Breeding Waders of English Upland Farmland (BWEUF): survey and data analysis for breeding waders on in-bye land
Report to Natural England
Gavin Siriwardena, Greg Conway (British Trust for Ornithology)
Andrew Stanbury, Mark Eaton (Royal Society for the Protection of Birds)
Summary
1. Building on the significant conservation concern for waders breeding in in-bye farmland in England, this project assessed the importance of in-bye land and the management thereof under agri-environment schemes (AESs) for these species, using a combination of analyses of existing data sets and a new survey of the relevant habitats in 2016. The focal species were Curlew Numenius arquata, Golden Plover Pluvialis apricaria, Lapwing Vanellus vanellus, Oystercatcher Haematopus ostralegus, Redshank Tringa totanus and Snipe Gallinago gallinago. 2. Throughout this project, an operational definition that ‘in-bye’ farmland consisted of all agricultural land below and within 1km of the ‘moorland line’ (latter as delimited by Natural England) was used. ‘In- bye tetrads’ (2×2km grid squares) are nominally those with greater than 20% cover of in-bye farmland. 3. The strongest historical analysis available made use of data from Bird Atlas 2007-11 (Balmer et al. 2013) and changes in distribution from the data collected under the previous atlas project (Gibbons et al. 1993). Analyses of the influence of AES management on the local colonization and local extinction of the target species in in-bye 2×2km tetrads throughout England were conducted. 4. Despite presence-absence analyses providing only coarse measures of population change, there were several significant relationships between AES management and the target species. Overall, Environmentally Sensitive Area (ESA) management tended to be associated with negative changes (local extinction or lack of colonization), whereas Environmental Stewardship (ES)/Countryside Stewardship Scheme (CSS1) management was associated with positive effects. The negative ESA effects in respect of colonization prevailed for all species except Redshank, while ES and CSS showed significantly positive associations for Curlew, Lapwing and Snipe. Local extinction was apparently promoted by ESAs for Curlew and Lapwing, but restricted by ES/CSS for Curlew. It is important to note that all these results are correlative and that the associations do not prove either positive or negative effects of the different AES schemes. 5. A wide range of other data sources that potentially inform about changes in in-bye wader populations were collated, totalling 105 datasets, but the disparate survey/recording methods used and limited recording of survey effort in space and/or time in some surveys limited their utility for investigating long-term trends and the effects of management upon them. These data covered many local areas and discrete time periods within the range from the early 1980s to the present. As a compromise among the range of data resolutions available, data were pooled at the 2×2km tetrad level. Where they were not explicitly recorded, zero records for species were inferred where possible; data providing presence information only were discarded. Analyses then proceeded using models of presence-absence by year at the tetrad scale: the disparate survey methods used made comparing counts unreliable. These considered both temporal trends and local colonization/extinction.
1 As distinct from the latest AES in England, which is similarly named ‘Countryside Stewardship (CS)’.
6. Final analyses used data from 1994 onwards and were dominated by BBS data because the quality of the other data sources for the present analysis was often low. The data were less standardized than Atlas data, but revealed some of the same patterns in respect of local colonization or extinction. These analyses therefore add little to the inference already gleaned from Atlas data. Analyses of the variation in linear trends in presence over time with respect to AES management also produced little evidence for important patterns not detected by the Atlas analyses. 7. Novel population trends based on these presence-absence analyses revealed long-term declines in Curlew, Lapwing and Redshank, but increases in Oystercatcher and a noisy pattern, but overall stability, for Snipe. These patterns are similar to those reported for these species for terrestrial habitats in England by the BBS. 8. A new survey of waders in in-bye farmland was designed, set up for online coordination and data inputting, and conducted during 2016. The survey unit was the tetrad and candidates for coverage were selected with respect to recorded counts of the target waders in Bird Atlas 2007-11, with the rule that at least 80ha of the tetrad had to be ‘in-bye farmland’. Tetrads were selected for survey randomly, within strata based on Atlas counts. Coverage was then split between professional surveyors employed by RSPB and volunteers, such that a total of 522 of the total of 2837 candidate tetrads identified in England were surveyed, with a bias towards areas with larger wader populations. Tetrads were visited twice in Spring 2016, with birds and habitat being recorded by field within the in-bye area and controls in place to avoid bias due to partial coverage. 9. Data were analysed to address: (i) the abundance and distribution of the target species in in-bye habitats and how these relate to recent estimates for the total populations in England; (ii) how the patterns of distribution and abundance of the target species relate to those of key current and historical AES options. For testing, AES options were grouped into non-exclusive categories describing broad forms of management and/or nominal effects on waders: grazing, restoration, nesting, feeding, nesting and feeding, and hay. Tests were also conducted using total areas of ESA and ES/CSS. 10. Significant numbers of breeding pairs of all the target species were found in in-bye fields, with the exception of Golden Plover, but the latter species was frequently recorded foraging in in-bye areas. An estimated 14% of English Redshanks and 28% of Oystercatchers are found in in-bye, reflecting their primarily coastal distributions, whereas in-bye supported 48-67% of Curlew, Lapwing and Snipe, showing the importance of the habitat for these declining species. Estimates of the total populations found in in-bye land for each species were as follows (median number of pairs and 5th-95th percentile range): Curlew 15039.5 (10551.5 - 20747.7), Golden Plover 173.5 (28.8 - 440.9), Lapwing 27243.4 (18143.6 - 37917.9), Oystercatcher 6828.1 (4043.9 - 10603.7), Redshank 1604.4 (628.1 - 3092.2) and Snipe 4527.5 (2064.1 - 8090.6). 11. In analyses of associations between waders and AES management at the tetrad scale, there were significant, positive results for all of the target wader species, except Golden Plover. The patterns were especially strong for Curlew and for options for grazing and restoration, among the management types considered. The only negative association found was for Curlew and ESA management, which could show a failure of management or independent associations between particular regions and each of ESA management and wader abundance. The positive associations found suggest that AES management in the grazing, restoration and, to a lesser extent, feeding categories could be widely beneficial across species, but it should be noted that these associations are strictly correlative and do not necessarily imply that AES management has driven the variation in abundance. In particular, such patterns could arise through the targeting of AES towards existing areas of high wader density. 12. Analyses of associations between waders and AES management within tetrads show positive associations for all species except Golden Plover, reflecting the limited use of in-bye habitat only for foraging by this species. Curlew adult counts showed a significant negative association with ESA management. There were no significant associations for breeding pairs with hay. Management directed
at restoration/maintenance and provision of feeding and nesting habitat generally has significant positive associations for all species except Golden Plover, both for breeding pairs and total adult counts. 13. Analyses of associations between waders and the key features of in-bye field habitats mostly reflect well-established patterns in wader ecology, such as the avoidance field boundaries, preference for heterogeneous vegetation and reliance on wet ground. Variation in preferences for specific field types across species and between measures of feeding versus nesting birds reveal a diversity in species’ ecologies that underlines the importance of habitat heterogeneity at the field or landscape scale to support the entire wader community. 14. This study has confirmed that in-bye farmland is important for several wader species, with consistent results from long-term changes in distribution and fine-scale distribution in 2016, both at the tetrad scale and within tetrads, although populations continue to fall. There is evidence that conservation action via ES/CSS may be having positive effects, but ESAs are associated with negative effects. The causality of these correlative patterns with AES management remains uncertain. It would be valuable to repeat the survey conducted in this study in around five years to investigate the influence of management on population changes and to allow more confident inference about the causality of the changes observed.
Introduction
A number of breeding waders have been identified as being of significant conservation concern in England, and the UK more widely, as a result of population decline and range contraction (Hayhow et al. 2015, Robinson et al. 2016, Balmer et al. 2013). Recent monitoring and research effort has focused on waders in moorland and the remaining lowland wet meadows, which are generally now protected sites or nature reserves (e.g. O’Brien & Wilson 2011, Smart et al. 2014). However, a key habitat that has seen little attention is the enclosed farmland immediately below the moorland, which is commonly referred to as “in-bye” land (although note that this term technically refers to all enclosed farmland, and can also refer to land adjacent to farmsteads). This habitat may support important populations of waders, but is not sampled effectively by the BTO/JNCC/RSPB Breeding Bird Survey and has not been the subject of recent bespoke survey effort. There is therefore a need to establish the status of the key wader species in in-bye and to assess population trends, if data permit. There has been a range of local and small-scale or short-term survey activity recording waders in relevant habitats, which may be combinable into larger scale datasets suitable for measuring population change, but the quality of potential inference is dependent upon consistency in survey and reporting protocols.
