Use of falcons to displace nesting gulls from an urban area: final report

6. Discussion

Falconry had a clear impact on gull behaviour within the Campaign Zone in Dumfries but by the end of the study had failed to reduce the numbers of pairs attempting to breed. It is unlikely that falcons would have been responsible for an increase in gull presence (as confirmed by behavioural observations (see figures 2, 3a & 3b)), hence other factors must have influenced the overall increases in numbers. Strong breeding success in earlier years could have resulted in increased return rates or, similarly, the displacement of between 40 and 70 breeding pairs of gulls from an industrial estate in Cargenbridge (outwith Dumfries, approximately 4 Km to the south), may have combined to increase numbers. Without mark-recapture studies it is not possible to determine how these factors will have influenced the results.

The initial programme clearly resulted in gulls perceiving a significant threat thereby suggesting the technique has merit for development. During week one, for example, the mere presence of falcons on their perches (weatherings) was sufficient to have all gulls within a 300m radius circling in the air. At this stage, actual flights were not required to create a deterrent effect. The short falcon flights implemented from individual rooftops, therefore also had a long lasting impact and consistently kept gulls on the wing and off the ground throughout the day. This response level demonstrated gulls had either previous experience of predation from hunting Falcons, or were being impacted through a neophobic response caused by the novelty value of falcon presence within the townscape. It was apparent, however, that both species of gull began to tolerate the presence of falcons as the trial period progressed and no actual threat occurred.

It is likely that the start date for the trial was implemented too late to prevent returning birds from settling into the area to breed. This factor, combined with the testing of non-hunting falcons only, is thought to have resulted in breeding gulls firstly being settled into the area in the first place, and secondly developing a tolerance of the falcons by the end of the trial. By the time gulls were forced to either begin nest building and egg laying, or depart the area (half way through the trial programme), the majority were showing signs of habituation. By this point, the longer range (between different rooftops), longer lasting flights of falcons were not sufficient to cause significant disturbance. Several gulls remained sitting and exhibited relaxed, if observant behaviour as falcons flew past at short (<20m) range. Without any lethal reinforcement by hunting falcons or the deployment of other techniques as part of an integrated strategy, the end result was not affected by the initial disturbances created.

Early disturbance at breeding colonies is often cited as key to preventing an area being attractive to gulls when they arrive back from their wintering grounds (e.g. Abbeystead gull colony (Coulson, 1991)). Lesser Black-backed and Herring Gulls were already present within the townscape of Dumfries hence a recommendation for any future management activity would be to ensure deterrence methods are in place prior to the return of gulls. If immediately presented with a potentially hazardous breeding location, gulls would have a greater opportunity to relocate to a new area. The reduction in the numbers of sub-adult birds in the area suggests these birds were disturbed and may have been influenced to breed outside the Campaign Zone in future years. Further data would be required to determine if this is the case.

The total number of breeding gulls present across the town of Dumfries increased in 2009 over and above the numbers recorded in 2008. The programme implemented therefore failed to reduce the nesting population. As there is no literature suggesting that falconry should increase the presence of birds, however, (Baxter & Allan, 2006), a more general change in numbers between years must have occurred. The cause of this is unknown. Many populations of inland gulls are known to be rising and recruitment is considered a key driver of these changes. Urban breeding gulls are thought to have higher survival rates than coastal cliff nesting gulls (Rock 2005) hence successful nesting in 2005 may have provided an increase in the number of returning adults available in 2009. In line with this, very little egg control was implemented within the Campaign Zone in 2008 thereby making the area favourable to newly prospecting birds visiting that year. Similarly, destruction of an industrial area at Cargenbridge and the removal of habitat for between 40 and 70 nesting pairs of gulls could equally well have resulted in displacement into the Campaign and Control Zones. These factors are difficult to qualify and can only be measured via individual bird recognition and use of a mark-resighting scheme.

