Honey-buzzard in Scotland

The main aim in the new season is to write a book on Honey-buzzard, using my experience in British Isles, continental Europe and Africa, dating from 1993. The pioneering Scottish records below will form an integral part of the book

Field work in 2021 in Northumberland will complement this task, filling in gaps and reinforcing findings. Further lengthy trips in mind abroad are Israel/Tarifa/Gibraltar in September and Kenya Highlands and coast in due course in winter

Broad headings for book are: Motivation, Historical Records, Rediscovery in Britain, Migration Patterns, Arrival and Display, Breeding Activity, Habitat, Dispersal, Wintering Grounds, Four Identification Models, Value of Field Experience, Further Studies, Bibliography

The impression until recently has been that the Scottish population of Honey-buzzard is poorly known. However, as discussed below, publication of The Birds of Scotland in 2007 has revealed much more information than was in the public domain before. This development is very welcome. We also discuss below the exciting find of Honey-buzzard concentrations in central Scotland in from 2016-2019, as reported in Scottish Birds.

The table below summarises observations made by NR from 1998-2016 in Scotland. No visits were made to the country in 2005, 2011, 2013-2014, 2016-2020.

Because questions will inevitably arise as to whether the recent sharp increase in breeding Honey-buzzard reported for Scotland is an artefact or real, NR's full results from 1998-2015 are now published, in the form of a spreadsheet, sorted by date and latitude. 49 records from 32 places are itemised in a full report with no data withheld.

Maps are also published of sightings for Scotland as a whole (1998-2015), Highland (2000-2012), Perth and Kinross (2001-2015), Trossachs (2002-2008), Dumfries and Galloway (2004). Some further analysis of the results is planned for the book, including numbers, site density, habitat, seasonality, comparison with Northumberland and the recent results published in Scottish Birds.

Below is the extant broad-brush summary.

* total for distinct sites is less than than that obtained by simple summation of the data above; repeat counts of a site are eliminated in the totals. The reliability of this process has been increased by storage of all data on BirdTrack.

Area means roughly the land in a 40km radius of the named town or city. The site near Inverness at which birds have been satellite tagged is not included in the above totals. The overall totals are 32 sites including Perth 10; New Galloway 7; Inverness 7; Callander 5; Aviemore 2; Glasgow 1; Skye 0; Ullapool 0.

These are very much opportunistic sightings in visits lasting one week each year. It would be interesting to see more comprehensive figures if these are available.

The total of 32 sites above is significantly above the 14 reported for Scotland in the last UK results for 2000. It is though within the 30-50 pairs estimated from a recent habitat analysis for the whole of Scotland (see density studies). Certainly the nature of the above survey does not allow an extrapolation to be made for population purposes but the widespread distribution of the Honey-buzzard in Scotland is surely not in doubt.

The Birds of Scotland/Other Literature

Some interesting comments are in Scottish Bird News (no 86, December 2007, p.8, report on SOC Conference held on 2-4 November 2007). The population in Highland reached 15 pairs in the early 1990s and breeding was noted in Perth & Kinross and in Dumfries & Galloway at the same time. Most breeding attempts (93%) are successful. The species is probably overlooked as a breeding species in Scotland and may be more widespread, maybe reaching 50 pairs (From talk on European Honey-buzzard by Brian Etheridge with his summary).

These comments are based on the account for the European Honey-buzzard by Brian Etheridge (2007) in The Birds of Scotland. This account has a number of interesting points:

• Breeding population is currently 15-20 pairs, possibly up to 50 pairs.

• Regular breeding occurs in the four areas Highland, Moray, Inverness and Perth & Kinross, with irregular breeding in six others including Dumfries and Galloway.

• Decline in breeding pairs reported in Highland may be due to reduced effort by the few enthusiasts engaged in monitoring the species as monitoring new areas in response to forestry changes is time consuming. As casual sightings during the breeding season have not diminished, the overall population north of the central lowlands appears to be unchanged.

• Birds return from central African wintering grounds in the second half of May and most pairs are settled by the end of the month. High circling and sky-dance occur only briefly at this time of year, being mostly performed in July and August perhaps by non-breeding birds.

• Eggs are laid in late May or early June and the first young fledge in early August. Fledging has occurred as late as mid-September.

• Fledging success is high at 93% with 1.9 young/successful nest and 1.8 young/attempt.

• Preferred trees for nests were 55% Beech and 39% Oak in mixed woodland and 75% Douglas Fir and 19% Scots Pine in coniferous.

• Spacing between adjacent pairs in the same wood averaged 2.7km (range 2.2-3.6km, sample 5).

