E cacy of Using a Laser Device to Reduce Wild (Water)Birds Visits to the Free-range area of a Layer Farm Situated in an Avian In uenza Hotspot-region in the Netherlands


 Free-range layer farms have a significantly higher risk of introduction of avian influenza viruses (AIV) compared to indoor layer farms. Wild water birds, natural reservoir of AIV, likely play a role in the transmission of AIV to chickens by contaminating the farms’ free-range area. The use of a laser as a repellent device for wild birds has been studied since the 1970s, in particular around airfields as bird-strike prevention, but its use at poultry farms to keep wild birds away from the surroundings of the poultry barn has not been studied. Part (1.5 ha) of the free-range area of a layer farm that directly boarded the poultry barn was equipped with a video-camera recording system. Visits of wild birds to this free-range study area were recorded for a month without and subsequently a month with a Class-III B laser in operation in the winter period. The laser was operated in the free-range study area between 5:00 PM and 10:00 AM, chickens were present in the free-range study area between 10:00 AM and 5:00 PM. The laser was operated in surrounding grass pastures between 10:00 AM and 5:00 PM. Mallards (Anas platyrhynchos) and Western barn owls (Tyto alba) visited the free-range study area between sunset and sunrise; Blackbirds (Turdus merula), Western yellow wagtails (Motacilla flava), White wagtails (Motacilla alba), Common Moor-hen (Gallimula chloropus) and Common kestrels (Falco tinnunculus) visited the free-range study area between sunrise and 10:00 AM. The overall (all bird species) efficacy of the laser for reducing the rate of wild birds visiting the free-range study area was 98.2 %. This efficacy was for birds of the Order Anseriformes 99.7 % and for birds of the Order Passeriformes 96.1 %. The overall exposure time of the free-range study area to wild birds (all species) was 617 times lower with the laser in operation compared to a situation in which the laser was not in operation. Exposure time to wild birds of the Order Anseriformes – known as bird flu risk species - was 2,320 times lower compared to the situation without the laser in operation. Summarizing, it can be concluded that the laser equipment used in this study was highly proficient in keeping wild birds – in particular wild ducks - away from the free-range study area of the layer farm.


Introduction
Wild birds are the natural reservoir of low pathogenic avian in uenza viruses (LPAIV), and wild birds are implicated in the dissemination of both LPAIV and highly pathogenic avian in uenza viruses (HPAIV) 1,2 over longer distances. Surveillance studies in birds of the orders Anseriformes (several species of swans, wild ducks and geese) and Charadriiformes (several species of waders, gulls and terns) indicate high LPAIV prevalence. Especially dabbling ducks show high LPAIV seroprevalence [3][4][5][6][7][8] . Therefore, several wild bird species of these orders are considered AIV high-risk species 5,9,10 .
A signi cantly higher risk of introduction of LPAIV and HPAIV is observed for free-range layer farms in comparison to layer farms where the chickens are kept completely inside a poultry barn [11][12][13][14][15] . The majority of LPAIV introductions on Dutch poultry farms is observed on free-range layer farms 13,15 . Wild birds visiting the free-range area may infect chickens by direct contact 16 or poultry may get infected via indirect contact, like drinking water from water pools formed in the free-range area in the autumn and winter period due to excessive rain, that are contaminated by wild bird faeces 17 . Other indirect contact may be through contact of chickens with droppings from wild birds deposited on the soil in the free-range area 18 .
The use of a laser (an acronym for Light Ampli cation by Stimulated Emission of Radiation) as a repellent device for wild birds has been studied since the 1970s, in rst instance as bird-strike prevention 19 . Bird-strikes, collisions between an aircraft and one or more wild birds, were a prominent reason to look for methods and strategies to avoid wild bird presence at air elds [20][21][22] . Later on, lasers were used and evaluated for dispersing wild birds causing agricultural crop damage 23,24 and resolving human health and safety issues and nuisance problems at land lls, oil-rig platforms and recreational parks 25-29 : birds perceive the approaching laser beam on the ground as a physical danger and as a reaction leave and avoid the area.
For wild birds that are active during dusk and night, like wild ducks, of which their eyes are adjusted to vision under dark conditions 30 , it was shown that in particular mallards (Anas platyrhynchos) strongly reacted on a laser by showing avoidance behavior 31 .
Visits of wild birds to the free-range area of a layer farm were investigated recently using surveillance by a video-camera recording system (VCRS) 32 . Several species of gulls visited the free-range area almost daily from sunrise until 10:00AM between January and August. Mallards (Anas platyrhynchos) predominantly visited the free-range area from sundown till dawn in the period December -February. The investigation revealed no direct contact between the layers and wild birds in the free-range area. In addition, nearly all wild birds disappeared from the free-range area when the layers appeared from the poultry barn in the morning. Take-up of AIV by outdoor layers via water or soil in the free-range area, contaminated by wild bird droppings, was suggested as the most important route of transmission.
Reducing visits of wild birds to the free-range area seems to be a logic preventive measure in order to limit access of chickens to possibly contaminated soil and water. To our knowledge, the use of a laser device in poultry farming to reduce wild bird visits has not been tested in a scienti c setting yet. Hence, the objective of our study was to determine the e cacy of a laser device to reduce wild (water)bird visits to the free-range area of a layer farm situated in an avian in uenza hotspot-region in the Netherlands.

