A PRE-BREEDING SURVEY ON AREAS PROPOSED FOR BRACKEN CONTROL WITHIN THE BLACK MOUNTAINS – MAY/JUNE 2018

A REPORT COMMISSIONED BY BEACONS NATIONAL PARK AUTHORITY

Author and Surveyor – Andrew King, AvesEnto Surveys

Summary

Introduction

The Black Mountains Land Use Partnership, a partnership of land-owners, graziers and public bodies, successfully bid for funding to assist and develop the sustainable management of the Black Mountains. Over £1 million was awarded by the European Agricultural Fund for Rural Development, and the Welsh Government to restore and improve natural habitats, to improve access to the public and to diversify economic opportunities for local communities.

Included in the project is controlling bracken, which threatens to encroach onto grazing land and into important wildlife habitats. This will assist the major land-use which is hill-grazing by sheep owned and managed by a large number of local farmers that exercise their common- land hill grazing rights. Not only will bracken control ease access to the hill for sheep, but also assist tourism which in the last twenty years has steadily increased in line with the popularity of hill-walking and mountain-biking. These have brought new pressures for both the uplands and the valleys that thread their way through the Black Mountains, but have undoubtedly stimulated the local economy.

Bracken can provide suitable cover and feeding and breeding opportunities for a range of upland breeding . These differ according to the density of bracken which ranges from forming a dense and impenetrable barrier, to being rather open with a grass sward beneath. The fern can harbour good numbers of invertebrates especially in late-summer, but it is the layer of decaying bracken litter at its base that provides small (, chats, warblers) a safe and relatively undisturbed nest-site.

Surveyor Experience

Lead-surveyor and author of this report was Andrew King (AvesEnto Surveys) commissioned to carry out ornithological survey of breeding birds that had the potential, or were physically nesting, within defined areas destined for bracken control. Andrew has over 40 years’ experience of carrying out bird surveys across many habitats and locations. Working in a voluntary capacity, he has participated in many British Trust for Ornithology (BTO) and Royal Society for the Protection of Birds (RSPB) surveys, a significant proportion in the uplands of Britain. He also holds a BTO bird-ringing licence specialising in summer reedbed/scrub species and on wintering farmland birds in , and participates in the monitoring of several species: Peregrine Falcon, Dotterel, Golden Plover and Curlew. In the last 15 years, he has expanded his considerable experience into commercial chargeable surveying, normally on a sub-contract basis, specifically on ornithological and entomological (butterflies and moths) projects.

Breeding birds and the Law

There is a general requirement under Law that breeding birds should not be disturbed, and that nests, eggs and developing chicks must not be wilfully destroyed. The Wildlife and Countryside Act (1981) describes an offence is committed if intent can be proven for activities that knowingly lead to disturbance and destruction of nests, eggs or young. While some damage is considered inevitable by undertaking such works as bracken control by ground methods during May/June, the fact that preliminary monitoring of breeding birds has taken place, and advice given and followed to mitigate against wholescale destruction of breeding sites, then the work may be considered lawful.

Background

Bracken control proposals to treat approx. 320 hectares at 25 locations were first developed in spring 2017. Funding delays and decisions regarding eligibility of capital works items within the project, meant that by March 2018, a reduction in treatment area to 156 hectares at 13 locations was required. A further 75 hectares at 3 sites were destined for aerial treatment. The lack of ground disturbance and optimal timing for aerial spraying being outside of the breeding season, these sites were not included for checks on breeding birds, and therefore do not feature in this report.

Treatment methods across the 13 locations involved combinations of rolling bracken, cutting by tractor mounted flail/mulcher, or a ‘roboflail’ remote-controlled cutter. Depending on progress and ease of working in 2018, the proposal is to treat bracken on these sites for three consecutive years.

Survey Methodology

Preliminary discussions between Natural Resources Wales (NRW), Brecon Beacons National Park (BBNP) Ecology team, and the lead surveyor had discussed various options for assessing breeding bird use of the bracken control plots. There is little historical ornithological data to guide this work on the Black Mountains, but the criteria for selecting the position and extent of the 13 locations for bracken control by ground treatment included ease of access, slope, minimal extent of ‘ffridd’ across the site, and density of bracken trash. Ffridd is a term used to describe upland/montane habitats in Wales that includes upland heath (Gorse, Heather, Bilberry, etc) and where there is regeneration of young trees and shrubs (most commonly Hawthorn, Rowan, , Ash and Holly. Presence of this habitat-type is attractive to a wider range of both resident and summer migrant birds, with a consequential increase in expected density of breeding birds.