• The main threat is timber operations from June-August. There are no perceived threats from egg collectors and disturbance by birdwatchers is not a problem.

• Based on habitat preferences in the north and abundance of this habitat elsewhere in Scotland, Honey-buzzards may be far more widespread in Scotland than current knowledge suggests. Further survey work by experienced field workers is needed to reveal the true picture.

The report of the Scottish Raptor Monitoring Scheme for 2006 showed that effort was rather low in searching for breeding Honey-buzzards in Scotland with just 5 home ranges visited by Highlands RSG and none by other RSG.

In 2011 it was estimated in The Scottish List that 15-20 pairs bred in Scotland (Scottish Birds 31, Supplement 1, pp.10-11 (2011)).

Short note: McInerny, Christopher J, Honey-buzzard carrying a reptile during display flight, Scottish Birds 34(2) p.143 (2014). Author believes this is the first report of a reptile being used as a visual prey item during display flight by Honey-buzzard in Scotland. Comment by Brian Etheridge: “I have not witnessed this before, though I have watched this species display many times. Display in the late summer is quite frequent and appears to be made by non-breeding adults. It is often followed up the next year by a breeding attempt in the wood over which the display occurs.”

Comments by NR: NR has readily found birds in the main breeding areas cited by Etheridge (2007) but has also found significant numbers in Dumfries & Galloway where further study is required (by everybody) and around Callander (including the Trossachs) in Stirlingshire. In Northumberland the few times in which forestry operations have affected birds has caused the loss of the nest for that year and the birds have not moved into the nearest available wood: they appear to seek available optimal habitat within the same general area. So the apparent decline in the Highlands may be due to the whole area understandably not being searched each year.

Display in spring is very variable in Northumberland, being dependent on fine weather. The activity in July and August noted above (and in Wales) is very rare in Northumberland, perhaps because it is due to contention when optimal nesting sites are scarce and this state has not been reached yet in Northumberland. Fledging in Northumberland runs from mid-August (not early August) but fledging is counted as birds above the canopy which will inevitably be later than the strict time when the bird first flies within the tree cover. Routinely some birds fledge in Northumberland in early September. Fledging success is also high in Northumberland at c90% per breeding attempt: it is difficult to give a more precise figure as attempt needs to be carefully defined. The number of juveniles fledging per brood is apparently lower in Northumberland (1.3/successful nest in 2007) but the methods are different. Ringers often measure the number ringed whereas the number appearing above the canopy is the measure used in Northumberland. The fatality rate between ringing and appearing above the canopy may be significant and not all juveniles may rise readily above the canopy.

There is considerable difference between the tree species used for nests. In Northumberland Scots Pine and Norway Spruce are preferred in coniferous plantations and Oak in deciduous woodland. Northumberland preferences seem closer to those found in the work in Holland (Bijlsma 1993) than in Scotland: it does depend on the trees available and the areas selected by the observers to actually find nests. Spacing between nests in Hexhamshire in Northumberland, where a large area provides suitable habitat, is 2.5km, very close to the 2.7km found in Scotland.

Timber operations have so far in Northumberland been a very minor threat but if the private owners are tempted at some stage by high timber prices to do extensive felling this could cause significant problems. The success rate in Northumberland does not suggest that egg collectors are the problem they are made out to be. Birdwatchers in Northumberland have also not been a problem as the time that needs to be invested and the subtleties of identification have proved to be major barriers.

By 2014 when McInerny published his note, it appears that at least some raptor workers in Scotland are being more candid about the growing population of Honey-buzzard in the country. Brian Etheridge is based in Inverness.

Scottish Birds paper: KD Shaw, CJ McInerny, A Little, K Little, JS Nadin, R Goater, An exceptional season at a central Scotland Honey-buzzard study area, Scottish Birds 37(1) 3-13 (2017).

It is very satisfying to see a report of another successful Honey-buzzard study area in the UK. The site is kept secret but it's possibly the Tay Valley, Perthshire, which I have claimed several times as an area rivalling the Tyne Valley for Honey-buzzard. Indeed I would take this paper as confirmation of my earlier work. The study was made last summer (2016) so it's very topical.