This project assessed the importance of in-bye land for threatened waders by estimating the total numbers present in these habitats, relative to national populations measured from Bird Atlas 2007-11 data, while also measuring differences in numbers between Agri-environment Scheme (AES) and non-scheme-managed habitats, in order to assess the value of AES management for breeding grassland waders that nest and/or forage on in-bye farmland in the English uplands. The study focused on Curlew Numenius arquata, Lapwing Vanellus vanellus, Oystercatcher Haematopus ostralegus, Redshank Tringa totanus, Snipe Gallinago gallinago and Golden Plover Pluvialis apricaria. The aim was to indicate the extent to which AES management protects these target species in these habitats and/or is selected by them, noting that positive results may also reflect the targeting of AES management at habitats that are good for waders.
The project was divided into two parts: first, the available historical and current data on wader abundance and presence were collated and analysed, including a new analysis of Bird Atlas 2007-11 data and changes since the previous Atlas (Gibbons et al. 1993); second, a new volunteer-based survey of in-bye land was conceived, designed, undertaken and the resulting data were analysed. Both parts of the project considered the status of wader populations in in-bye habitat and investigated the impacts of AES management on population and distribution change.
Part 1 Analyses of historical data
1.1 Analyses of Bird Atlas 2007-11 data
1.1.1 Methods
Historical data with known, controlled and comparable survey methods were available from the Bird Atlas projects in 1988-91 (Gibbons et al. 1993) and 2007-11 (Balmer et al. 2013). Although standardized abundance data were collected during the more recent Atlas project, only presence-absence data were available for waders from the earlier project, so only a presence-absence-based comparison could be made. Hence, data on Lapwing, Curlew, Oystercatcher, Redshank and Snipe were extracted from both Atlases for 2×2km tetrads with more than 80ha of “in-bye”, defined as the area of farmland (separated from urban, woodland open water using the Centre for Ecology and Hydrology Land Cover Map 2000) within 1km below Natural England’s delimitation of moorland.. Where tetrads had been surveyed for both Atlases, one-hour timed tetrad visit data were used to identify presence and quasi-absence (it is inevitable that some actual presences will have been missed during these short survey visits, but birds are likely to have been more numerous, on average, in tetrads where presence was confirmed).
The data were analysed using three model structures: extinction (present in 1988-91 and present or absent in 2007-11), colonization (absent in 1988-91 and present or absent in 2007-11) and environmental improvement (local loss, local maintenance and local colonization expressed as an ordinal variable with levels 1-3, respectively). For each model structure, the influence of agri-environment scheme (AES) management on distributional change was investigated by using the area of relevant options in each tetrad as a predictor variable, thus testing whether the persistence or spread of the target species’ distributions was associated with the area of AES coverage in a tetrad. Tests were conducted separately for Environmentally Sensitive Area (ESA) and Countryside Stewardship Scheme (CSS)/Environmental Stewardship (ES) management. These data were supplied by Natural England from historical outputs from the Genesis database for 2005, 2009 and 2013. For ESAs, both total area of ESA agreement overlapping with focal tetrads and the area of wader-relevant management of in-bye in focal tetrads2 were considered. The maximum area under ESA agreements in the Genesis outputs for 2005, 2009 and 2013 was used to represent ESA quantity in each tetrad. ESA options are not recorded or defined in as much detail as those in newer AESs, so the wader-targeting options used here may be a conservative selection of potential management impacts. Hence, the total ESA area variable provided a more inclusive approach to assessing potential ESA effects, while the ESA options variable was more specific and avoided the potential for ESA effects to be obscured by the inclusion of options not targeted at waders. For CSS and ES, together, the area of wader-relevant options in focal tetrads that could be applied to in-bye was summed within and across the years during which each agreement was in place, such that the total quantity of AES management considered reflected both the areas involved and the length of time over which they had been in place. This
2 Defined as agreements including options 1BB, 1BC, 1BM, 1CG, 1DR, 1DU, 3BR, 4CW, BWS, EGS, MCS, MMS, O1A, O1B, O1C, O1D, O2B, OO2 and OO3; note that option data were not available for older agreements reported only in 2005. was done across all years for both schemes (2002-2013) and post-2005 alone (to focus on ES management, which might nominally be expected to be more effective).3
The analyses were conducted using logistic or ordinal logistic Generalized Linear Models in SAS (www.sas.com), assuming binomial or multinomial errors. Logistic (“yes/no”) models were used for colonization and extinction, while ordinal logistic models were used for the three levels for environmental improvement. The area of “in-bye land”, as defined elsewhere in the project, was included in all models to account for the bias that more management associated with in-bye land is expected to be found where there is more in-bye land.
1.1.2 Results4
Presence-absence is a coarse measure of population change: records of “presence” may conceal considerable variation in local abundance, so local increases or declines in density cannot be ruled out. Moreover, Atlas field surveys have been designed to reveal complete species lists as closely as possible within a timed visit (Balmer et al. 2013), so actively aim to detect species that are at low abundance. Therefore, species could decline or increase a great deal at a local scale without any change being registered in the available data. This means that the power of the tests of Atlas tetrad changes might be expected to be low. However, there were good numbers of significant associations between areas of ESA or ES/CSS management and the three metrics of change in wader distributions that were considered (Table 1.1). Overall, ESA management tended to be associated with negative changes (local extinction or lack of colonization), whereas ES/CSS management was associated with positive effects (Table 1.1).
Specific results showed predominantly significantly negative associations between ESAs and colonization for all species except Redshank, whereas ES and CSS showed significantly positive associations for Curlew, Lapwing and Snipe (Table 1.1).
ESAs were significantly associated with local losses of Curlew and Lapwing, but there was some evidence that losses were prevented for Snipe (ESA option area) and Redshank (total ESA area). Conversely, ES and CSS restricted losses significantly for Curlew, and weakly for Snipe (Table 1.1).
The ordinal measure of general environmental improvement assumes that changes in conditions for the target species improve in equal steps from those that drive local extinction through those that support population maintenance to those that promote colonization, and hence the analyses test the extent to which AES management has achieved this environmental improvement. Therefore, ESAs seem to have driven environmental deterioration for Lapwing, but improvements for Snipe (ESA option area) and Redshank (total ESA area; Table 1.1). The combined effects of ES and CSS across the environmental improvement gradient were positively associated with Snipe presence, but there was no effect on the other species (Table 1.1).
It is important to note that all these results are correlative and that the associations do not prove either positive or negative effects of the different AES schemes. However, it may be significant that there was a clear contrast between the results found for ESAs and ES/CSS, given that the latter schemes feature more sophisticated management options, as well as specific targeting of management towards breeding waders. It is also noteworthy that positive and negative parameter estimates in these analyses are relative, so a positive colonization effect, for example, does not necessarily imply range increase. Moreover, in the absence of experimental controls, positive or negative associations could reflect the location of management
3 The options considered for CSS were UH1, UP1, UP2, UP3 and UP4; those for ES were HL7, HL8, HK15, HK16, UL20, EL3, UL21, UL23 and UL22. 4 Note that these results differ from those presented in the March 2016 interim report, which were subject to data errors. in areas that are independently associated with local losses or gains of the target species. For example, apparently negative effects of management could occur if it is disproportionately located in areas where declines are occurring and it is either insufficiently effective or a long time lag for positive effects to become detectable is in operation.
Table 1.1. Results summary for responses of metrics of distributional change to ESA and ES/CSS management areas. Note that positive parameter effects indicate potential benefits for colonization and environmental improvement, but negative ones indicate benefits in respect of extinction. Significant results (P<0.05) are shown in bold.