For pure bracken sites, the following species that habitually nest on the ground (or within decaying bracken litter close to the ground) are the most likely to be encountered. They are listed in ascending order of preference for breeding in areas of dense bracken infestation:

 Wheatear (prefers rocky, short grazed swards with very low bracken density)  Skylark  Meadow (and Cuckoo that commonly use as nest host)  Tree Pipit  Reed Bunting  Whinchat  Stonechat  Wren (prefers thick bracken stands, deep litter, presence of other herbage eg nettles)

The likelihood of finding a nest, or observing behaviour of birds to suggest breeding on a bracken control plot, is expected to increase in direct proportion to proximity and quality of ffridd, depth of decaying bracken litter, and density of bracken stems/metre. These were not measured in detail for this project, but the monitoring results bear a direct relationship to these factors.

Additional species that also nest on the ground (or within bracken litter) expected where ‘ffridd’ occurs at the margins of bracken control sites:

 Robin  Willow Warbler  Yellowhammer

Ffridd species that would prefer to nest within the small tree or bush itself (rather than on the ground) includes Chaffinch, Redstart, Dunnock, Blackbird, Mistle Thrush, Garden Warbler, Linnet, Bullfinch, Blue Tit, Carrion Crow and Buzzard. These were all found in ffridd habitats on or close to the Black Mountains bracken control sites.

The sampling methodology agreed was as follows:

1. Observe an area of uniform bracken within the boundaries of a bracken control location, away from ffridd trees/bushes if possible, for 20 minutes. 2. Once the timed observation period is complete, for bracken control locations >10 hectares in extent, then mark out four corners of a 100m x 100m study plot. (For locations <10 hectares in extent, mark out a plot measuring 50m x 50metres). 3. On each study plot, make two diagonal walks linking the corner boundary canes. 4. Scan again at each corner marker, and at the mid-point. Note down GPS measurements at each of the five points. 5. Record species present, and any breeding activity. 6. Up to five of these plots should be marked out and monitored at all treatment sites.

Table 1: Bracken Treatment sites

Site Number Site Name Area (ha) Proposed treatment

1 Blaenyoy, 50.6 ha Tractor and Roboflail 2 Trevelog, Llanthony 5 ha Roboflail 3 Cwm Buchel, Llanthony 9 ha Rolling 5 Ffawydd, 6 ha Cut, roll and collect 6 Cadwgan, Grwyne Fawr 8 ha Cut, roll and collect 8 Llwyn-y-Brain, 9 ha Tractor 9 Cwm Milaid, Llanbedr 10 ha Roboflail 16 Cwm-Rhos, Cwmdu 5 ha Tractor and Roboflail 18 Blaen-ffirnant Farm, 4 ha Roboflail Pengenffordd 19 Grafog, Pengenffordd 4 ha Tractor and Roboflail 21 (south) Blaenau Uchaf, Tractor and Roboflail 21 (north) Penybryn, Talgarth ca. 7 ha Tractor and Roboflail 23 The Grange, Capel-y-Ffin 3.8 ha Roboflail 25 Llewenau Draw, Cwmdu 13 ha Tractor and Roboflail

Survey constraints

A number of constraints were recognised during the pre-breeding bird survey phase of the work.

The survey methodology was drawn up based around the Brown and Shepherd ornithological survey model, and adapted to provide assessments of breeding birds on a given bracken treatment site. These surveys at each site were carried out fully, with the exception of Site 5 (Ffawydd, Grwyne Fawr) for which the habitat appeared to have been incorrectly assessed and/or mapped. The methodology adopting multiple one hectare (10,000 sq metres) or quarter-hectare (2,500 sq metres) sub-plots across each control area generally worked well, but served only two of the three objectives. Those met were to identify bird species within and adjacent to each study plot, and to determine the likelihood (and from that the actual or likely nest-stage and timing) of breeding. However, extrapolating this further to provide a density figure of nests per hectare or control area proved much more difficult within the time allocated, and any quantification of density would have too many variables to be meaningful. The observation period at each sub-plot gave little insight into birds breeding within that sub- plot, but did give an opportunity to assess species within the immediate area. The physical act of walking through to measure out the sub-plot, and by making the required diagonal sweeps across it had the effect of flushing both feeding and breeding birds. Nests were found (or suspected) during these procedures.

