Gull Deterence Trials

*******************************************

As part of the proposed plans to construct a new reservoir site in Oxfordshire Thames Water have apparently been required by the MOD to provide evidence that large numbers of roosting gulls can be dispersed to avoid the possibility of aircraft strike. They have therefore enlisted the services of the CSL Bird Management Unit to carry out an eight week trial. They will be using a so far un-tested technique of casting laser beams across the reservoir at various stages throughout a certain number of evenings during the period. This will obviously cause disturbance to other species as well as gulls. Regular progress reports from both CSL and Thames Water will be posted on this page.

******************************************

The following update received on 6th February.
THE GULL DETERENCE TRIALS WILL FINISH ON 22ND FEBRUARY

The bird strike trial has been progressing well and we have made some
minor
changes to the methods we were employing in the last week or so. Our
consultants have provided a report this week on the proposed methods
for
the final few weeks of the trial, we expect to be able to complete
earlier
than originally expected.

We have intensified the deployment of the laser through the evening to
test
whether more intensive methods deterred the birds from returning so
quickly
as had been happening earlier in the trial. This seems to have been
successful.

Having established that the lasers will disperse birds that have
settled on
the reservoir and prevent them returning during the night, however, we
now
want to continue with 12 consecutive days dispersal to determine if
lasers,
when deployed over several nights, are sufficient to stop birds coming
in
in the evening without having to suggest using pyrotechnics or other
dispersal methods as part of the strategy for the proposed reservoir as
these might have a more adverse effect on non-target species. The 12
day
continuous night-time deployment of lasers (as opposed to 2 days a week
in
the original protocol) will shorten the trial significantly and the
roost
will, based on the responses of the birds so far, return to its former
numbers quite rapidly after the laser deterrence is discontinued.
There is no intention to permanently disperse
the
Farmoor gull roost beyond the completion of this trial.

**********************************************

The followinig update was received from CSL on 23rd January:

Update - week beginning 15th January 2007.

After a successful first week reducing the numbers of gulls using Farmoor reservoir, birds had retuned to pre-control numbers within three days of control being removed.

By Wednesday the 17th, over 13,500 gulls had returned to Farmoor at dusk. Lasers again resulted in all birds being cleared. To reduce the probability of an overnight build-up, a second visit was made in the late evening. Approximately 300 Black-headed gulls had returned by this time (11pm) and were rapidly dispersed with the lasers. By 0715 at first light, it became apparent that 2/3rds of the preceding nights birds had returned. It was not clear whether the birds present in the morning had roosted overnight or were using the site as a pre-departure site. Results may also have been confounded by a bout of particularly strong windy weather.

Monitoring the following evening showed a return of approximately 3500 gulls to the roost although particularly rough weather conditions may again have been an influencing factor. As with all previous efforts with the lasers, they were again successful at deterring all the gulls from the site in a short period of time (c.20 minutes). In contrast, Dix Pit had over 10,000 birds present with large numbers still arriving as darkness fell. This represented a record count for this site.

A later visit night was again undertaken at Farmoor. Based on the previous nights findings, and on the previous weeks radar data, the later deterrence visit was undertaken at 3am. An estimated 1500 gulls (all gulls present) were successfully dispersed. As with the previous mornings, however, approximately 5,000 birds had returned to the roost by first light. (Numbers at Dix Pit were even higher with 19,300 gulls recorded roosting there at 7.10am).

The majority of non-target species at Farmoor were again unaffected. Mallard showed a 50% decline after the 3am deterrence session. Whilst this decrease was within the variability recorded during the pre control period, non-target species will continue to be monitored carefully especially if additional sessions of deterrence are implemented overnight.

Summary
The study has set out to answer three primary questions in relation to flight safety.

1 - can laser deterrence techniques be used to scare gulls from a roost?
2 - can lasers deterrence successfully prevent gulls from returning to the roost overnight without affecting the presence of non-target species?
3 – In the event that gulls stop using a site as a roost, will they habituate to lasers and eventually ignore their use?


1. The initial monitoring during the first week of deterrence and the monitoring undertaken during this week of deterrence has confirmed that the laser system used is indeed highly effective at scaring gulls from the roost at Farmoor. Some birds appear to be moving to alternative roost sites and remaining there overnight.

2. Whilst the laser was effective at instantaneous scaring of gulls from the reservoir, they have returned to the roost during the night and several thousand have always been present the morning after deterrence has been implemented. The additional deterrence implemented during the middle of the night failed to prevent birds from returning by dawn. It was hoped that a single session of deterrence that scared gulls from the roost at dusk (in the manner of the earlier trial conduced by Kenward et al.) would be effective whilst reducing the likelihood of disturbance to non-target species.