Free-range layer study farm
The free-range layer farm was selected in an earlier study 32 to investigate visits of wild fauna to the free-range study area using a VCRS. Our study farm experienced several LPAIV introductions in the last 12 years since start of production. As standard procedure for all free-range layer farms in the Netherlands, serological surveillance is performed every three months for presence of AIV antibodies within the national serological surveillance program 15 . About 38,000 laying hens are present at the start of egg production, housed during the night in a barn measuring approximately 120 m length x 40 m wide. According to regulation 33 , chickens have at least 4 m 2 per chicken space in the total free range area. The part of the free-range area that is directly connected to the poultry barn (size: approximately 1.5 ha) is fenced except for the large right side of that area. In this part of the free-range area (Fig. 1), a video-camera recording system (VCRS) was installed to monitor wild bird visits (we refer in the rest of the manuscript to this area as the free-range study area). Although a large open connection to the grass pastures surrounding the free-range study area exists, no layers would be found in the connecting grass pastures. This is not an unusual situation: in ocks of free-range layers, generally only a small proportion of the ock uses the free-range area at any one time, and in addition often with a very skewed distribution across the free-range area, with the majority of the poultry going outside remaining in the area immediately outside and close to the poultry barn [34][35][36] . Layers can access the free-range study area through pop holes in the side-wall of the barn.
In the free-range study area, a total of 22 willow trees were present, providing cover, shade and shelter from weather or predators. The private house of the poultry farmer, close to the poultry barn, is surrounded by some large trees. In the freerange area, several waterway ditches (about 1m wide, lled with water) are present, which are connected to wider waterways (see Fig. 1). In general but depending on weather conditions and season, the free-range area is made available to the layers mostly from 10:00 AM until sundown. The layers were never fed and watered outdoors. A dairy herd (60 dairy cows and 60 young stock) and 50 sheep are also part of the responsibility of the farmer.

Video-camera recording system (VCRS)
The free-range study area was equipped with a total of eight Hikvision low light Turret 4Mp video-camera's tted with a xed focus wide-angle 2.8mm lens (Hikvision, Hangzhou, China; www.hikvision.com) at a height of 4m above ground-level on poles (Fig. 2). It was practically impossible to also have VCRS available in the grass pastures, adjacent to the free-range study area, at hundreds of meters away from the poultry barn. Experience with the effect of the laser on wild bird visits in the grass pastures was anecdotal by daily observations of the poultry farmer.
Video cameras were connected to a TruVision NVR10 network video recorder with HDMI/VGA video-output and a 12 Tb hard disk for storage (Interlogix, United Technologies Corporated, NC, USA; www.interlogix.com/truvision). The cameras were equipped with IR LEDs enabling night-recording. Recording was done at a speed of 2 frames/s. Video recording covered the time the laser was operating in the free-range study area.