Response type Parameter Species Effect SE Wald Chi- P Estimate Square (1df) Colonization ESA option area Curlew -0.1615 0.1548 1.09 0.297 Lapwing -1.6124 0.3033 28.25 <0.0001 Oystercatcher -0.9991 0.2129 22.03 <0.0001 Redshank -1.4492 0.9036 2.57 0.109 Snipe -0.4103 0.2447 2.81 0.094 Total ESA area Curlew -0.0071 0.0024 9.08 0.003 Lapwing -0.0200 0.0040 25.56 <0.0001 Oystercatcher -0.0096 0.0026 13.57 <0.0001 Redshank -0.0034 0.0063 0.29 0.589 Snipe -0.0065 0.0035 3.38 0.066 ES & CS option Curlew 0.4251 0.1887 5.07 0.024 area Lapwing 0.5086 0.1721 8.73 0.003 Oystercatcher 0.1415 0.1242 1.30 0.255 Redshank 0.3734 0.2252 2.75 0.097 Snipe 0.5487 0.1407 15.20 <0.0001 Post-2005 ES & CS Curlew 0.6535 0.3170 4.25 0.039 option area Lapwing 0.8364 0.2812 8.85 0.003 Oystercatcher 0.1596 0.2150 0.55 0.458 Redshank 0.5394 0.4425 1.49 0.223 Snipe 0.8867 0.2237 15.71 <0.0001 Extinction ESA option area Curlew 0.2787 0.1090 6.53 0.011 Lapwing 0.7561 0.1608 22.10 <0.0001 Oystercatcher 0.2765 0.2639 1.10 0.295 Redshank -0.6737 0.4542 2.20 0.138 Snipe -0.7899 0.2642 8.94 0.003 Total ESA area Curlew 0.0039 0.0015 6.71 0.010 Lapwing 0.0068 0.0021 10.49 0.001 Oystercatcher -0.0018 0.0035 0.27 0.603 Redshank -0.0149 0.0067 4.85 0.028 Snipe 0.0005 0.0042 0.01 0.908 ES & CS option Curlew -0.2963 0.1126 6.92 0.009 area Lapwing -0.1307 0.1074 1.48 0.224 Oystercatcher 0.0448 0.2079 0.05 0.829 Redshank -0.2814 0.3211 0.77 0.381 Snipe -0.4114 0.2256 3.32 0.068 Post-2005 ES & CS Curlew -0.4811 0.1783 7.28 0.007 option area Lapwing -0.2981 0.1791 2.77 0.096 Oystercatcher 0.0426 0.3394 0.02 0.900 Redshank -0.6898 0.5515 1.56 0.211 Snipe -0.6224 0.3564 3.05 0.081 Environmental ESA option area Curlew -0.0618 0.0930 0.44 0.506 improvement Lapwing -0.6660 0.1421 21.97 <0.0001 Oystercatcher -0.3399 0.2024 2.82 0.093 Redshank 0.4360 0.4260 1.05 0.306 Snipe 0.4849 0.2002 5.86 0.015 Total ESA area Curlew -0.0022 0.0013 2.81 0.093 Lapwing -0.0056 0.0018 9.33 0.002 Oystercatcher -0.0004 0.0027 0.02 0.877 Redshank 0.0158 0.0054 8.44 0.004 Snipe 0.0010 0.0025 0.15 0.698 ES & CS option Curlew 0.0916 0.0915 1.00 0.317 area Lapwing 0.0791 0.0971 0.66 0.415 Oystercatcher -0.1050 0.1660 0.40 0.527 Redshank 0.3476 0.2607 1.78 0.183 Snipe 0.3773 0.1401 7.25 0.007 Post-2005 ES & CS Curlew 0.1381 0.1506 0.84 0.359 option area Lapwing 0.2005 0.1584 1.60 0.206 Oystercatcher -0.2041 0.2718 0.56 0.453 Redshank 0.6600 0.4692 1.98 0.160 Snipe 0.5728 0.2317 6.11 0.013
1.2 Analyses of all available historical data
1.2.1 Methods
In addition to the Bird Atlas datasets analysed as described above, we collated data from 124 historical surveys of breeding waders in relevant parts of England from sources in Natural England and RSPB. Each data set was assessed for quality and relevance with respect to informing about bird abundance/presence, in-bye land and recent AES management. The full inventory of datasets collated and results of the assessment of their value are described in Appendix 1. The data were in a range of formats, and describe a wide range of survey results from the early 1980s to the present. The quality of these data and their suitability for further analysis are highly variable. This is not to pass judgement on these data for their original purpose, just for their relevance for the present study and, in some cases, on the way in which they have been stored, with zero or null records not being clearly identified; consequently only 105 datasets were used. In particular, some data sets are in GIS shapefile format and some are in spreadsheet form, while some consist of counts associated with defined survey areas, and others are location records for individual birds, groups of birds or territories. The latter are more difficult to interpret because the full survey extent (i.e. areas covered and survey areas where zero counts were returned) is not always clear (survey extent listed as not defined in Appendix 1). Bird records have been recorded at a range of spatial resolutions, such as grid references accurate to 100m or just as associations with spatial survey units; for most datasets these are the only data included (i.e. there is no habitat information, for example). Some surveys were focused on waders and others have some wader information collected within more general monitoring or surveys focusing on other species. Habitat/spatial coverage is also variable, with some surveys focused on spatial units like grid squares and others on “plots”, such as habitat patches, or larger areas such as farms/land holdings; the potential for the extraction of data specific to in-bye farmland from each, therefore, varied. As shown in Appendix 1, both the areas covered by individual surveys and our knowledge of those survey areas varied between datasets; no supplementary information revealing the survey areas actually covered was available for certain surveys. The survey areas given in Appendix 1 are mostly indicative and reflect minimum areas covered, but they reflect the uncertainty over survey coverage.
The key challenge with the interpretation of the historical data sources was to find a common “currency” with which to combine or to compare data across datasets, time periods and locations. Given the variability in the spatial resolution of the data, some aggregation of the smaller sample locations was necessary to produce spatial matching over time and to allow datasets with coarser spatial resolution to be compared to those with finer resolution. Further, larger spatial units then allow the combination of more data sources, albeit at the cost of a loss of spatial specificity and potential habitat matching. Ultimately, we decided to pool data at the tetrad scale, because it allowed a wide range of data sources to be combined, as well as potentially permitting the Atlas data sets to be included in the same, single analyses (although, ultimately, we did not do this because the power in the Atlas analysis turned out to be fairly high). In addition, while BBS coverage of in-bye alone does not support analyses, there are some relevant BBS squares, and the annual data from these locations were also included, by assignment to the tetrad in which they fell.
Spatially explicit survey areas were assigned to the tetrads in which they were found, with the total area and counts covered by each being divided pro rata between any two or more tetrads that they overlapped, i.e. assuming that the target species were evenly distributed throughout. This means that counts may actually derive wholly or partly from tetrads adjacent to the focal tetrad to which they were assigned, leading to noise in the data and analyses. Further assumptions had to be made about the spatial coverage of other surveys, notably that Atlas data referred to whole tetrads and that point records referred to coverage of a 100m square defined by the grid reference and its precision. Therefore, further noise in the data will arise if the areas actually surveyed to produce these data were larger or smaller than these assumed units, and bias could be caused if the areas actually covered differ systematically from those in the assumed spatial units
(e.g. if particular habitat types have been selected or avoided). Data sources recorded at 1km square resolution (typically BBS or those using the method of Brown & Shepherd 1992) were assumed to cover the whole of the 1km square indicated, which is implicit in the survey protocols used to generate such data. Survey effort per tetrad per year was then calculated as the total area surveyed by methods with these various approaches and the associated area coverage within the focal tetrad.
A critical factor in the analysis of survey data is the knowledge or reliable inference of zero counts or absences, against which positive counts and recorded presence can be compared. Fundamentally, zeroes are critical for distinguishing evidence of absence from absence of evidence. Where record data do not include zero counts, inference typically requires analytical solutions based on the presence of records of other species to identify where sampling effort has been made but a given species has not been found (e.g. Isaac & Pocock 2015). Most data from organized schemes include records of zero counts, or consist of complete lists of the species detected, such that zeroes for other species can be inferred. However, this does not apply to records of individual birds, or to survey data that are stored as individual point records, as opposed to sample data associated with grid squares or other spatial units. With such data sources here, zeroes could not be inferred for single-species data sets but, where multiple species were recorded as points, records across all species considered in the survey were summarized at the tetrad level, showing tetrads with and without recording effort in a given year, and zeroes for a given species were inferred for the latter tetrads if the species was not among those recorded in the tetrad. Those of the available datasets that were used in the final analysis are indicated in Appendix 1. The final list of species that had sample sizes sufficient to support analyses was Curlew, Lapwing, Snipe, Redshank, Oystercatcher and Golden Plover. However, a large proportion of Golden Plover records in many surveys are likely to relate to birds on open moorland, rather than in-bye, such that they do not provide reliable inference about the target habitat. Analyses of these historical data were therefore not conducted for Golden Plover.
