaching large numbers nor evidence of negative impacts to native species, human health and safety, or economies. If negative impacts are identified, future studies could focus on lethal and nonlethal techniques to minimize damages. Until then, we recommend more studies evaluating potential negative impacts of Eurasian collared doves as a nonnative bird in the United States.

Rose-ringed Parakeets
Three culling campaigns have been reported in the literature for roseringed parakeets (Hawaii, USA; Seychelles, Canary Islands) with all campaigns using shooting and trapping to capitalize on the roosting and flocking behavior of this gregarious species (Table 1, Figure 1, see Table S1, Appendix 1). Although lab studies have been conducted on the efficacy of avicides and fertility control, field implementation is not feasible for these approaches until nontarget species can be excluded. All nonlethal damage management tools (i.e., exclusion, visual deterrents, and auditory deterrents) investigated for rose-ringed parakeets occurred in agricultural systems in their native range (Table 1, Figure 1, see Tables S2, S3, Appendix 1). Most studies evaluating nonlethal management in agriculture focused on modifying the crop or surrounding habitat with exclusion or visual deterrents available for small-scale agriculture (e.g., ribbons, bagging crops, and nets). When invasive parakeets are impacting agriculture (e.g., Hawaii), nonlethal tools should be evaluated in concert with lethal methods (e.g., establishing population sizes in which nonlethal methods are effective since eradication is not likely). We did not identify any studies testing exclusion devices, visual deterrents, auditory deterrents, or chemical repellents to reduce urban damage aside from assessments of habitat use or food sources (Table 1, Figure 1, see Table S2, Appendix 1). Thus, more studies are needed to evaluate if declines in urban damage (e.g., reduced fecal matter or reduced public complaints) occur after population reductions or implementation of nonlethal management. Of the 20 field studies evaluating damage, all but four were in agricultural settings (habitat = 6, exclusion = 3, visual = 6, auditory = 4; Table 1, Figure 1, see Tables S2, S3, Appendix 1), whereas urban studies including damage often evaluated the ecological impact to other species.

Monk Parakeets
We found two studies that evaluated shooting to reduce monk parakeet populations, and these were to either mitigate damage to utilities from nest structures or reduce damage to crops within their native range (Table 1, Figure 1, see Table S1, Appendix 1). As found with rose-ringed parakeets, lab studies have been conducted on the efficacy of avicides and fertility control for monk parakeets, but methods to exclude nontarget species are needed (Table 1, Figure 1, see Table S1, Appendix 1). In South America, studies have tested avicides at the nest, but environmental concerns and the status of monk parakeets as native species requires alternative approaches. Of the 11 field studies evaluating damage, 10 were in agriculture (avicide = 2, falconry = 2, habitat = 4, crop variety = 2, visual = 2, bioacoustics = 1, chemical repellent = 2) and one was in an urban setting and assessed nest building behavior (Table 1, Figure 1, see Tables S1, S2, S3, Appendix 1). Most studies assessing monk parakeet damage have included nest destruction and its impact on subsequent bird behavior such as nest building or nest site selection. We found few reports that evaluated nonlethal damage management tools for monk parakeets in urban or agricultural settings, apart from modifying human structures (e.g., electrical towers) to minimize nests and thus damage (Table 1, Figure 1, see Tables S2, S3, Appendix 1). Most studies evaluating surrounding habitat, and the resulting bird populations or crop damage, have been conducted in the native range of South America. Most invasive populations of monk parakeets in the United States are contained to periurban areas, but as populations increase and their distribution expands, we will likely see more studies evaluating tools to reduce agricultural damages, especially for periurban agriculture within expanding human development. Until then, we recommend more studies evaluating human dimensions to increase public support of invasive bird management.