Some specific comments:

p.4 History of Honey-buzzards in Scotland. The statement “However, many ornithologists who study Honey-buzzards, including the authors of this paper and the RBBP, believe that published UK population numbers are an underestimate” is strongly supported.

p.5, p.9-10 Study area has extent of 90 km². Five pairs bred: how does this compare to the Tyne Valley study area? It's good to deal in the distance between pairs, which is given as 3.5, 3.5. 3.0, 3.0, (all km) in the Scottish study, giving an average of 3.25km. The comparable distance is 2.5km in the Tyne Valley study area. In terms of habitat Northumberland has less large conifer blocks in the Tyne Valley than in the central Scotland study area, with more mature woodland in deep valleys.

p.6 Study area. Only two raptors, Common Buzzard and Goshawk, were seen interacting with Honey-buzzard. In the Tyne Valley study area, Hobby is the main interacting species at display time with Common Buzzard keeping a very low profile. At fledging time Common Buzzard is the main interacting species, particularly if the Honey-buzzard have nested fairly close to the Common Buzzard, but the aerial battle with Hobby goes on through the whole season. Hobby are scarcer in the Tay Valley than in the Tyne Valley but I have found a few there. Sparrowhawk also occasionally mob Honey-buzzard.

p.8 The number of non-breeders is much higher than in the Tyne Valley study area and the peak activity in July for aerial activity is not the pattern in Northumberland. It appears that the intense interaction between breeders and non-breeders causes the activity when the young have hatched. The Honey-buzzard in the Tyne Valley study area are almost invisible in July, only becoming active in the air a few days before fledging from mid-August. The number of non-breeders may be determined by how much surrounding habitat exists for a natural dispersion of 'new' breeders. Obviously this is an area for further study. There is a possibility that an unusual number of non-breeders were present in 2016, giving much more interaction than in a normal season and making the population much more visible than usual.

p.10 Young fledged from third week of August. This is exactly the same as in the Tyne Valley study area where at the lower altitude areas in the valley, young fledge from 15/8-22/8. At higher altitude, on the moorland edges in the Tyne Valley study area at up to 450m asl, young fledge significantly later, mainly in early September with a few as late as mid-September. Juveniles in the Scottish study were seen on 21/8, 24/8, 27/8, 6/9, 7/9, 11/9. This is very similar to the picture in Northumberland for lowland sites but at upland sites juveniles can be seen for about a further 2 weeks up to almost the end of September (27/9 in 2016). It is very significant that the juveniles were seen above the canopy: Roberts in his review of the Northumberland Honey-buzzard claimed that juveniles were never seen above the canopy: a result perhaps of a failure to recognise the juvenile's different shape and plumage.

Juveniles continue to appear in Northumberland until early November. These are thought to be migrants, from Scottish bred sites. The numbers vary considerably from year to year. 2016 was the 2nd best year on record with 22 juveniles noted in Northumberland (9/9-4/11) and 4 in the Lake District (9/10-12/10). Other good years have been 2014 (20 juveniles in Northumberland), 2012 (27, best to date), 2011 (12), 2008 (14). There may be a correlation with breeding success in Scotland. Since these are just opportunistic sightings, a substantial Scottish breeding population is indicated. The juvenile migrants (and locally-bred birds) seen in 2016 have been analysed in terms of colour-phase.

p.11 Breeding success is given as 7 juveniles raised by 5 pairs, that's 1.40 young per successful nest. In the Tyne Valley study area 1.69 young were raised per nest in 2016 with all 49 sites being successful.

p.12 Advice to others … Suggested survey time of July/August is very suitable for raptor workers, involved in other studies of earlier breeders such as Goshawk and Merlin, indeed of all other raptors except for Hobby. You are though vulnerable to being criticised as in the old saw: “Why are you looking over there: I thought it was lost over here: because the light's better over there!”. Someone targeting Honey-buzzard as top priority should definitely get out in the spring as you get a clearer perspective on what is going on.

It is worth noting that for birds fledging in mid-August, the laying date will be 80 days beforehand, at the end of May. So the adults will be sitting at this time, severely reducing visibility. The adult male in fine weather does do high circling right over the nest site at the start of the incubation period but this is difficult to detect on a casual basis as unless the bird is seen soaring it may be overlooked; such display is presumed to be a deterrent to the settling of late migrants, mainly young birds without territory. The pair apparently cease mutual display just before the first egg is laid. A fledging date of mid-August is associated with lowland sites in the Tyne Valley study area, where display is conspicuous in mid-May. At higher altitudes fledging is later so display might well be noted from late May to mid-June. So spring-watch is to be encouraged, with mid-May for lower altitudes and late May to mid-June for higher altitudes. Birds breeding on the edge of the grouse moors appear to breed the latest, maybe to coincide fledging with the peak productivity of the heather moor.