Abundance data potentially provide far greater sensitivity to environmental change than presence-absence data, so it would be advantageous, nominally, to extract count data from those data sets that include such information. However, the use of multiple survey/sampling methods over time and incomplete records of the time spent surveying different areas of habitat (e.g. no record at all for point bird records and no splits with respect to habitat in many grid square data) means that counts from different data sets are neither comparable as raw numbers nor readily convertible into a common currency. Hence, especially given the pattern of survey occurrence over time with data commonly being collected over periods of around three years, there is a strong risk of bias in the average counts available from different time periods. Hence, we decided to limit the analyses to patterns of presence and absence, at the tetrad scale, extending the analyses described in Section 1.1 to incorporate further data sets. This also means that it was not possible to extract a meaningful baseline against which to compare new survey data from 2016 (see Part 2).
Using the principles outlined above, wader presence-absence data were collated, as annual data points per tetrad, for as many in-bye tetrads in England as possible. These data were then analysed using the same approaches used for Atlas data in Section 1.1, with the addition of a weighting factor for survey coverage, such that presence or absence records from tetrads that had received more survey effort were given more weight in the analyses. Specifically, analyses considered (i) from where local populations have been lost and gained, and (ii) associations between local changes and the extent and nature of local AES implementation (considering the same lists of options described in section 1.1.1). These analyses considered whether patterns of colonization/extinction (as modelled with respect to Atlas data in section 1.1) before and after 1995, when ESA establishment was completed, have been influenced by AES management. Secondly, whether AES management has influenced the slope of linear trends in presence over time was investigated by testing for interactions between year (as a continuous variable) and the AES management in the tetrad. A control for the interaction between year and the area of in-bye in each tetrad was also included, because
more in-bye-relevant AES management would be expected to occur by chance where there is more in-bye land, so relationships involving the management could otherwise merely reflect those with the background habitat. The latter model was fitted in a repeated measures framework using generalized estimating equations, accounting for the temporal autocorrelation between the presence records in successive years. All analyses were conducted using SAS 9.4 (www.sas.com).
An additional analysis made possible by the spread of records over the years considered temporal trends in the probability of wader presence. Using an analytical method adapted from that used as standard in indexing trends in abundance using BBS data, logit-linear models were fitted to the annual tetrad-year data from the historical collation process for the five wader species. In each model, tetrad and year were fitted as fixed effects, thus allowing spatial and temporal variation in presence. These models allow analyses with unbalanced data, such as different runs of years being available for different tetrads, to be conducted, but they do not control for the influences of gross changes in sample distribution (such as larger proportions of tetrads in the south in one part of the time series), which could then cause bias by confounding spatial and temporal variation in populations. To contribute to a temporal trend, tetrads must vary in occupancy status between the years for which data are available, so those in which a species was always present or always absent were omitted.
Given that the data had to be limited to simple presence-absence records at a large spatial scale, it was not appropriate to attempt to model more subtle influences on populations, such as other habitat influences or biological interactions.
1.2.2 Results
The numbers of tetrads ultimately available and used in analyses for each species are shown in Table 1.2. There were large numbers of individual tetrads with potentially relevant data, but many data were not suitable for use in formal analyses. Data quality with respect to long-term, large-scale monitoring was, ultimately, rather low in respect of many of the data sets considered, either because of lack of temporal replication or because there was no temporal variation in presence-absence among the years for which data were available for specific sites. This means that the final annual trends estimated refer only to 1994 onwards (see below): none of the earlier data were suitable. However, the data recording presence in tetrads in all years for which data are available for a tetrad remain useful for comparison with those locations from which birds have disappeared, in the extinction analyses.
Table 1.2. Sample sizes for analyses using collated historical data
Total tetrads with Tetrads for colonization/ Tetrads for temporal records of the species extinction analyses trend analyses Curlew 691 20/160 170 Lapwing 716 35/148 195 Oystercatcher 523 67/47 154 Redshank 535 76/41 88 Snipe 594 63/79 184
The test of colonization and extinction patterns with respect to AES management revealed only one significant and three near-significant results, involving negative associations between Snipe and Lapwing colonization, and ESA management (Table 1.3). This is consistent with two of the patterns identified using Atlas data above, but the others were not detectable using this analysis. It should be noted that power was higher in the analyses of Atlas data and that the standardized data collection approach in the Atlas data sets means that there is much less potential for spatio-temporal biases to affect the analyses. Therefore, the analyses of collated data are likely to be less reliable and should not necessarily be interpreted as superseding the Atlas ones. However, the ESA results in particular could be more informative on distributional change effects because the dividing year around which colonization and extinction were defined was targeted at the inception of the scheme. Thus, negative associations between Curlew and ESAs found in the Atlas analyses were less well supported, but those for Lapwing and Snipe were reinforced.
Table 1.3. Results summary for responses of metrics of distributional change, from historic surveys and BBS data, to ESA and ES/CSS management areas, pre and post ESA establishment. Note that positive parameter effects indicate potential benefits for colonization and environmental improvement, but negative ones indicate benefits in respect of extinction. Significant and near-significant results (P<0.05) are shown in bold.
Response Parameter Species Effect SE Wald Chi- P type Estimate Square (1df) Colonization ESA option area Curlew 0.0326 0.1139 0.08 0.775 Lapwing -0.1047 0.0568 3.4 0.065 Oystercatcher 0.0015 0.0438 0 0.972 Redshank -0.0328 0.0597 0.3 0.582 Snipe -0.3151 0.1379 5.22 0.022 Total ESA area Curlew 0.0002 0.0008 0.1 0.753 Lapwing -0.0016 0.0008 3.7 0.054 Oystercatcher -0.0005 0.0005 1.32 0.250 Redshank -0.0008 0.0008 1.09 0.297 Snipe -0.0048 0.0026 3.39 0.066 ES & CS option Curlew 0.0109 0.1982 0 0.956 area Lapwing 0.0263 0.0819 0.1 0.748 Oystercatcher -0.0095 0.0432 0.05 0.827 Redshank -0.0207 0.0614 0.11 0.736 Snipe 0.0529 0.0514 1.06 0.303 Extinction ESA option area Curlew 0.1626 0.1217 1.78 0.182 Lapwing -0.0873 0.1415 0.38 0.537 Oystercatcher 0.2617 0.2103 1.55 0.213 Redshank -116.933 276508 0 1.000 Snipe -0.0049 0.0971 0 0.960 Total ESA area Curlew 0.0015 0.0014 1.13 0.288 Lapwing 0.0009 0.0009 0.86 0.354 Oystercatcher 0.0564 195217 0 1.000 Redshank -1.5091 3028.18 0 1.000 Snipe 0.001 0.0008 1.43 0.232 ES & CS option Curlew -0.4584 0.9711 0.22 0.637 area Lapwing -0.0099 0.097 0.01 0.919 Oystercatcher -4.349 8.8123 0.24 0.622 Redshank 0.0152 0.0938 0.03 0.871 Snipe 0.1101 0.1097 1.01 0.316
The present data offer more new potential inference over and above Atlas data when inter-annual variation is considered. The limitations of data quality meant that these analyses could only consider the period from 1994 onwards. This timing reflects the dominance of the BBS in the final dataset considered, so the analyses are best viewed as involving BBS data enhanced with the other available, relevant data. Nevertheless, the analyses of the variation in linear population trends with respect to AES management revealed a few significant relationships across species and AES variables (Table 1.4). There were negative relationships between each of Lapwing and Oystercatcher presence trends and ESA management, probably reflecting the same pattern revealed for the colonization and extinction analyses above for both the collated data and Atlas data. Similarly, negative relationships for Curlew and Oystercatcher with total ESA area (Table 1.4) reflect some of those found in the Atlas analysis (Table 1.1). The lack of any effect of ES/CSS option area could result from the use of linear time trends, if ES/CSS management affected only a part of the time period considered: if the management were effective, we would expect a change in trend direction and, by definition, a non-linear trend.
Table 1.4. Results summary of temporal trends in the probability of wader presence, from historical surveys and BBS data, to ESA and ES/CSS management areas.