3. In order to be able to evaluate habituation to the lasers it is necessary to find a laser deployment strategy that completely deters the target population. Whether birds are returning to Farmoor to use the site as a pre-departure site is immaterial as long as gulls continue to arrive at dusk. Based on the results to date, the level of deterrence required to achieve this will need to be increased. A change of protocol is therefore recommended.

Protocol
Previous trials to deter birds from this roost involved preventing birds from settling on the site from c.2pm onwards (Kenward et al). As the use of lasers has been restricted until after all people have left the site, it has not been possible to adopt this strategy. Given the effectiveness of the lasers at instantaneous dispersal, we anticipate that a move to an evening dispersal, followed by approximately 15 minutes scaring each hour during the night would both disperse the birds and prevent them returning, hopefully with the result that they will abandon the roost site altogether and stop arriving at dusk.

We therefore propose to carry out a through-the-night dispersal exercise commencing on Wednesday 24th Jan. TW Staff or other users on site should notice no difference as dispersal will not start until the site is vacated. All other protocols will remain the same.

We will evaluate the effectiveness of this dispersal strategy at the end of this week and, if necessary, suggest a longer term change in protocol for the remainder of the trial.

Regards

John Allan

Ps Andy Baxter is on leave for the next 2 weeks, please direct all communication through me until his return. Thanks

*********************************************

The following email was received from CSL on 15th January:

"Laser deterrence trial: Week 1 update.

1. The meeting held with the Oxfordshire Ornithological Society (OOS)
on
the 9th January 2007 successfully discussed the issues of concern.
Trial
deterrence of the gull roost at Farmoor reservoir using laser torches
was delayed by two days. Deterrence began on the 10th January at
approximately 1730 hrs.

2. Radar and visual observations during the preceding week confirmed
the
movements of gulls into the reservoir during the hours of darkness.
Increases of approximately 30% occurred overnight. Waterfowl numbers
remained comparatively stable overnight although reductions in the
numbers of Mallard and Tufted duck did occur over the course of the
night on some occasions. The majority of wildfowl moved from Farmoor
2
to Farmoor 1 during the night, apparently in response to increasing
numbers of birds in the gull roost on Farmoor 2.

3. On implementation of the laser scaring, gulls were successfully
dispersed. An instant response to the beam occurred resulting in the
departure of gulls either into the airspace above the reservoir or to
alternative roosting areas. Deterrence was continued whenever any
recently deterred gulls attempted to return for approximately 90
minutes
each evening after which the site was clear of gulls.

4. The impact on target and non-target species was evaluated based on
total numbers present before and after deterrence and via the use of
night vision equipment. Tufted duck numbers remained relatively
steady
between evening counts. Numbers declined (as per pre-trial findings),
between dusk and dawn. Attempts will continue to be made to identify
groups of non-target birds and avoid pointing the laser in their
direction. No disturbance or deterrence of specifically identifiable
individuals such as the Black-necked grebe that arrived before this
initial trial period began, were observed.
Gulls such as the Glaucous
gull and Yellow-legged gulls that were present within the main roost
itself were, as expected, disturbed. Numbers of other species present
did not appear to be affected by this first phase of the trial. One
Black headed gull was found dead at Farmoor on the second day after
control was implemented. This is consistent with numbers noted during
the pre-trail period.

5. Radar monitoring proved extremely effective in identifying
nocturnal
movements of birds between roost sites. Almost all gulls dispersed by
the laser moved to the Stanton Harcourt complex or in the direction
of
Port Meadows (Oxford). Subsequent movements, however, showed a return
to
Farmoor reservoir occurring at night. Pulses of birds returned to
Farmoor between approximately 8pm and 1.30am. The result was that
approximately half the numbers of birds observed at dusk, were
present
the following dawn. This contrasted favourably with the numbers
observed
during the pre-trial period when overnight increases of approximately
30-50% occurred.

6. The scaring protocol was maintained for this initial week of
deterrence as numbers arriving at dusk fell from c.15,000 gulls on
the
first night of deterrence to c.5000 gulls by night 4. Different
individual gulls will use the roost on different nights. This initial
phase will aim to 'educate' those gulls that do use the site
intermittently that they will have to expend additional energy should
they attempt to roost here. Should the numbers returning to the roost
between dusk and dawn fail to decrease, or the numbers arriving at
the
roost fail to decrease, additional further deterrence sessions at
later
times during the night will be included within the protocol.

Summary:

1. Hand held laser torches mounted on a tripod and swept across the
roost at Farmoor proved effective at deterring gulls during this
first
week of the study.
2. Gulls returned, in reduced numbers, during the night.