Laser equipment and application
We used an automated laser (Laser Class-III B), Avix Autonomic Mark II (www.birdcontrolgroup.com, Delft, The Netherlands) for this study. The laser was placed on a steel, three-legged tower (the top was 6m above ground level) that was secured to a concrete slab (1.5m x 1.5m) weighing approximately 2,000 kg (Fig. 3). The laser was positioned in the right upper corner of the free-range study area (Fig. 4). It uses a laser with a wavelength of 532 nm (green color), and has a variable power of 50-499 mW (depending on installation preference). Studies of the avian retina indicate that birds have a spectral sensitivity range of approximately 300-700 nm 37 . In general, A Class-III B laser is not capable of creating hazardous diffuse re ection except for conditions of intentional staring into the laser beam at distances close to the diffuser 38 . By regulation, the use of the laser is limited to the property of the farmer, and the laser beam is not allowed to shine on public roads and houses/buildings of neighbors. Sign boards should be put out at the boarders of the property, warning the general public: "Class-III B laser controlled area -avoid eye or skin exposure to direct or scattered radiation". Safety requirements to be met The laser has a horizontal projection range of 360˚ (continuous), and a vertical projection range of -50˚ to + 30˚. According to the manufacturer, the laser beam can shine 100m away for every meter the laser is placed above ground level. The movement of the laser beam over the ground is guided by a small computer, which is part of the laser; the laser-movement navigator is computer-programmed, by manual-visual selection of geographical waypoints in the landscape; the laser beam moves from programmed waypoint to waypoint.
The free-range study area was split into two parts, which were laser-beamed separately (one part laser-beamed in 3:30 min., the remaining other part in 5:30 min.), with an interval of 1 min. without laser-beaming between the two parts. The free-range study area was laser-beamed between approximately 5:00 PM in the evening -depending on changing length of daylight over time-(just after the chickens went inside the poultry barn by themselves due to arrival of sunset) until 10:00 AM the following morning, the time the chickens would be allowed from the poultry barn into the free-range area by the farmer. The grass pastures -approximately up to a distance of 600m from the laser -were laser-beamed between 10:00 AM in the morning and approximately 5:00 PM in the evening.
The total surface of the grass pastures was divided into 10 separate parts, that were laser-beamed separately, with an interval of 1 min. A total cycle to laser-beam the different parts took approximately 45 minutes. We choose not to laser-beam the grass pastures also during the night period because it would leave the free-range study area for a large period of time without laser-coverage.

Presence of wild birds in surroundings
In waterways and grass pastures surrounding the free-range study area, ocks of several different wild birds species can be observed in the autumn-winter period: barnacle goose (Branta leucopsis), greater white-fronted goose (Anser albifrons), greyleg goose (Anser anser), mute swans (Cygnus olor), Eurasian wigeons (Mareca penelope), mallards (Anas platyrhynchos), black-headed gull (Chroicocephalus ridibundus), lesser black-backed gull (Larus fuscus), cormorant (Phalacrocorax carbo) and oyster catcher (Haematopus ostralegus) 32 . The free-range layer study farm is located at close distance to the coast in the northern part of the Netherlands. This region is known for wild waterfowl ying along the coast during migration in the Autumn, Winter and Spring time.