Parameter Species Estimate Standard Wald Chi- P Error Square (1df) Year × ESA option area Curlew -0.0005 0.0816 -0.01 0.9947 Lapwing -0.2314 0.1233 -1.88 0.0605 Oystercatcher -0.419 0.1119 -3.74 0.0002 Redshank -0.1372 0.2176 -0.63 0.5285 Snipe -0.1321 0.1131 -1.17 0.2425 Year × Total ESA area Curlew -0.0023 0.0011 -2.15 0.0314 Lapwing -0.0039 0.0016 -2.41 0.0161 Oystercatcher -0.0053 0.0013 -4.08 <.0001 Redshank -0.0008 0.0033 -0.24 0.8123 Snipe -0.0026 0.0019 -1.38 0.1664 Year × ES & CS option Curlew -0.0113 0.0508 -0.22 0.8244 area Lapwing 0.0251 0.0295 0.85 0.3945 Oystercatcher 0.015 0.0535 0.28 0.7797 Redshank 0.1076 0.1634 0.66 0.5104 Snipe 0.0021 0.043 0.05 0.9615
Clearly, considering long-term population changes only as linear trends has the potential to be misleading and is often inaccurate, but the patterns revealed by the simple trend analysis of changes over time in the probability of presence in in-bye tetrads lend support to this simplification: all appear rather linear (Figure 1.1; note that this does not suggest that trends in abundance have necessarily been linear). There were clear declines in Curlew, Lapwing and Redshank presence whereas Oystercatcher presence increased and Snipe appear to have remained more stable. There is considerable scatter around the broad temporal trends, probably reflecting spatial changes in the samples for each species over time.
Figure 1.1. Temporal trends in probability of presence within tetrads for five breeding wader species, from historic surveys and BBS (1994 to 2015) data. Note that removal of tetrads with constant occupation status meant that no data from before 1994 could contribute.
a) Curlew b) Lapwing 1.0 0.9
0.9 0.8
0.8 0.7
0.7 0.6
0.6 0.5
Probability of presenceProbability of 0.5 presenceProbability of 0.4
0.4 0.3 1995 2000 2005 2010 2015 1995 2000 2005 2010 2015
Year Year
c) Oystercatcher d) Redshank 0.6 0.9
0.5 0.8
0.7 0.4 0.6 0.3 0.5 0.2 0.4
Probability of presenceProbability of presenceProbability of 0.1 0.3
0.0 0.2 1995 2000 2005 2010 2015 1995 2000 2005 2010 2015
Year Year e) Snipe 0.6
0.5
0.4
0.3
Probability of presenceProbability of 0.2
0.1 1995 2000 2005 2010 2015 Year
Part 2 Survey of waders in in-bye farmland in England
2.1 Methods
2.1.1 Survey site selection
A survey aiming to reveal the fine-scale distributions and patterns of relative abundance of breeding waders in in-bye farmland in England was conducted in 2016. For operational purposes, “in-bye farmland” was again defined as the area of farmland (separated from urban, woodland open water using the Centre for Ecology and Hydrology Land Cover Map 2000) within 1km below Natural England’s delimitation of moorland. During the process of the selection of specific survey areas, this habitat definition was refined after contact from local and National Park staff to reflect the actual location of enclosed land adjacent to moorland areas. Changes were required, in particular, in Northumberland, where unenclosed moorland habitats are found at lower altitudes than further south. The in-bye area, as finally defined, was intersected with the 2×2km tetrads of the national grid across England using ArcGIS 10.3 (www.arcgis.com), revealing 2837 tetrads with at least 20% (80ha) cover of in-bye land. These areas comprised the “population” of locations for sampling from which survey areas were selected and are shown in Figure 2.
Figure 2.1. Locations of all candidate in-bye tetrads: a) all candidate tetrads, grey squares (n=2837), with stratified sample, blue squares (n=1000), and b) survey coverage achieved within stratified sample, red squares=covered and blue squares- not covered, in 2016 (n=522)
a) b)
A subset of 700 of in-bye tetrads was selected at random for priority survey coverage (Figure 2), with stratification towards high-abundance areas of Lapwing and Curlew (which occur mostly in northern England), but also coverage of low and zero abundance areas, which are critical as counterfactuals for comparison with the high-abundance locations. Initial information on abundance was taken from the timed tetrad visit (TTV) data collected as part of Bird Atlas 2007-11 (Balmer et al. 2013). These data are tetrad- specific, but not habitat-specific, so waders in in-bye areas could not be separated from those in other habitats, notably open moorland. This means that some areas of in-bye land are likely to have been included in the survey sample in an unrepresentative stratum, because counts for the tetrads in which they fall reflect populations in moorland (especially), not the target habitat. The counts are also based on short (one-hour) visits to tetrads, so are likely to underestimate total numbers present and to be subject to considerable stochasticity in respect of the counts actually recorded. However, some stratification is important to ensure an efficient distribution of survey effort with respect to the detection of the target species and the Atlas tetrad data are the best data set available for this purpose. Details of the stratification are shown in Table 1. Note that it was important to include a selection of “no atlas data” tetrads in the survey sample because TTV data were not available for all tetrads in the in-bye area and it is possible that the selection of tetrads for survey by Atlas volunteers was biased towards better-quality habitat. The occurrence of target Environmental Stewardship (ES), Countryside Stewardship Scheme (CSS) and Environmentally Sensitive Area (ESA) management (non-zero areas within tetrads; see section 1.1.1 for a list of options) was high within the full selected sample of tetrads (1942/2837), and correspondingly high within the subset selected and made available to surveyors (551/700), as well as that actually surveyed (410/522). The power to detect effects of management via comparison with counterfactual areas (with different quantities of management) should, therefore, be high (dependent upon the actual, biological effects of the management in question).
Table 2.1. Details of the survey square dataset selected by random sampling stratified by Lapwing and Curlew abundance in Atlas TTV data.
Strata Region Total Potential Covered Sample Top 25% Curlew & Lapwing North 162 162 130 Top 25% Curlew North 51 51 37 Top 25% Lapwing North 92 92 76 25%-75% Curlew & Lapwing North 179 113 21 25%-75% Curlew North 320 180 50 25%-75% Lapwing North 278 137 35 ALL Curlew & Lapwing Southwest & West 42 42 34 North 877 81 52 Zero Curlew or Lapwing Southwest & West 416 31 20 North 299 68 53 No Atlas coverage Southwest & West 121 43 14 Totals 2837 1000 522
It is typically difficult to achieve high rates of volunteer surveyor participation in northern and upland areas in England, so this survey was organized with the aim of using a combination of volunteer and professional survey effort, with total estimated achievable survey coverage of 600 tetrads. Of these, 240 were covered professionally by RSPB, with a focus on the areas of highest wader abundance (and therefore also survey square density). The remainder were advertised to volunteers by BTO and RSPB, chiefly through the BTO’s
network of regional representatives and central communication channels. An option for ‘expedition’ surveys for volunteers who do not live near to the target areas was publicized, but achieved no uptake, despite initial interest from some birdwatching groups.
2.1.2 Survey protocol
National grid tetrads were selected as the survey unit because the TTV data that informed sampling were at this spatial scale. These are large areas that could take several hours to survey thoroughly, but many featured far less than complete cover of the habitat of interest, making coverage more tractable. Nevertheless, it was important, especially for volunteers, to allow partial coverage of the in-bye areas within tetrads so that useful data could still be collected where some of the in-bye area could not be visited, either because the in-bye area was too big to cover in the time available, or because access to some of the area of interest was restricted. The survey protocol was, therefore, designed for flexibility, with data collection by field within the designated area of in-bye land (enclosed farmland up to 1km from the moorland line). It is possible that this approach encouraged a bias towards more promising fields for waders, depending on surveyor psychology, but identifying better areas will usually entail some checking of a wider range of areas, and the protocol was designed to make recording basic data about these areas easy and quick, which should have ensured that data on them were collected. On each of two visits to each tetrad (the first between 1st April and 31st May, and the second between 1st June and 15th July, undertaken between 30 minutes post- sunrise and midday), observers were asked to record which fields they visited, and the counts of the target species and of a set of other species of potential interest in each of these fields. They were also asked to record the habitat within each visited field, using a set of pre-defined categories. Coverage of a field nominally consisted of walking within 50m of all points in the field, subject to surveyor discretion given the vegetation present (i.e. shorter vegetation and more favourable topography mean that complete coverage of an area could be achieved by scanning from a greater distance). If access to fields was not available, but some of the areas within were visible from the boundary, observers were asked to record the area that they were able to survey, and the birds and habitats found in that area. Full survey documentation, as provided to all observers, is included here as Appendix 2.
For efficiency of data collection and validation, the survey was set up online (http://www.bto.org/volunteer- surveys/breeding-waders-english-upland-farmland), with respect both to tetrad allocation to surveyors and to data entry. Within tetrads, individual fields within the nominal “in-bye” area were identified, in order to show surveyors the target area for coverage. Provision was included for the collection of records of the target species from non-in-bye areas of tetrads so that surveyors were not asked to ignore such records. Key pages from the online survey form are shown in Figure 3.