Common Mynas
Distress call are the most common nonlethal damage management tool investigated for common mynas, although most studies were not explicitly testing such sounds as an auditory deterrent to avoid damages (Table 1, Figure 1, see Tables S2, S3, Appendix 1). Common mynas are a popular test subject for behavioral studies due to their highly developed communication systems. Thus, studies evaluating response to distress calls or learning from observing conspecifics being trapped or attacked are common. We suggest future studies capitalize on the mynas' ability to learn and behavioral responses to auditory cues to effectively reduce damage to agriculture or native species. Due to common mynas' sociability, traps have been the most common and successful lethal tool. Fertility control, via understanding of nesting preferences and behaviors, is also prevalent in the literature to reduce impacts to native species. Additionally, evaluations of habitat use and food availability in urban to periurban sites have been investigated for mynas due to most negative impacts being directed at native species in or adjacent to human-modified environments. Lethal control by firearm is more common compared to European starlings because some eradication campaigns found the use of pesticides impractical or needed to target trap-shy individuals. Of the 11 field studies evaluating damage, seven were in agriculture (habitat = 1, netting = 4, visual = 4, bioacoustics = 1, chemical repellents = 2); whereas the other four studies evaluated damage in the form of ecological impact to other species (i.e., number of native birds or nest competition; Table 1, Figure 1, see Tables S2,  S3, Appendix 1).

European Starlings
European starlings exploit a range of habitats from urban to rural. Their nesting behaviors negatively impact cavity nesting birds while their omnivorous diet makes it an invasive species that inflicts significant damage to agriculture as well as human health and safety. Thus, research on lethal techniques has a long history (Table 1; Figure 1, see Table S1, Appendix 1). Avicides were first used in the late 1960s and are well-studied in starlings due to their gregarious flocking and foraging behavior. When starlings form large roosts during winter, they focus their daily activities in relatively confined areas making lethal tools an effective option to reduce bird numbers at livestock facilities or urban sites. Lethal control by firearms is not as common, but future studies should evaluate culling via relatively quiet air rifles when pesticides or trapping are not options. Fertility control approaches are limited due to large population sizes and lack of bait dispensers that target breeding starlings and avoid nontargets. Nest boxes have been used to capture starlings when the birds are dispersed during the breeding season. Additionally, competition with cavity nesting birds has resulted in nest box designs to exclude starlings. Starlings are problematic for crop damage in the summer and fall requiring research to gauge the efficacy of exclusion tools, frightening devices, and chemical repellents with many techniques also being tested in urban areas. Although research in chemical repellents has a long history in the laboratory, difficulty in reaching the necessary residue levels on crops with current application methods limits field efficacy. Exclusion devices (e.g., nets) are often the best method to reduce damage to crops but cost, labor, and farm size constraints cause managers to look for alternatives capitalizing on antipredator behavior, especially bioacoustics. Ultimately, starlings can become habituated to tools even with integrated techniques that incorporate auditory, gustatory, and visual senses; granted if the damage period is short these tools can be effective. Modifying habitat effectively reduces bird density, but the ability of managers to affect change beyond their jurisdiction limits a landscape approach. Thus, apart from regional initiatives to organize against an invasive species, most studies evaluate tools that can be implemented on site (e.g., feed pellet size for livestock). Of the 24 field studies evaluating damage after tool implementation, all were in agriculture (avicides = 2, natural predators = 4, exclusion 6, cultural practices = 5, habitat = 1, visual = 8. bioacoustics = 1, chemical repellents = 6), highlighting the need of future studies to include damage reduction estimates in urban areas (e.g., public complaints, cost of cleaning). Most field studies on auditory repellents (distress calls = 8) focused on the behavior or number of birds and did not evaluate damage. Thus, we recommend that future research include damage reduction estimates in addition to bird behavior and numbers.

House Sparrows
Lethal campaigns against house sparrows have a long history, but most recent studies aim to understand declines in their native ranges. As far as lethal control, two eradication campaigns have been reported in the literature for house sparrows. Most studies use or evaluate trapping given house sparrows' propensity to enter traps and affinity for human development. The number of avicide (n = 13) and chemical repellent (n = 19) studies is a function of the past work to limit house sparrow populations or their damage to agriculture, which is not as common today. Fertility control in the form of nest destruction or nest box design is still prevalent given the species potential impact on native cavity nesters. Although opposite goals, studies evaluating declines can inform methods to reduce house sparrows in their invasive ranges. Hence, we included the 14 urban studies and 5 rural studies evaluating habitat or supplemental food (Table 1, Figure 1, see Tables S2, S3, Appendix 1). Of the 31 field studies evaluating damage after tool implementation, all but one was in agriculture (traps = 2, avicides = 2, nest destruction = 1, cultural practices = 9, habitat = 7, netting = 1, visual = 5, auditory = 3, chemical repellents = 8). Most of the nonlethal damage management tools (i.e., habitat modification, exclusions, visual deterrents, and auditory deterrents) investigated for house sparrows occurred in agricultural systems outside of the United States (Table 1, Figure 1, see Table S3, Appendix 1).