Translating video recordings into data for analysis
Video recordings were replayed using the video-recorder with hard-disk storage, as done in an earlier investigation 32 . Features of wild bird visits and operation of the laser were entered into a database: observation date; laser in operation in the freerange area or not; identity of visiting wild bird(s) (Family, Order, Species); time (hh:mm:ss) wild bird(s) start presence in the free-range study area; time (hh:mm:ss) wild bird(s) end presence in the free-range study area; the total number of wild birds in the wild bird visit.
An unique wild bird visit was de ned as one or more of wild birds from the same species landing in -or sitting on the fence of -the free-range study area, and subsequently remaining until departing the free-range study area 32 . Wild birds visiting the free-range study area per observation day was de ned as the sum of wild birds in wild bird visits per observation day.
Exposure time -per observational day -of the free-range area to wild birds, which is the total time wild bird(s) visit the freerange study area per observation day, was calculated as the number of wild birds of the same species, multiplied by the total time of this wild bird visit.

Study design
The manufacturer of the laser equipment claims that the laser is able to decrease wild bird visits for 75% or more. The eld data from the earlier study 32 indicated that we could expect on average 8 wild ducks visiting per day (standard deviation: 10 duck visits) in the period November -February (situation without the use of the laser). With 95% con dence and 80% power and the above-mentioned expectations, the calculated minimum sample size (number of observation days) needed to test the e cacy of the laser to keep wild birds -in particular wild ducks -out of the free range study area, is 24 days with and 24 days without the use of a laser 39 . The study design was as follows: start with approximately one month of video-camera monitoring of wild bird visits without the use of the laser (December 2019; sunset was depending on the observation day between 4:28 PM and 4:35 PM; sunrise was between 8:21 AM and 8:48 AM), followed by one month of observations with the use of the laser (January 2020; sunset was between 4:35 PM and 5:25 PM; sunrise was between 7:52 AM and 8:48 AM). We choose the months of December and January because our previous study in this farm 32 showed that not only these were the months of peak number of daily visits of birds of the Order Anseriformes but also there was no differences in the average/median number of birds visiting between these two months, hence we would expect a negligible to very low in uence of natural temporal changes in the number of birds visiting: eg. reduction of birds visiting in January, which could not be attributed to the presence of the laser.

Observer agreement
Two observers converted the video-data into analyzable data. The identi cation of wildlife species was conducted by the observers, and if needed with support of a wild bird eld guide of Europe 40 and if needed by a colleague, who has had many years of experience with eld identi cation of birds. If exact species determination was not possible from the video recordings, the general species category was indicated with the addition of an unspeci ed mark. Inter-observer agreement in observations of wild birds as a measure of data quality was calculated using the Cohen's Kappa statistic 41 by having pairs of observers independently noting visits of wild birds for the same observation days. Cohen's Kappa was calculated for a) observations on a given night of the visits of wild ducks, which are considered AI risk birds; b) observations on other wild birds (e.g. song birds) that would possibly appear between sunrise and 10:00 AM in the morning after which the chickens would enter the free-range study area. Based on an a priori estimated inter-observer agreement of approximately 0.95 and a maximum acceptable error in the estimated inter-observer agreement of 0.10 and a 95% con dence level, a sample size of approximately 16 observation days is required 39 .

Statistical analysis
To assess e cacy, the rate of wild birds visiting per observation day was compared between the period with and without the laser in operation on the free-range area of the layer farm. To estimate these rates and compare them, a negative binomial regression model was tted, which corrected for overdispersion observed when tting a Poisson regression model 42 . Two models were tted, the rst model aimed to compare the overall rate of wild birds visiting between periods with and without the laser in operation. In this model the response variable were the total number of birds visiting per observation day and the explanatory variable was a variable identifying the presence or absence of the laser. The exponent of the estimated regression coe cient of this model is the relative risk (RR) of birds visiting the free-range when the laser was present.
The second model tted used as explanatory variables the taxonomic Order of the birds visiting and the interaction with the variable identifying the presence/absence of the laser. The exponent of the bird order:Laser-presence interaction represents the RR of birds from each Order visiting the free-range study area when the laser was present. The threshold for signi cance of the estimated parameters was p < 0.05. Next we estimated the percentage e cacy of the Laser for reducing the daily rate of wild birds visiting as: Laser e cacy=(1-RR_Laser)*100.
Comparisons between the period with the laser absent and present were also made on the number of days with observed wild bird visits within each experimental period and the free-range exposure time. Number of days with visits were compared using the Chi-square test and the exposure time was compared by tting negative binomial regression models using the same approach as that described for assessing the rates of wild birds visiting per day.
All analyses were done using the statistical software R version 4.0.2 (R Development Core Team, 2020).