Identification is more challenging in autumn than in spring as Common Buzzard juveniles are readily confusable with Honey-buzzard juveniles. Identification is easier in May/June as Honey-buzzard is often the only displaying raptor, other than Hobby, with Common Buzzard on eggs or with small young. So understanding the spring position gives a solid baseline from which to work in the autumn.

p.12 Acknowledgement to FCS for sympathetic management of forestry. There are also some very good private forests in central Scotland, e.g. those belonging to Duke of Atholl.

Scottish Birds paper: CJ McInerny, KD Shaw, A Little, K Little, K Hoey, K Gibb. B Kerr, Honey-buzzards in central Scotland: observations and comparisons from two study areas during 2017, Scottish Birds 38(1) 15-29 (2018).

This paper reports on the following season, 2017, with two study areas now identified, the one is eastern central described for 2016 (SA1) and a new one in west central Scotland (SA2). SA1 is believed to correspond approximately to my Perth (Tay Valley) area above where I found Honey-buzzard at 10 sites; SA2 is believed to correspond approximately to my Callander/Glasgow area above where I found Honey-buzzard at 6 sites. Numbers found by The Scottish fieldworkers have been summarised in another paper The honey-buzzard in Scotland: a rare, secretive and elusive summer visitor and breeder, C.J. McInerny & K.D. Shaw, The Glasgow Naturalist (online 2018) Volume 27, Part 1, available at https://www.glasgownaturalhistory.org.uk/gn27_1/McInerny&Shaw_HoneyBuzzard.pdf:

In central Scotland the most recent population estimate for 2017 suggested up to 49 birds and 17 territories, including up to 24 breeders, seven non-breeders, nine non-assigned birds and eight or nine juveniles (McInerny et al., 2018). These numbers are an increase from 2016 when 23 birds were recorded, including 16 breeders and non-breeders and seven juveniles (Shaw et al., 2017). In years previous to 2016 even fewer were detected, despite similar observer effort, so it is possible that numbers are increasing, at least in central Scotland.

My own totals were obtained from 2001-2015 in the Perth area and from 2002-2008 in the Callander/Glasgow area, seemingly reaffirming the long-term presence of Honey-buzzard in the area, rather than simply a recent colonisation. Increasing experience with Honey-buzzard facilitates the finding of new pairs and I respectfully suggest that this is the situation here.

The paper in Scottish Birds provides some interesting photographs, drawings and description of the wing-clapping display, of individual birds and of habitat. Up to 9 territorial pairs were noted in SA1 and up to 3 in SA2. Habitat was generally managed woodland, in an upland situation with south facing slopes. In 2017 in Scotland 'a few' further territories were detected in similar habitat elsewhere and 2-3 territories in lower, mixed, agricultural areas. In Northumberland Honey-buzzard breed from high grouse-moors at c450m asl through dale woodland to lowland fields, right up to the outskirts of the Newcastle conurbation in the 'edge lands'. They use unmanaged deciduous woodlands in addition to managed woodlands.

The breeding densities are given as inter-nest distances of 2-4km apart. In Northumberland the nests are typically 2.5km apart where woodland is widely distributed. My studies in the Tay Valley showed a similar distance with the birds alternating the territories between one side of the Tay and another to give maximise packing.

The presence of Goshawk in the study areas is acknowledged and does not appear to have affected the Honey-buzzard success. This is interesting as in SW Northumberland Goshawk are heavily persecuted by the game estates and I had attributed their near-absence as a positive feature for Honey-buzzard. It may well be that Goshawk only take significant numbers of Honey-buzzard if other prey is scarce.

McInerny, C.J., Shaw, K.D., Hoey, K., Little, A., Little, K., Gibb, K., Kerr, B & Nadin, J.S. Honey-buzzards in central Scotland during 2018: a different breeding year. Scottish Birds 38, 308-313 (2018).

This paper reports on the season, 2018. This is labelled as a different breeding year, mainly because it was late starting and the birds in some territories proved difficult to monitor as to their status. Birds appeared to arrive later, were less visible, and performed little wing clapping during May and June, compared with 2017; the first bird was seen on 12/5. In Northumberland birds arrived back on 4/5 at my local site, in fine fettle. On 20/5 in Hexhamshire at typically 100-200m asl 6 birds were noted at 4 sites, in active display and in floating mode, but without wing-clapping. On a trip to Norfolk from 12/5-19/5 8 Honey-buzzard were noted at 6 sites, with some in display. So the arrival did not seem late and the birds seemed in good condition in England. However, in upland areas in Northumberland on the edge of the grouse moors up to 450m asl, the Honey-buzzard season often runs 2-4 weeks behind areas such as Hexhamshire and laying can be delayed until mid-June. These sites will not fledge young until early September. Display at these sites is still vigorous if the weather is good and even occurs in drizzle if prolonged bad weather occurs. The later start may be designed to take advantage of the heather flowering season in August/September and increased numbers of insects but it also enables the birds to avoid some of the wet, cold spells that can particularly affect the uplands in May/June in a 'bad' season. With the reduction in the coverage of the study area, the upland sites are being studied less in Northumberland so no pertinent observations were made in 2018 but observations in past seasons confirm the timing. It may be that the Scottish birds arrived on time but took time to get into condition; this would result in the birds being seen sporadically but not in full display. They may well have eventually got into breeding condition and gone onto breed (see later).