Figure 2.2. Representative pages from the online form for the survey: (a) tetrad selection, (b) visit data entry, (c) bird data entry and (d) habitat data entry. a)
b)
c)
d)
Land access for surveyors had the potential to be a significant limiting factor for this survey, so every effort was made to secure landowner contact information and access permissions from local Natural England staff, National Park offices and other contacts known to BTO and RSPB staff to assist surveyors, within the constraints of data protection rules. Nevertheless, with a survey for which organization began in February 2016, inevitably there were significant areas for which access could not be secured by either professional or volunteer surveyors in time for survey work beginning in April or May. This has limited survey coverage in some tetrads.
2.1.3 Data analysis
Data from the field survey were analysed with respect to three specific research questions: (i) what are the abundance and distribution of the target species in in-bye habitats and how do these relate to recent estimates for the total populations in England?; (ii) how do the patterns of distribution and abundance of the
target species relate to those of key current and historical AES options? (iii) how are patterns of abundance and distribution related to in-field habitat characteristics? Ideally, the data would also have been analysed in conjunction with historical abundance and distribution data to investigate changes over time and their dependence upon AES management. However, as discussed in Section 1.2.1, the data collected on the target species previously have been disparate in structure, sampling protocol and spatial scale, making any potential attempt to extract comparable count data likely to be unreliable and to produce misleading results. Hence, such quantitative analyses were not attempted. The data collected here, however, provide a reliable baseline against which to measure changes in in-bye wader populations at some future date.
Analyses for question (i) above considered count per unit area surveyed per tetrad and scaled counts up within strata to produce a national estimate for total populations in in-bye land in England for each species. Bootstrapping was used to calculate mean, median and 95% confidence limits for the total population of each species, using records identified as the maximum number of breeding pairs and, separately, adults counted, across visits. First, the ratio of total in-bye area in each surveyed tetrad to the surveyed area in the tetrad was used to scale counts up to represent the total in-bye area per tetrad, i.e. assuming that the surveyed area was representative of all in-bye in the tetrad, but not necessarily that in other tetrads. Second, within each stratum of survey tetrads, 999 bootstrap samples were selected (by resampling, with replacement) to match the numbers of tetrads in each stratum that were actually surveyed (Table 2.1). For each bootstrap, the average count per unit area of in-bye land surveyed was used to predict the total population estimate for all tetrads in the total area of in-bye considered, by multiplying up by the total area of in-bye in the stratum. The total area of in-bye in tetrads each containing less than 80ha of in-bye was calculated by region and added to the area used in the extrapolation for the zero Atlas coverage stratum for each species. These bootstrap samples were then combined to calculate summary statistics per stratum, which were then summed across strata to provide the mean, median and 95% confidence intervals for each species in in-bye across all of England. All analyses were conducted using SAS 9.4 (www.sas.com).
Analyses for question (ii) were conducted in two ways. First, counts were analysed at the tetrad level, asking whether they were influenced by the quantity of management in the tetrad. These analyses used generalized linear models appropriate for count data, i.e. using a log link function and assuming a Poisson error distribution, with a correction for overdispersion if necessary, to model count as a function of the area of relevant AES options, considering the various schemes both individually and in combination. The AES variables considered are described in Table 2.2. The individual options considered to provide beneficial management (determined on the basis of scheme handbooks and ecological knowledge) for in-bye waders that provided similar management or benefit (e.g. grazing to increase feeding habitat, hay meadows providing nesting habitat, etc.) were combined into composite variables for testing, so that similar management types did not feature in the counterfactual areas considered in tests of a given option (Table 2.3). Note that individual options may be included in more than one group if they provide more than one type of resource (Table 2.3). Options were considered in three categories dividing them from the perspective of similarities in management and three more (overlapping) from the perspective of common potential effects on birds. For these analyses, only AES management options that were active in 2013 and later were included; the 2013 cut off was justified on the assumption that management from 2013 onwards would still be likely to provide a positive benefit to breeding waders. All models included the area of in-bye land surveyed as a control, so all results referred to the effect of management, rather than of area of grossly suitable habitat. Tetrads with similar management and bird populations were highly spatially autocorrelated, which could lead to artificial inflation of the precision of the results via pseudoreplication because the data from adjacent tetrads were not independent. Therefore, models were fitted using generalized estimating equations and a repeated measures structure. To account for the autocorrelation in the most efficient way that was supported by the data, four alternative correlation structures were compared, each explicitly accounting for the correlation between data within spatial units at different spatial scales: (i) discrete regions determined by the gross distribution of in-bye tetrads of the north, west (Welsh borders) and south-west, (ii) seven broad regions chosen by eye to produce approximately equal areas/sample sizes per region, based on the distribution of tetrads and administrative boundaries (Cumbria, Lancashire, North-east, South Pennines, South and West, Yorkshire and Yorkshire Moors), (iii) national grid 100km square identity (twelve regions), and (iv) national grid 10km square identity (137 squares). Model fits were compared using quasi-information criterion (QIC) values (Pan 2001). Note that these regions were used only to account for autocorrelation in the models and not make inferences about variation in abundance or AES effects). Preliminary analyses showed that accounting for autocorrelation at the 10km square level was best-supported by the data for all species except Golden Plover, for which the broadest region classification (i) was preferred. Results for these, preferred, models are, therefore, reported and inference here is based upon these models.
The second set of analyses for question (ii) considered counts at the field level, asking whether birds were more likely than expected by chance to be recorded in in-bye habitat areas managed under an AES option. All AES option analyses considered multiple option types, grouped into themes of similar management and as combinations by scheme (Table 2.3). All bird registrations were assigned the AES option identity for the field in which they were found that had been provided in the spatial option uptake data by Natural England, assuming that point records for options referred to the field parcels in which they fell. Total numbers of adult birds and total numbers of pairs per tetrad and survey visit were then calculated for birds that were recorded within each option type, or outside it and elsewhere in the in-bye area. By chance, the proportion in the managed area should be equal to the proportion of the in-bye area that is managed under that option type. Hence, statistical tests asked whether the difference between the number of birds or pairs in managed areas observed and that expected by chance was significantly related to the absolute area of AES management present. Positive results would indicate selection of the habitat.
Although the survey protocol nominally specified two survey visits to each tetrad, many observers appeared to have covered a reduced area on the second visit, or did not make a second visit at all. Second visits were considered in deriving maximum counts at the tetrad level, as described above, but were therefore omitted here to avoid biases towards better quality habitat and to make the data consistent across tetrads. Squares with single visits only mostly occurred for tetrads in low-abundance strata, where observers found no birds and unpromising habitat on a first visit timed to coincide with maximum detectability. In some instances, professional surveyors were asked to cover additional squares with late, single visits before they conducted their second visits to an initial selection of tetrads. This may have led to some underestimation of numbers of breeding pairs if birds bred early, failed and did not re-nest, but it is still likely that they would be detected in counts of feeding adults, because the surveys still took place in May.
Analyses were conducted using generalized linear models with an identity link function and a normal error distribution, modelling the difference between observed and expected counts as a function of the area of relevant AES options. All analyses were conducted using SAS 9.4 (www.sas.com).
Table 2.2 AES variables considered in tests of associations with wader counts in in-bye land in 2016
Scheme Option Code Option Description CSS UH1 Upland Hay Meadows UP1 Upland In-Bye Pasture UP2 Upland Rough Grazing Pastures (Enclosed upto 20Ha) UP3 Upland Rough Grazing (Over 20Ha) UP4 Upland Limestone Grassland ESA 1BB Unimproved permanent grassland 1BC Enclosed rough grazing 1BM Meadows O1B Extensive Permanent Grassland Extensive permanent grassland Extensive permanent grassland and rough grazing Inbye Unimproved grassland and enclosed rough grazing (existing agreement holders only) O1C Enclosed permanent rough grazing Enclosed unimproved permanent grassland Extensive permanent grassland Wet grassland O1D Low input permanent grassland O1G Low input permanent grassland O2A Meadows Reversion of improved grassland to extensive permanent grassland Species-Rich Hay Meadows Traditional pastures Wet Grassland O2B Traditional hay meadows ESS EL3 In-bye pasture & meadows with very low inputs: SDA land HK15 Maintenance of grassland for target features HK16 Restoration of grassland for target features HL7 Maintenance of rough grazing for birds HL8 Restoration of rough grazing for birds OL3 In-bye pasture & meadows with very low inputs: SDA land(organic) UL21 No cutting strip within meadows UL22 Management of enclosed rough grazing for birds UL23 Management of upland grassland for birds UOL21 No cutting strip within meadows UOL22 Management of enclosed rough grazing for birds UOL23 Management of upland grassland for birds
Table 2.3 AES variables, options combined by common functions with respect to wader ecology (determined by ecological knowledge) considered in tests of associations with wader counts in in-bye land in 2016. Note that the nesting & feeding category consists of options that are all also listed under each category individually. The first three categories consider the options from the bird perspective and the second three from the management perspective.