Results
Inter-observer agreement

General observations
Mallards ((Anas platyrhynchos) visited the free-range study area without the laser in operation between just after sunset (when the layers were already back into the poultry barn) and sunrise. Western barn owls (Tyto alba) hunted in the free-range study area on mice and rats during the night. Blackbirds (Turdus merula), Western yellow wagtails (Motacilla ava), White wagtails (Motacilla alba), Common Moor-hen (Gallimula chloropus) and Common kestrels ( Falco tinnunculus)) visited the free-range study area limited to the period after sunrise in the morning until 10:00 AM without the laser in operation. When the layers would ow out of the poultry barn and into the free-range study area at around 10:00 AM, these wild birds would disappear, scared by the large amount of layers in the free-range study area. It was expected to have a number of gulls of different species present in the free-range study area in the morning between sunrise and 10:00 AM. They were not observed this time, but were seen present in the surrounding pastures in the time period that the laser was operational in the free-range study area and at the same time not sweeping in the surrounding pastures.
Without the laser in operation, mallards (Anas platyrhynchos) visited the free-range study area of the poultry farm on 28 out of 29 observation days, while with the laser in operation, mallards visited the free-range study area on 1 out of 28 observation days (Fig. 5), which is a highly (p < 0.001) signi cant reduction in the number of days mallards visited the free range. Furthermore, without the laser in operation, other birds (Blackbird (Turdus merula), Western yellow wagtail (Motacilla ava), White wagtail (Motacilla alba), Western barn owl (Tyto alba), Common Moor-hen (Gallimula chloropus) and/or Common kestrel (Falco tinnunculus)) visited the free-range study area on 18 out of 29 observation days, while with the laser in operation, other birds visited the free-range study area on 4 out of 28 observation days (Fig. 4), which is a highly (p < 0.001) signi cant reduction in the number of days these wild birds visited the free range area.
Daily rate of wild birds visiting A summary of descriptive statistics on daily rate of wild birds visiting the free range study area with and without the laser in operation, grouped by taxonomic Order, are shown in Table 1. Wild birds of the Order Anseriformes, in this case only mallards (Anas platyrhynchos), visited the free-range study area often and with considerable day-to-day variation in the period without the laser in operation while virtually no mallards were observed in the period with the laser in operation (Fig. 5). A similar effect of the laser can be observed for visiting wild birds of the Orders Passeriformes, Strigiformes, Gruiformes and Falconiformes, although the number of visits are considerably smaller compared to those by the mallards (Fig. 5). Table 1 Summary of daily number of wild birds visiting during the experimental periods (28 observation days with and 29 days without the laser in operation), grouped by taxonomic Order, observed in the free range study area. The mean daily rate of wild birds (all species) visiting was signi cantly (p < 0.001) lower when the laser was used: RR = 0.0183 (95% CI: 0.007-0.043), which means that with the laser in operation, the mean rate of wild birds (all species) visiting is 55 (1/0.0183) times lower than when the laser is not in operation.