In central Scotland 5 pairs raised 7 juveniles in the area SA1 (east central) with 2 pairs failing; there was one 'failed' territory in SA2 (west central); in new areas SA3 'central' central there were 2 sites in May/June and at SA4 south-east central a single bird was noted in June. So that is 11 territories of which 5 were successful, raising 7 young. The weather was dry and warm in the summer of 2018; this is thought by the authors to increase Honey-buzzard productivity but not sure this is the case as wasps thrive in damp, humid weather and Honey-buzzard can dig more easily into wet soils than dry ones, particularly clay.

From experience in Northumberland this is apparently a very poor performance but do wonder about some of the failures. The site in SA2 where an adult was seen on 30/8, possibly on 2/9, does not seem like a failure. Birds in general vacate their breeding sites if the nest fails and start their migration to the wintering grounds. In central Scotland the birds arrived late in 2018 and displayed little: this might suggest that some were in poor condition and delayed nesting. Honey-buzzard in upland areas of Northumberland can delay their nesting to mid-June if conditions dictate, thus delaying fledging until the first half of September. Both birds in plates 252-253, taken on 11 August, look very much in territory; I respectfully suggest there is a well-hidden nest and juvenile(s) fledged in early September, quite possibly the bird seen there on 2/9. The site in SA3 where many positive observations were made in May could also have fledged young in early September. When I studied extensively upland sites in Northumberland, September was a very busy month for looking for recently fledged juveniles. I would suggest the season in central Scotland be extended until at least mid-September to look for juveniles above the canopy where occupancy established earlier in the season but no juveniles seen in August.

The observed increase in central Scotland in aggression by Goshawk is very interesting as this is a very major difference between the 2 study areas, Goshawk being scarce in the SW Northumberland study area because of persecution by keepers on game estates. It may be that the presence of Goshawk gives more variation in Honey-buzzard breeding success as the Goshawk may predate Honey-buzzard if other food becomes scarce. The presence of Goshawk may also increase the secrecy of the Honey-buzzard as they keep more under the canopy.

In the acknowledgements the main landowners appear to be the Forestry Commission. I think in the Tay Valley where I made many observations that the Atholl Estates is an important landowner. So if the central Scotland study area does not include much of the Tay Valley then our totals in east Central (SA1) are summative, significantly increasing the Scottish population. Secrecy is a hindrance to effective analysis.

Finally there is an independent measure of Honey-buzzard productivity in Scotland: migration counts of juveniles in Northumberland in September and October and even early November. This is discussed on the page http://nickrossiter.org.uk/hbweb/northumberlandhb.htm. Basically high numbers of juveniles were observed in SW Northumberland on migration in 2011, 2012, 2014, 2016, 2017 and low numbers in 2013, 2015, 2018, 2019 with the last year qualified by more limited fieldwork than usual. I think the pattern from 2016-2019 of high and low numbers matches reported breeding success in central Scotland. Full details of each migrant seen is given in the link above.

McInerny, C.J., Shaw, K.D., Mason, G., K., Little, A., Little, K., Observations on nesting Honey-buzzards in central Scotland during 2019. Scottish Birds 40, 72-75 (2020).

Studies have revealed two separate breeding populations, one in east central Scotland and the other in west central Scotland, with other pairs thinly distributed through the region. During the 2019 breeding season there were fewer pairs and non-breeders than in previous years; no detailed figures are given for the central Scotland study area but a poorer breeding season is indicated. A second nest was located giving 2 nests found in the study to date, both in conifers. In 2019 a series of photographs is given on the latest find in east central Scotland in a Forestry Commission wood. The nest is large with sprays, both deciduous and coniferous, on the rim. The female was much more attentive to the nest than the male, who was absent for long periods. The main food observed was wasp comb but 2 unidentified bird fledglings were also supplied. The older nestling was clearly more advanced throughout than the younger one, both in size and plumage. The young were aged at 2-3 weeks old, hatched in early July.

Honey-buzzard Home Page

Nick Rossiter 2007-2022