AES variable Options included Likely benefits to waders tested Nesting Habitat UH1 UP2 UP3 UP4 1BB Provides tussocks or tall portions 1BM O1B O1C O1D of sward that are required for O1G O2A O2B EL3 HK15 nest sites. HK16 OL3 UL21 UL23 UOL21 UOL22 UOL23 Feeding UP1 UP2 UP3 1BC O1B Mixed sward structures with O1C O1D O1G O2A EL3 short swards permitting assess to HK15 HK16 HL7 HL8 OL3 ground-dwelling invertebrate UL22 UL23 UOL22 UOL23 prey. Nesting & UP2 UP3 1BB O1B O1C Affords both nesting habitat Feeding O2A EL3 HK15 HK16 features and elements of short OL3 UL23 UOL22 sward for feeding. Grazing UP2 UP3 1BC O1B O1C Creation of heterogeneous HL7 HL8 UL22 UOL22 structure, a short sward providing access for feeding and tussocks for nest sites. Hay UH1 1BM O2A O2B UL21 Nesting habitat, primarily for UOL21 Curlew. Restoration & O2A HK15 HK16 HL7 HL8 Increased heterogeneity of sward maintenance structure providing improved feeding and nesting opportunities.
Analyses for question (iii), relating to the distributions of birds with respect to the habitat types recorded within in-bye fields, were conducted at the field scale, comparing counts by species between fields with different habitat characteristics, as described in Table 2.4. The field data were taken from the records produced by surveyors in the field. Analyses modelled average densities (counts of adults and pairs, separately, per unit field area) as a function of the categorical or continuous variables listed. Continuous variables were tested as both linear and quadratic functions, in order to reveal both linear and non-linear responses. All analyses used log-normal models to account for data skew towards a few large values for most species, employing a repeated measures structure (and fitted using generalized estimating equations) to account for possible autocorrelation among multiple fields in the same tetrad. Analyses were again conducted using SAS.
Table 2.4 Habitat characteristic variables against which field-specific densities were tested. Variables are described as continuous, or with respect to the levels of the categorical variables considered. Full details of the definitions for each category and cues are provided in the survey instructions.
Variable Type/levels Sward type Uniform, Tussocky (majority short with minority tall) or Patchy (majority tall with minority short) Sward height Short, Medium or Long Grassland type Improved, Semi-improved, Rough grass, Hay Meadow or Silage Grazing type Cattle, Cattle & Sheep, none, Other (mostly horses) or Sheep Boundary type Treeline, Hedge, Wall, Fence or Stream Waterlogged Yes or No Bare ground area Continuous Rush area Continuous Water coverage area Continuous
2.2 Results
2.2.1 General survey results
Completed survey data were returned from 522 tetrads, with a good reporting rate for the target species using the in-bye habitat (Table 2.1). A high proportion of the tetrads in strata containing the top 25% of both Curlew and Lapwing were covered (Table 2.5). The area of in-bye covered during the survey was 73,050 ha (Table 2.6), which represents 13% of the total in-bye habitat identified within the 2837 tetrads considered and 12% of the total area of in-bye in England.
Table 2.5 Summary of species occurrence in 2016 survey tetrads: the number of tetrads from which each species was reported, the total number of adults and breeding pairs recorded, and the number of adults and breeding pairs counted within in-bye fields alone. Both adult and pair numbers were derived from tetrad- specific maxima across survey visits.
Species No. of tetrads in Total adults Total pairs In-bye only In-bye only which species adults pairs recorded Curlew 427 8333 3724 7393 3211 Golden Plover 106 2068 137 2004 111 Lapwing 375 18127 7369 17583 6874 Oystercatcher 306 4141 1715 3919 1621 Redshank 141 1161 610 1104 579 Snipe 246 1816 1408 1673 1303
Table 2.6. Details of area of in-bye coverage achieved by the survey within each of the survey strata.
Strata Region Total Total in- In-bye tetrads bye area covered (ha) (ha) Top 25% Curlew & Lapwing North 162 31354 18536 Top 25% Curlew North 51 9827 4879 Top 25% Lapwing North 92 19998 12349 25%-75% Curlew & Lapwing North 179 35563 3367 25%-75% Curlew North 320 64602 7512 25%-75% Lapwing North 278 55591 4815 All Curlew & Lapwing Southwest & West 42 7227 3298 North 877 58385 7336 Zero Curlew or Lapwing Southwest & West 416 26790 1632 North 299 162606 6833 No Atlas coverage Southwest & West 121 81593 2492 Totals 2837 553537 73050
Curlews were found in more tetrads than Lapwings, but the total numbers of the latter were considerably higher, reflecting their colonial habit (Table 2.5). Oystercatcher and Snipe were also widespread and found on approximately half the tetrads surveyed. Redshank and Golden Plover were the least numerous but were still each found in over 100 survey tetrads. Note that totals of adults are not simply double those of pairs because birds recorded in flocks will not have been registered as being breeding pairs and because some breeding pairs will have been recorded on the strength of sightings of the behaviour of single adults (see Methods).
2.2.2 Population estimates for target species in in-bye in England
Population estimates, accounting for non-covered habitat within 2837 tetrads containing in-bye, are shown in Table 2.7 for both the number of adults and the number of breeding pairs identified for each species. The 95% confidence interval s for all figures are wide, reflecting the highly variable counts between in-bye tetrads. A breakdown of these population figures by Government Office Region is presented as Appendix 3.
Table 2.7. Population estimates for adults and breeding pairs in all in-bye habitat in England, as obtained from bootstrapping of the survey results with 999 iterations. A correction factor was applied to each stratum to account for wader numbers in in-bye habitat that was not covered by the survey across all 2837 tetrads.
Species Summary statistics for population size combined across strata Numbers of pairs Numbers of adults Mean Median 5th 95th Mean Median 5th 95th percentile percentile Percentile Percentile Curlew 15233.3 15039.5 10551.5 20747.7 38822.5 38227.0 26897.3 53296.4
Golden Plover 184.7 173.5 28.8 440.9 6422.2 6044.0 995.3 15305.7
Lapwing 27498.5 27243.4 18143.6 37917.9 74507.4 73830.2 49387.1 101856.6
Oystercatcher 6993.7 6828.1 4043.9 10603.7 19609.4 18916.0 10586.3 31562.9
Redshank 1661.5 1604.4 628.1 3092.2 3552.6 3410.4 1226.1 6930.1
Snipe 4705.0 4527.5 2064.1 8090.6 6191.2 5979.1 2967.7 10266.4
2.2.3 AES associations at the tetrad level
There were significant associations between AES management and bird counts at the tetrad scale for all species except Golden Plover, and all but one of these significant associations were positive (Table 2.8), indicating either bird selection of AES-managed areas, or successful targeting of management to cover areas of higher abundance for these species. ESA management was significant only for Curlew and provided the only negative result (Table 2.8). Conversely, CSS & ESS management as a whole was positively associated with this species, but was not associated with the other species (Table 2.8). This general pattern was also detected in respect of the individual option categories of feeding, grazing and restoration for Curlew, although there was no evidence for effects of the hay or nesting option classes (Table 2.8). Snipe showed positive associations with the same option categories, but not with the overall scheme measures (Table 2.8). The patterns for the other species were less strong, but there were significant, positive associations with grazing and restoration options for Lapwing, Oystercatcher and Redshank (Table 2.8).
Table 2.8 Results of tests of associations between waders and AES variables. The variables tested are defined in Table 2.3. Parameter estimates show the statistical effect on total counts at the tetrad level of the area of the management concerned in the tetrad, together with the standard error (SE) for this effect. Chi-square test statistics and P-values refer to generalized estimating equation (GEE) analyses accounting for the autocorrelation associated with repeated measures, so significance values differ from those implied by the empirical SEs. Results significant at below P=0.1 are highlighted in bold.