Daily rate of wild birds visiting by taxonomic Order
The mean daily rate of wild birds of the Order Anseriformes visiting the free-range study area was signi cantly (p < 0.001) lower when the laser was used: RR = 0.00327 (95% CI: 0.00018-0.0159), which means that with the laser in operation, the mean rate of wild birds of the Order Anseriformes visiting was 306 (1/0.00327) times lower than when the laser is not in operation. Furthermore, the mean daily rate of wild birds of the Order Passeriformes visiting was signi cantly (p < 0.001) lower when the laser was used: RR = 0.0394 (95% CI: 0.0149-0.0919), which means that with the laser in operation, the mean rate of wild birds of the Order Passeriformes visiting was 25 (1/0.0394) times lower than when the laser is not in operation.
To estimate the RR for birds of the Orders Falconiformes, Gruiformes and Strigiformes, they were combined into one group because of their low number of daily visits. Exposure time of free-range study area to wild birds A summary of descriptive statistics on exposure time of the free-range study area to visiting wild birds with and without the laser in operation, grouped by taxonomic Order, are shown in Table 2. The overall exposure time (all bird species) was signi cantly (p < 0.001) lower when the laser was used: RR = 0.00162 (95% CI: 0.000295-0.0089), which means that with the laser in operation, exposure time was 617 (1/0.00162) times lower than when the laser was not in operation. Exposure time to wild birds of the Order Anseriformes was signi cantly (p < 0.001) lower when the laser was used: RR = 0.000431 (95% CI: 0.0000246-0.00756), which means that with the laser in operation, exposure time of wild birds of the Order Anseriformes was 2,320 (1/0.000431) times lower than when the laser is not in operation. Furthermore, exposure time to wild birds of the Order Passeriformes was signi cantly (p < 0.001) lower when the laser was used: RR = 0.0575 (95% CI: 0.0129-0.2566), which means that with the laser in operation, exposure time of wild birds of the Order Passeriformes was 17 (1/0.0575) times lower than when the laser is not in operation.
To estimate the RR with respect to exposure time for birds of the Orders Falconiformes, Gruiformes and Strigiformes, they were combined into one group because of their low number. Exposure time to wild birds of the Orders Falconiformes, Gruiformes and Strigiformes together was 18 times (1/0.0544; not signi cant, p = 0.14) lower when the laser was used: RR = 0.0544 (95% CI: 0.00115-1.04175).

Observations by farmer in surrounding pastures
Since there was no VCRS monitoring in the surrounding pastures, we could only rely on anecdotal daily observations by the poultry farmer. He indicated that without the laser in operation in the pastures, sometimes large congregations of geese (barnacle goose (Branta leucopsis), greyleg goose (Anser anser), greater white-fronted goose (Anser albifrons), gulls (lesser black-backed gull (Larus fuscus), black-headed gull (Chroicocephalus ridibundus) and occasionally some swans (mute swan (Cygnus olor)) were observed to be present. With the laser in operation, the pastures were swept clean: no geese, gulls or swans were observed by the poultry farmer anymore during daytime.