The potential for falconry to have displaced gulls into the surrounding area is also difficult to quantify. Anecdotal evidence suggests that fewer non-breeding, prospecting gulls were present thereby suggesting return rates in 2010 could be reduced. Behavioural data from the monitoring programme also suggests that gulls were generally less disturbed within the Control Zone and that such locations may therefore have offered a better breeding area. Data from the censuses, however, show that birds actually moved into, rather than out of, the whole area. Increases observed within the Control Zone cannot therefore be attributed to decreases in the Campaign Zone.

The lack of day roosts observed in the Campaign Zone and anecdotal evidence of reduced levels of sub-adult and juvenile birds suggests the falcons may only have been tolerated by birds that had already elected to breed. If prospecting birds in 2009 did not perceive a credible breeding location due to the disturbance levels created, 2010 could be positively affected by the efforts of 2009, although it will not be possible to tease apart the relative effects of the falconry from the egg and nest removal programme.

What was also evident from the results was that Lesser Black-backed Gulls showed higher levels of response to the flights of falcons than Herring Gulls. As a less robust species, it is possible that Lesser Black-backed Gulls could be targeted more frequently by wild falcons and may have developed a greater sense of 'fear' at their presence. During studies in Devon in 2004 and Aberdeen in 2009, Peregrines were seen taking Common, Lesser Black-backed and Black-headed Gulls but not Herring Gulls (pers obs). This indicates that Herring Gulls may indeed be aware of the threat falcons could create. As the study progressed, the lack of actual threat from the falcons resulted in fewer gulls of either species exhibiting a disturbance response. Disturbance was momentarily increased in week 8 when a single, large, Gyr x Saker hybrid was briefly introduced from one of the buildings. Whilst the results of this flight are masked by the generally tolerant behaviour exhibited at that time, all gulls within view of the bird took flight. Variation of species used alongside creation of an actual threat (hunting bird), may thus assist with developing and continuing disturbance.

Neither gull species were observed exhibiting aggressive behaviour to people. Observations undertaken by falconers, monitoring staff and census staff all failed to witness such behaviour despite gulls routinely being observed in close proximity (within 5 metres) to the public when food was made available. The majority of aggressive activity is, however, normally associated with the period between late incubation and the fledging of young when the adult birds will protect the investment they have made in the development of their young. This period did not arise during these studies.

Falconry clearly impacted, therefore, on the behaviour of gulls but was insufficient to create an overall reduction in breeding numbers in 2009. The deployment of any novel scaring technique will always elicit a response from a newly targeted population (Bomford & O'Brien, 1990), but maintenance of response rates over time becomes increasingly difficult. Companies offering falconry-based services routinely suggest that target birds have an innate fear of falcons but this is clearly not the case if routine deployment of falcons that do not hunt is deployed. The opportunity to use hunting falcons was not possible, however, as the method was aimed at a highly public area and did not want to result in capture of a gull within a busy public area. The inability to re-enforce the use of falconry with an element of lethal control, however, prevented a full assessment of standard falconry methods from being evaluated.

The methods developed by the bird control team were new methods that, to our knowledge, have not previously been used in an urban environment. Coverage of a wide area achieved by flying birds between different rooftops and members of staff represents a breakthrough in this field. Being able to present the level of coverage that was achieved by the end of the programme from the outset of any future studies may therefore help to increase the value of such activity across many sites. It would also benefit the control strategies if gulls were to have the perceived threat from the falcons reinforced whenever indications of habituation or tolerance were observed. This could be through direct lethal control, through deployment of hunting / novel falcons or via other techniques and strategies. It may be possible, for example, to use dead gulls as lures for the falcons so that birds in the area are able to witness the falcon's apparently 'catching and dispatching' conspecifics. Further discussion is required to determine the most appropriate forward strategy but evidence from many other studies confirms that a lethal element or perception of a lethal element of control can create a significant improvement in non-lethal methods (Baxter & Allan, 2008).

Further study to streamline the implementation and timing of the falconry programme, trial the inclusion of a lethal element of control and allow the adoption of complementary strategies to facilitate sustained reductions over a wider area is therefore recommended.