The preceding 2017-18 had been dry and mild up until early January. Two periods of severe cold temperatures and snow occurred during March, followed by wet weather until mid- April. These had dual impacts of driving birds off common-land (commoner resident species had already arrived back in late-February) and reducing body condition. For resident species, this forced a relatively late breeding season, and even then some birds appeared not to have bred at all. Summer migrants arrived back at dates close to normal, but the main arrival was slightly delayed. Emergence of bracken was also slightly delayed, and by mid-summer in the last week of June, the uppermost fern fronds had only just completely unfurled. The general impression was that numbers of commoner species (Meadow Pipit and Skylark) in all upland habitats were lower than normal.

Dry and settled, warm weather assisted the task of accessing and carrying out the surveys. During the first weeks of hot, dry conditions in May, first broods generally did well, but later breeding attempts found that invertebrate supplies as required by parents feeding chicks reduced significantly. Consequently, fewer second clutches/broods are expected, which has meant that reduced numbers of active nests were in the bracken when contractors started work on cutting/rolling bracken on 24th June 2018.

Survey results

Bird species seen within bracken, or using adjacent ffridd habitats, are tabulated below in separate tables according to their habitat preference – note some species are shown under both habitats, but an asterisk * indicates its likely preferred nest location. Those species featured and asterisked in the ‘pure bracken’ column nest on or close to the ground in decaying bracken trash, and are therefore more vulnerable to bracken cutting before the end of the breeding season. The British Trust for Ornithology code is also shown (this being used in Table 3 and 4) and the current population trends to reflect levels of decline in the period 1995-2016. Red reflects those species in most severe decline, amber indicates some decline, and green as stable or increasing.

Table 2: Species found in each habitat, with their preferred nesting choice:

Pure bracken sites Ffridd habitats RED Listed species RED Listed species Tree Pipit (TP) * Tree Pipit (TP) Skylark (S.) * Yellowhammer (Y.) * Whinchat (WC) * Linnet (LI) * Cuckoo (CK) * Cuckoo (CK) Yellowhammer (Y.) Mistle Thrush (M.) * Linnet (LI) AMBER Listed species AMBER Listed species Dunnock (D.)* Meadow Pipit (MP) * Willow Warbler (WW) Dunnock (D.) Redstart (RT) * Reed Bunting (RB) * Bullfinch (BF) * Willow Warbler (WW) *

GREEN Listed species GREEN Listed species Stonechat (SC) * Stonechat (SC) Wren (WR) * Blackbird (B.) * Pied Wagtail (PW) Robin (R.) * Woodpigeon (WP) * Garden Warbler (GW) * Chaffinch (CH) * Carrion Crow (C.) * Buzzard (BZ) * Blue Tit (BT) * Wren (WR)

Each treatment site visited is listed below, summarising the condition of bracken, and those ground-nesting species seen within the pure bracken habitat:

Table 3: Bracken characteristics and species found using pure bracken, by site

Plot Location Visit Litter Species in Nests? Comments date depth bracken 2 Trevelog 14/05/18 Shallow - No No ground- nesters 6 Cadwgan 02/06/18 Medium MP/CK MP Highest MP numbers 8 Llwynbrain 23/05/18 Shallow - No History of bracken cutting? 9 Cwm Milaid 23/05/18 Deep MP/SC/WW/TP 18 Blaenffirnant 28/05/18 Medium WW No History of stock- feeding? 21S Blaenau Uchaf 08/05/18 Patchy Y./PW No 3 Cwm Buchel 07/05/18 Medium WC/MP/RB No & 29/06 23 Grange 14/05/18 Deep MP/TP No Banks & trash = increased chance of nests 25 Llewenau 30/05/18 Shallow MP/CK/TP/WW No Draw 1 Blaenyoy 21/05/18 Medium WC/TP/Y. TP Banks & trash = increased chance of nests 16 Cwmrhos 30/05/18 Patchy Y./S./WW/MP S. 19 Grafog 28/05/18 Shallow WC/TP/WW WC 21N Penybryn 08/05/18 Medium MP/TP/Y./WC/S. No

Similarly for ffridd, the following table illustrates those species seen at each location most likely breeding in or close to adjacent ffridd. Many were noted singing or ‘on guard’ in single or groups of trees/bushes within the bracken treatment area. It is therefore likely that these birds were spending at least part of their time foraging in bracken destined for cutting/mulching, as well as convenient branches in the ffridd to act as song-posts. No time was devoted to searching for nests in ffridd, or within ten metres of ffridd habitats.