Management Implications
We recommend an integrated pest management strategy including lethal and nonlethal tools specific to the damage problem, species, and environment. The effects of nonlethal tools are temporary given birds are capable of learning and habituating to threats that do not pose a consequential negative stimulus. Thus, success with nonlethal tools requires combining and rotating multiple techniques and negative reinforcement (i.e., shooting). To achieve population reduction, a coordinated and sustained lethal campaign is required at broad scales. Primary management tools for culling invasive birds include shooting and, if feasible, trapping and toxicants at foraging sites and hand net capture at roosting sites. We recommend shotguns for birds in flight and air rifles for precise removal of perching birds. Research is needed on fertility control via contraceptives given functionality is limited by the difficulty in establishing feeding stations, especially when natural alternative food is available. The primary nonlethal management tools for reducing bird damage at agricultural sites include 1) modifying the crop and surrounding habitat, 2) exclusionary devices, and 3) frightening devices. For crop management, we recommend 1) growing sensitive crops away from flight lines, loafing sites, and night roosts, 2) eliminating early and late-maturing crops in the same locality to avoid establishment of feeding sites, 3) advancing harvest to shorten the damage period, 4) delaying destruction of unused crops to provide alternative forage, and 5) using large plots with limited space between plots to reduce damage at field edges. Habitat suitability can be reduced by altering the landscape by 1) removing loafing habitat near crops and 2) providing alternative forage via decoy crops where the birds are not harassed. Devices can exclude birds from entire crop fields and orchards (e.g., netting over trees and plots) or limit access to parts of the plant (e.g., bags, netting, or plastic over fruiting bodies). Promising tools for hazing and bird exclusion include lasers due to visual sensitivity in many bird species, drones due to the ability to access hard to reach areas, and playback of naturally occurring bioacoustics (e.g., distress calls) that reduce habituation. Primary nonlethal management tools at urban roosting sites include 1) modifying habitat (e.g., antiperch devices, alternative landscaping, or trimming roost trees) and 2) deploying frightening devices (e.g., lasers or water devices) that make the roost undesirable. The mobility and cognitive ability of pest birds along with temporal and spatial variation in damage patterns makes estimating damage difficult and costly (Sausse et al. 2021). Nevertheless, future research should include adaptive management plans for population suppression or eradication in addition to efficacy tests of nonlethal management tools that include animal numbers and behaviors with subsequent damages. Linking damage reduction with culling effort, pest densities, or tools in urban and agricultural areas will inform efficacy and thus tool adoption.

Funding declaration
The State of Hawaii Department of Land and Natural Resources (DLNR) Division of Forestry and Wildlife and The United States Department of Agriculture, Animal and Plant Health Inspection Service, Wildlife Services, National Wildlife Research Center (USDA-APHIS-WS NWRC) funded this project. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. The findings and conclusions in this publication have not been formally disseminated by the US Department of Agriculture and should not be construed to represent any agency determination or policy. Any use of trade, firm, or product names is for descriptive purpose only and does not imply endorsement by the U.S. Government.

Authors' contribution
PEK, WPB, ABS, SRS research conceptualization; PEK, WPB, ABS, BMK, SRS ethics approval; ABS, SRS funding administration; PEK writing original draft; PEK, WPB, ABS, BMK, CJA, SCH, EWR writing review and editing. Table S1. References listed chronoligcally for studies evaluating population control tools at nesting, foraging, loafing, and roosting sites for the seven small-bodied, invasive birds inhabiting periurban habitats (* = field studies; D = field studies where damage was recorded). Table S2. References listed chronoligcally for studies evaluating nonlethal damage management at urban nesting, foraging, loafing, and roosting sites for the seven small-bodied, invasive birds inhabiting periurban habitats (* = field studies; D = field studies where damage was recorded). Table S3. References listed chronoligcally for studies evaluating nonlethal damage management at agricultural foraging sites for the seven small-bodied, invasive birds inhabiting periurban habitats (* = field studies; D = field studies where damage was recorded). Appendix 1. References for studies evaluating lethal and nonlethal damage management tools for the seven small-bodied, invasive birds inhabiting periurban habitats.