3. Unusual or rare gulls within the roost were deterred.

4. Some disturbance of Tufted ducks may have occurred but further
information will be required to confirm this. Other species appeared
to
be unaffected suggesting the lasers can be targeted specifically.

5. The presence of one dead Black-headed gull during this initial
phase
of the trial is entirely within the expected parameters for gull
mortality observed during the pre-trial period.

We will provide weekly updates on the progress of the trail as we
progress."

This statment recieved from CSL on 15th January:

Interim Update report 1.
Laser trial pre-control monitoring.

Andy Baxter
Issue Date: 12/01/07

CSL Bird Management Unit
Sand Hutton
York
YO41 1LZ
a.baxter@csl.gov.uk
Tel 01904 462071

Summary

· Trials to test the suitability of laser deterrents for use on the proposed Upper Thames Major Resource Development (UTMRD) reservoir were targeted at Farmoor reservoir for winter 2006/07.

· Farmoor reservoir consistently attracted over 6000 gulls to its roost throughout the pre-control period.

· Weather conditions appeared to be the key factor affecting fluctuations in the numbers of gulls using the roost.

· Secondary roosts existed at Dix Pit (Stanton Harcourt), Port Meadows (Oxford), and Abingdon.

· Small numbers of other non-target waterfowl were present at Farmoor reservoir in comparison to the numbers at Dix Pit.

· Large numbers of gulls roosting on the southern Farmoor reservoir appeared to result in a movement of non-target waterfowl to the northern reservoir.

· Numbers of non-target waterfowl at Farmoor reservoir fluctuated between nights but remained relatively constant during the night.

· Mallard, Tufted duck and Cormorant were the exceptions with numbers varying significantly between dusk and dawn.

· Dead gulls and other species were located after periods of strong wind or cold overnight weather.

· Pre-control monitoring at Farmoor reservoir confirms this as an ideal site in which to evaluate the effectiveness of laser light systems for deterring gulls.

1. Outline work programme.
Pre-trial monitoring for the testing of laser deterrents as part of the suitability assessment for the Proposed UTMRD began during the week beginning December 11th. Monitoring each week consisted of two morning and two evening sessions at both Farmoor reservoir and Dix Pit (within the Stanton Harcourt complex of water bodies). Morning counts involved five half hourly or three hourly counts of each species of gull on the water from dawn. Evening counts involved identical half hourly or hourly counts from mid afternoon through to dusk. Counts were undertaken using standard census techniques by trained CSL ornithological staff. Counts of all other species present were recorded at least once during each visit along with measurements of weather parameters, flightline directions where possible, and notes of deceased birds when found.

2. Issues log

2.1. Visibility (thick fog) during the week prior to Christmas (beginning 18th December) prevented a full suite of counts being undertaken. Nevertheless, sufficient non-target species counts and roost counts were obtained. Flightline movements could not be observed.

2.2. It was apparent from differences between final evening counts and initial morning counts that significant numbers of gulls were arriving to roost at the reservoirs by night. This will be investigated by radar monitoring prior to the deployment of deterrence.

2.3. Oxford Ornithological Society (OOS) raised concerns over the possible impacts of the scaring trial. Thames Water (TW), Central Science Laboratory (CSL) and OOS met on the 9th January 2007 to discuss the trial in more detail and to address the issues raised by OOS. TW will provide OOS with regular updates on the progress of the trial.

3. Monitoring timetable

Visits were made to Farmoor and Dix pit to undertake the monitoring work on:

11th/12th/13th December 2006
20th/21st/22nd December 2006
27th/28th/29th December 2006
3rd/4th/5th January 2007

Each visit consisted of the scheduled pm, am, pm, am counts at each reservoir.


4. Pre-control results
The purpose of this briefing note is to provide a summary of information gathered and an indicative report on findings and variables rather than a full statistical analysis of the data.

4.1. Use of roost sites by gulls
4.1.1. Farmoor
Farmoor provides the main roost location for gulls in the local area. The southern (main) reservoir holds almost all the gulls present during the night. The northern reservoir routinely acts as a pre-roost and holds some birds over night but these are vastly outnumbered by birds on the larger site. Total numbers have varied significantly depending on weather conditions but ranged between approximately 6000 and 29,000 birds. An approximate 2:1 ratio of Black-headed gulls to Lesser Black-backed gulls occurred throughout the monitoring period. Approximately 60% of gulls arrive at the roost within 1.5 hours of dusk and depart within one hour of dawn. The remaining birds arrive during the hours of darkness. Acoustic records and observations with night vision equipment and radar have confirmed large numbers of gulls arriving during the hours of darkness. Early morning counts showed roost size routinely increased by approximately a third during the night. Strong wind or thick fog appeared to be the main factor that reduced the numbers of gulls using Farmoor as a roost.