Table 4: Ffridd characteristics, and species found using ffridd habitats, by site

Plot Location Visit date Ffridd Species in Comments present ffridd 2 Trevelog 14/05/18 Minimal B./BZ/RT 6 Cadwgan 02/06/18 Medium WC/SC/LI/ Good quality ffridd to W WW/RT/GW and N of Nant Bran gully 8 Llwynbrain 23/05/18 Minimal WC/S./CK/ WW/SC/MP/ TP/CH 9 Cwm Milaid 23/05/18 Gorse WR/BF/BZ/ Ribbons of ffridd. Most WC patches C./WC at east end. 18 Blaenffirnant 28/05/18 Degraded CH/RT/WR 21S Blaenau 08/05/18 Degraded CH/RT/WR/ Uchaf CK 3 Cwm Buchel 07/05/18 Minimal LI/CH/GW/ Good quality ffridd on steep & 29/06 WP/WW/TP bank below trt area 23 Grange 14/05/18 None 25 Llewenau 30/05/18 Patchy B./CK/Y./WR/ Draw R. 1 Blaenyoy 21/05/18 Yes WC/TP/Y. Ribbons of good quality ffridd, and bilberry 16 Cwmrhos 30/05/18 Patchy RT/B./CH/R. Ffridd on west side 19 Grafog 28/05/18 Patchy RT/B./BT/CH/ Y./WC 21N Penybryn 08/05/18 Patchy CH/RT/M./B./ Good quality, esp close to WR/R. Penybryn boundaries

Observation and search-time constraints within the methodology did not permit a full nest search of species breeding in the study plots, from which an average breeding density could be calculated. However, the following ranking can be given for the frequency (> one contact) by which species were flushed from pure bracken habitats within the one or quarter-hectare study plots:

Meadow Pipit > Tree Pipit >Willow Warbler >Whinchat >Yellowhammer Conclusions

While the overall number of species found using bracken were to be as predicted, the number of birds seen was lower than expected. This was partly a function of the variability of the bracken sites chosen for treatment (some were richer than others), but also a poor and late return to the hills for several of the common resident species. It may also have been a consideration and objective at the project planning stage to select plots for ground treatment that were likely to be less attractive to breeding birds. This appears to have been relatively successful.

Two periods of severe cold during March (‘Beast from the East’ and ‘Mini-Beast’) coupled with a wet end to the winter appeared to reduce numbers of birds returning to breeding sites. Meadow Pipit and Skylark, always likely to be the most numerous on bracken and bracken- fringe areas, appear to have suffered over-winter mortality, or to have been in poor condition. Migrant species also returned slowly, but by the last week of May nesting behaviour was indicated by both residents and migrants alike.

Those sites with a history of bracken-cutting for bedding or feeding stock over-winter (plots 2,8,16 and 18), tended to have thinner bracken (over grass) with minimal bracken trash beneath. Consequently, few birds were seen feeding or flushed, and even fewer found these conditions suitable for breeding.

Timing of optimum bracken control varies according to the methods selected, but all depend on weakening the plant, and therefore vigour, by allowing it to expend resources into developing top-growth which is then damaged by rolling or cutting. Rolling was the suggested method at several sites, this ideally occurring during rapid stem extension suggesting the first half of June, and therefore having the greatest potential conflict with breeding birds. Cutting should be carried out at or close to full frond, whereby the plant had completed formation of top growth using its stored resources, but yet to receive any benefits back to the underground rhizomes. It was noted during the final survey visit on 29th June, that the last leaflets on the top frond were unfurling.

During ornithological survey, only four nests were found on or close to measured study plots, of which three were in sites where bracken litter had progressively built up (see six photographs of three nests in Appendix 2). Bracken litter provides some protection for nest contents and a position just a few centimetres off the ground, particularly important for first clutches to raise them off a wet ground surface. The nests were not followed up to check for success, nor to determine whether the chicks had fledged prior to bracken treatment. By reference to the date and contents at nest discovery, and the date on which contractors started bracken treatments (cutting and rolling), the period was probably sufficient in length, but the margin was tight. Those nests found, all with eggs, suggest a first-egg date ranging from 20th - 30th May. Assuming a typical requirement for upland passerines of 27-31 days in which to complete incubation and feeding of chicks in the nest, bracken treatment should only be considered from 1st July. Accepting that 2018 has been a late breeding season, contractors starting work from 24th June will have put a small number of nests (first broods) at risk, but to have been well ahead of birds building nests for a potential second clutch. The dry conditions and evident difficulty that some species have experienced in finding sufficient invertebrates to feed chicks will probably limit the take-up of second broods in 2018. In consideration of these events, and the bird species mix and bracken characteristics at each site, recommendations were forwarded to the BBNP Ecology team on 1st July to suggest that timing of control treatments should be split into 10-day periods throughout July.