Discussion
The measures of inter-observer agreement in this study are considered perfect 41 or excellent 43 , indicating the availability of high quality data for a sound scienti c analysis.
The objective of our study was to determine the e cacy of a laser device to keep wild birds -in particular wild ducks -out of the free-range study area of a layer farm situated in an avian in uenza hotspot-area of the Netherlands.  32 , particularly regarding the expected numbers of daily visits in December and January (see methods), we expect that the estimated e cacies can be mostly attributed to the effect of the laser. The overall exposure time of the free-range study area to wild birds (all species) -a direct measure of potential risk for contamination of the free-range study area -was 617 times lower with the laser in operation compared to a situation in which the laser was not in operation. Speci cally, exposure time of the free-range study area to wild birds of the Order Anseriformes -known as bird u risk species -was 2,320 times lower with compared to without the laser in operation.
Summarizing, it can be concluded that the laser equipment used in this study was highly pro cient in keeping wild birds -in particular wild ducks -away from the free-range study area of the layer farm.
Most earlier studies investigating the e cacy of a laser in relationship to wild birds were focused on hazing and dispersal, and in majority executed in experiments in small con ned rooms with a limited number of captured wild birds and over a limited time period, mostly a few days. Blackwell et al. 25  the cases they ew off. In summary most of these studies, using different lasers, reported e cacies higher than 80%, particularly for water birds (geese and gulls). Our study provides eld evidence for further con rmation of the reported high e cacies, in particular for mallards. Furthermore, the anecdotal observations of the farmer in the pastures are also an indication of the e cacy of the laser for reducing visits of geese and gulls in those areas.
An important requirement for habituation of wild birds to the presence and operation of the laser is that the stimulus will be repeated and is predictable 47,48 . Because the laser-beam is moving through the landscape in a seemingly random fashion, its predictability for wild birds is probably limited. In the 28 consecutive days that the laser was in operation in our study, we did not observe any possible habituation of wild bird species to the presence and operation of the laser. This is in accordance with other studies 24,25,28,44,45 although many were of limited length (only a few days).
There is very scarce data concerning health effects and risks for birds by lasers with laser-power equal or larger than the laser used in our study. Glahn et al. 44 conducted an ocular hazard assessment with a class IIIB laser with a power of 5 mW (moderate power). Five captive double-crested cormorants (Phalacrocorax auritus) were used for an exposure experiment: the birds were hand-held and their eyes exposed to the laser at a distance of 1, 13 and 33 m. Exposure distances were chosen based on the nominal ocular risk distance (NORD) to humans for this speci c laser (approximately 13 m) with the assumption of an eye-blink reaction time of 0.25 seconds; 24-and 48-h post-treatment examination of the eyes of the experimental birds revealed no ocular changes at the distances investigated. The manufacturer of the laser indicates that there is virtually no risk of eye damage for the wild birds because birds react strongly on the laser beam that is coming in their direction on the ground and y away before the laser beam might touch their eyes. In our study, similar to other investigations 21,24,27,46 , wild birds that were dislodged by the laser, settled down in pastures nearby, but outside the area that was laser-beamed. Indirect indications for limited to neglectable ecological damage to wild birds due to the use of a laser can be derived from the fact that after laser activity was stopped in experimental settings, wild birds would return back to the area mostly within one to a couple of days 21,23,25,45 . When we stopped the laser activity after 28 days of operation in our experiment, it took two and six days for wild birds of the Order Passeriformes and Anseriformes, respectively, to show up again in the free-range study area. The number of birds visiting returned to levels comparable to the period before the laser was put into operation in our experiment.
About 20% of free-range layer farms in the Netherlands have experienced one or more introductions of LPAIV in the period 2007-2018 (Fig. 6). We know so precisely, because the surveillance systems in operation in the Netherlands give us a complete insight into what is happening 13,15 . Close distance to waterways (a proxy for an attractive environment for wild water birds) and close distance to nature reserves with presence of AIV high-risk (water) birds were among the important risk factors for introduction of LPAIV on free-range layer farms 15 . So, for a limited number of free-range layer farms, situated in AIV high-risk areas and with a history of (recurrent) LPAIV introductions, the use of a laser might be helpful as preventive measure to keep wild water birds away from the free-range area of the layer farms.
There may be ecological implications for the mass application of these lasers by other poultry production type farms, these implications may need to be assessed before a recommendation of mass application is made. For now given both 1) the high risk of HPAIV introductions from wild water birds contaminating the area around poultry farms being experienced yearly since 2014 49 , and 2) the e cacy of the laser to reduce wild bird visits, it could be considered the use of the laser as a preventive measure in AIV hot-spot farms, similar to the one of this study.
In conclusion, in this study we con rm the high e cacy of using lasers to reduce the daily number of wild bird visits to the free-range area of a layer farm situated in a AIV-hotspot area. Given this high e cacy, the application of these lasers becomes a viable alternative for the prevention of introduction of avian in uenza infections in poultry.