4.1.2. Dix Pit
The secondary roost at Dix pit numbered between approximately 2000 and 8500 birds again depending on weather conditions. Thick fog increased the numbers present at Dix significantly. The roost here was confused by the numbers of day loafing gulls present that used the water body to bathe following feeding bouts at the adjacent landfill site. Numbers of each species present mirrored those at Farmoor with a predominance of Black-headed and Lesser Black-backed gulls.

4.1.3. Others
Whilst not visited as part of this study, strong wind conditions appeared to result in large numbers of gulls passing due east over Farmoor reservoir in the evening. These birds could be turning and moving on towards Abingdon further south or heading for the Portmeadows wetland area to the west of Oxford city. Radar evidence prior to deterrence being implemented suggested the latter and will be discussed in greater detail in the next update.


4.2. Use of roost sites by non-target species
4.2.1. Farmoor
Numbers of non-target species present on Farmoor varied significantly between weeks during the four week pre-control period. The majority of species showed little change in the numbers present at dusk compared to the numbers present at dawn. Exceptions included Cormorants, Mallard and Tufted ducks. Less than 20 Cormorants were present at Farmoor at dusk as opposed to a peak count of 363 arriving in the mornings to feed. Tufted Duck and Mallard both remained either stable or demonstrated approximately 50% reductions in the numbers present overnight. This suggests movements to alternative roost sites or foraging in wetland areas may occur at night.

Many non-target species also moved from the southern to the northern reservoirs as the numbers of gulls at the roost increased each evening but the total number present did not change. Fourteen species of waterfowl were consistently present at Farmoor during the pre-control monitoring period. Several other species were recorded, but not on all visits. Numbers of non-target species present at Farmoor were significantly less than numbers recorded at Dix Pit.

4.2.2. Dix Pit
Peak and mean counts of non-target birds at Dix Pit routinely exceeded those at Farmoor. The shallow, marginally vegetated pit with islands and significant amounts of aquatic vegetation provides a much less sterile environment for dabbling ducks etc. Peak counts of 890 Coot were recorded at Dix compared to 12 at Farmoor. Likewise, 68 Shovelor, 74 Pochard and 32 Wigeon compare to 2 Shovelor, 5 Pochard and 13 Wigeon at Farmoor. Sixteen none target species were routinely recorded at Dix Pit with little or no movements away from the site at night.

4.3. Dead birds
The welfare of birds is a key component of any study that involves harassment techniques. Observations of dead or moribund birds were therefore made throughout the study period. A total of ten dead gulls and one moribund gull were recorded during the pre-control period. Four Lesser Black-backed gulls (one ringed), one Herring gull and one Black-headed gull were recorded at Farmoor during sub-zero temperature nights on the 20th/21st December. Two further Black-headed gulls were recorded at Dix pit on the same occasion. A single adult Lesser Black-backed gull was also recorded in a moribund state during the same period. During the extremely strong winds recorded in early January, two further Black-headed gulls were recorded dead at Farmoor.

Five cormorants were also recorded dead in the water at the same time. None of these birds appeared to be freshly dead, however, with site staff suggesting these may have been long dead birds recently extracted from entanglement in underwater netting.

5. Discussion
Observations undertaken during the pre-control monitoring period confirm that a large gull roost suitable for testing the effectiveness of laser bird deterrence torches is consistently present at Farmoor reservoir. The objective of the trial is to assess the suitability and potential of laser deterrence for deterring gulls from the proposed UTMRD as opposed to any risk reduction at Farmoor itself.

Roost site selection by gulls in the local area appears to be influenced by weather conditions. Strong windy weather or thick fog appears to result in fewer birds using Farmoor reservoir, nevertheless, several thousand were present throughout the pre-control period with numbers routinely increasing over the course of the night. This provides a suitable baseline from which to assess the effectiveness of deterrence when it is implemented.

Farmoor reservoir provides a 24 hour resource for non-target species. A flock of feral Greylag geese utilise the short cropped grass around the perimeter of the site to feed during the day. Structures on the water provide roosting areas for small numbers of Cormorants whilst the water itself provides food for these and other piscivorous birds such as Grey Heron, Great-Crested grebe and Goosander. Dabbling ducks are also present although numbers in relation to other local sites such as Dix Pit are significantly fewer. Nevertheless, the impact of the laser deterrence on non-target waterfowl should be assessable. It was apparent that several of the non-target birds move away from the area used by roosting gulls (routinely off the southern reservoir and onto the northern one) once the main roost has settled. It is entirely possible, therefore, that lasers will not be directly aimed at any non-target birds.

After a slight delay, laser deterrence was implemented from the evening of the 10th January. Findings from the study will be updated on a weekly basis from Monday 15th onwards.