Discussion, and Considerations for Years 2 and 3

The project proposes that bracken control by these various ground treatments are repeated on the same areas again in 2019 and 2020. Assuming that treatment on these plots in 2018 has weakened the plants, as well as some deterioration in the depth and structure of bracken litter by the passage of vehicles, etc the requirement to assess potential for breeding birds in spring 2019 may not be necessary. However, the Method Statement provided to contractors in 2018 should again be employed in 2019. This has included an initial observation period such that activity of birds likely to be breeding on site can be investigated and potential effects on them mitigated.

Notwithstanding that 2018 has been a late season for both breeding birds and bracken development, it is recommended that 1st July is used as an effective start date for ground treatments in subsequent years. A completion window of 33-35 days should be given for the likely range of breeding passerines in bracken habitats in the Black Mountains. Of the species seen using bracken areas in 2018 (albeit at very low occurrence), the Cuckoo has the longest requirement of ca. 45 days from egg-laying (most likely within a Meadow Pipit nest) to the chick fledging and moving clear of the host nest. 1st July is considered to offer the best balance between completion of first broods and effectively controlling bracken shortly after frond emergence is complete.

Any new areas selected for bracken treatment by ground method should undergo some pre- breeding ornithological assessment. Not only is this advisable for legal reasons, but 2018 experience has shown there is considerable site-by-site variation of bracken characteristics and bird species mix. The possibility of selecting new areas that are of importance for Red- listed species, or indeed in which a rare or protected species may be breeding (or may arrive to hold territory in that year) should not be discounted. There has been a history of rare species breeding in the Black Mountains, most recently in ffridd-type habitats.

As part of any 2018 Pause and Review process that the project management team may undergo, the accuracy and appropriate scale of initial mapping of treatment areas should be considered. My own casual observations undertaken during other work in the Black Mountains since bracken treatments were completed, suggests that bracken treatment has extended beyond mapped boundaries. Whether these were planned, accidental or by local negotiation once contractors arrive on site, they could lead to potential difficulties. Not only are breeding birds put at some risk (as no pre-breeding assessment will have taken place on the added area), but it puts at risk the good objectives and planning of the project and open it up to adverse comment.

References

Harris, SJ et al. 2018. The Breeding Bird Survey 2017. BTO Research Report 706, BTO, Thetford.

Appendix 1:

Locations of all sub-plots on which breeding bird assessments have been made:

Plot Location Date Sub-plot Central GPS grid ref. No. size 1 Blaenyoy 21.5.18 1 ha SO30617 24958 SO30701 25010 SO30809 25054 2 Trevelog 14.5.18 0.25 ha SO28125 29583 SO28132 29502 SO28237 29652 3 Cwm Buchel 29.6.18 0.25 ha SO27600 27013 SO27598 27081 SO27706 27227 SO27819 27365 5 Ffawydd 2.6.18 0.25 ha No assessments made 6 Cadwgan 2.6.18 0.25 ha SO27603 25277 SO27620 25162 SO27649 25042 SO27677 24919 8 Llwynbrain 23.5.18 0.25 ha SO27078 23461 SO27180 23474 SO27498 23250 SO27598 23096 SO27623 23024 9 Cwm Milaid 23.5.18 0.25 ha SO25143 22347 SO25256 22656 SO25319 22668 SO25610 22619 SO25708 22625 16 Cwmrhos 30.5.18 0.25 ha SO19229 24938 SO19291 25022 SO19316 24953 18 Blaenffirnant 28.5.18 0.25 ha SO19131 28339 SO19248 28375 SO19387 28396 19 Rhydycar 28.5.18 0.25 ha SO18387 30292 SO18468 30403 SO18534 30458 21(S) Blaenau Uchaf 8.5.18 0.25 ha SO18830 31107 SO18919 31100 SO19060 31322 21(N) Penybryn 8.5.18 1 ha SO19061 32385 SO19370 32500 SO19401 32694 SO19489 32870 23 Grange 14.5.18 0.25 ha SO25039 31190 SO25083 31096 25 Llewenau Draw 30.5.18 1 ha SO19175 23926 SO19205 23787 SO19301 23638

Appendix 2 (six photographs of nests and nest contents)

These are of three (of four) nests located.

Whinchat nest, above Grafog sub-plot, 28th May 2018

Whinchat clutch, above Grafog sub-plot 28th May 2018

Skylark nest location, Cwmrhos 30th May 2018

Skylark nest, Cwmrhos, 30th May 2018

Meadow Pipit nest location, Cadwgan, 2nd June 2018

Meadow Pipit nest (partly laid clutch), Cadwgan, 2nd June 2018

20th July 2018

Andrew King, AvesEnto Surveys

Heddfan, Pennorth, Brecon, , LD3 7EX