Species AES variable Parameter Empirical GEE Score test P estimate SE Chi-Square Curlew CSS & ESS 0.003 0.001 6.87 0.009 ESA -0.006 0.002 7.2 0.007 Feeding 0.001 0.000 6.07 0.014 Grazing 0.002 0.001 10.42 0.001 Hay -0.007 0.003 1.96 0.162 Nesting and Feeding 0.000 0.001 0.28 0.599 Nesting 0.000 0.001 0.35 0.552 Restoration 0.002 0.001 11.26 0.001 Golden Plover CSS & ESS 0.008 0.000 1.09 0.297 ESA -0.016 0.002 1.65 0.199 Feeding 0.002 0.000 1.21 0.271 Grazing 0.006 0.000 1.11 0.291 Hay -0.012 0.002 0.95 0.331 Nesting and Feeding -0.003 0.000 1.24 0.265 Nesting -0.003 0.000 1.24 0.265 Restoration 0.005 0.000 1.19 0.275 Lapwing CSS & ESS 0.003 0.001 3.24 0.072 ESA 0.000 0.003 0 0.978 Feeding 0.001 0.001 2.54 0.111 Grazing 0.003 0.001 7.45 0.006 Hay -0.001 0.002 0.3 0.584 Nesting and Feeding -0.001 0.001 0.92 0.337 Nesting -0.001 0.001 0.65 0.419 Restoration 0.003 0.001 7.63 0.006 Oystercatcher CSS & ESS 0.002 0.001 2.65 0.104 ESA 0.000 0.003 0 0.988 Feeding 0.001 0.001 2.16 0.142 Grazing 0.003 0.001 4.9 0.027 Hay 0.001 0.002 0.04 0.843 Nesting and Feeding -0.001 0.001 0.57 0.451 Nesting -0.001 0.001 0.33 0.567 Restoration 0.003 0.001 5.19 0.023
Table 2.8, continued.
Species AES variable Parameter Empirical GEE Score test P estimate SE Chi-Square Redshank CSS & ESS 0.000 0.002 0.02 0.899 ESA 0.002 0.004 0.18 0.674 Feeding 0.001 0.002 0.61 0.434 Grazing 0.004 0.001 5.03 0.025 Hay 0.001 0.002 0.26 0.607 Nesting and Feeding -0.002 0.003 0.94 0.332 Nesting -0.002 0.003 0.73 0.395 Restoration 0.003 0.001 4.38 0.036 Snipe CSS & ESS 0.002 0.002 1.25 0.263 ESA 0.000 0.003 0 0.971 Feeding 0.002 0.001 4.22 0.040 Grazing 0.004 0.001 6.33 0.012 Hay -0.003 0.002 0.88 0.348 Nesting and Feeding 0.000 0.002 0.09 0.761 Nesting 0.000 0.001 0.07 0.788 Restoration 0.003 0.001 6.04 0.014
2.2.4 AES associations at the bird registration level
There were many significant associations between AES management and total bird counts at the field scale, and the majority of these associations were positive (Table 2.9), suggesting a positive influence of AES management on bird selection of fields. The significant negative associations were for Curlew and ESA management and Golden Plover for management associated with CSS & ESS, as well as management aimed at providing feeding and/or nesting benefits (Table 2.9). CSS & ESS management was positively associated with all species, with the exception of Golden Plover and Redshank, while all species showed significant positive associations with management directed at providing feeding resources and with habitat restoration/maintenance (Table 2.9). Grazing had a positive association with Curlew, Redshank and Snipe, while Hay was only positively associated with Lapwing. The counts of adults contained some large flocks of pre-breeding individuals, and the positive associations with managements geared towards creating breeding habitat and restoration of habitat appear to have attracted feeding birds.
For breeding pairs, all species showed positive significant associations between the number of birds on AES- managed fields, compared to non-managed ones, except Golden Plover (Table 2.10), which may reflect this species’ preference for moorland for breeding, its main use of in-bye being for feeding. All other species showed significant positive associations with both feeding and restored habitats (Table2.8). To varying degrees, these species also showed some positive association with management connected with nesting and feeding habitat delivery (Table 2.10). CSS & ESS management was only significant for Curlew, Oystercatcher and Snipe, but ESA management and hay was not associated with any of the wader species (Table 2.10). Management aimed at providing feeding and nesting habitat appears to have a broadly consistent, positive association with all breeding wader species examined, again with the exception of Golden Plover.
Table 2.9 Results of tests of preference by waders of AES management at the field level within tetrads. The variables tested are defined in Table 2.3. Parameter estimates show the statistical difference between proportions of total counts at the field level of the area of the management vs non-management in the tetrad, together with the standard error (SE) for this effect. Chi-square test statistics and P-values refer to generalized linear model analyses. Results significant at below P=0.05 are highlighted in bold.
Species AES variable Parameter Empirical Wald Chi- P estimate SE Square Curlew CSS & ESS 0.007 0.004 4.51 0.034 ESA -0.006 0.003 5.11 0.024 Feeding 0.016 0.004 20.08 0.001 Grazing 0.018 0.006 9.23 0.002 Hay 0.010 0.014 0.59 0.442 Nesting and Feeding 0.007 0.004 4.51 0.034 Nesting 0.007 0.004 4.18 0.041 Restoration 0.026 0.005 25.26 0.001 Golden Plover CSS & ESS -0.033 0.015 4.81 0.028 ESA 0.001 0.002 0.34 0.560 Feeding 0.044 0.014 9.09 0.003 Grazing 0.083 0.051 2.68 0.102 Hay -0.008 0.005 2.25 0.134 Nesting and Feeding -0.033 0.015 4.81 0.028 Nesting -0.031 0.015 4.58 0.032 Restoration 0.086 0.040 4.55 0.033 Lapwing CSS & ESS 0.024 0.011 5.09 0.024 ESA -0.018 0.014 1.75 0.186 Feeding 0.045 0.010 21.57 0.001 Grazing 0.015 0.016 0.82 0.366 Hay 0.408 0.082 24.92 0.001 Nesting and Feeding 0.024 0.011 5.09 0.024 Nesting 0.032 0.012 7.5 0.006 Restoration 0.048 0.018 7.6 0.006 Oystercatcher CSS & ESS 0.006 0.003 5.48 0.019 ESA 0.000 0.002 0.01 0.907 Feeding 0.010 0.002 17.33 0.001 Grazing 0.008 0.005 3.23 0.072 Hay -0.009 0.012 0.61 0.436 Nesting and Feeding 0.006 0.003 5.48 0.019 Nesting 0.006 0.003 4.41 0.036 Restoration 0.015 0.005 10.86 0.001 Redshank CSS & ESS 0.000 0.001 0 0.974 ESA -0.003 0.005 0.57 0.452 Feeding 0.007 0.002 25.4 0.001 Grazing 0.011 0.003 11.33 0.001 Hay 0.019 0.012 2.65 0.104 Nesting and Feeding 0.000 0.001 0 0.974 Nesting 0.000 0.001 0.01 0.930 Restoration 0.013 0.003 21.21 0.001
Table 2.9, continued.
Species AES variable Parameter Empirical Wald Chi- P estimate SE Square Snipe CSS & ESS 0.003 0.001 4.45 0.035 ESA 0.001 0.002 0.43 0.511 Feeding 0.013 0.002 39.53 0.001 Grazing 0.027 0.005 31.82 0.001 Hay -0.005 0.008 0.43 0.510 Nesting and Feeding 0.003 0.001 4.45 0.035 Nesting 0.003 0.002 2.84 0.092 Restoration 0.025 0.004 39.18 0.001
Table 2.10 Results of tests of preference by waders of AES management at the field level within tetrads. The variables tested are defined in Table 2.3. Parameter estimates show the statistical difference between proportions of total pairs at the field level of the area of the management vs non-management in the tetrad, together with the standard error (SE) for this effect. Chi-square test statistics and P-values refer to generalized linear model analyses. Results significant at below P=0.05 are highlighted in bold.
Species AES variable Parameter Empirical Wald Chi- P estimate SE Square Curlew CSS & ESS 0.005 0.002 7.63 0.006 ESA -0.002 0.002 1.55 0.213 Feeding 0.010 0.002 31.27 0.001 Grazing 0.013 0.003 16.98 0.001 Hay 0.011 0.009 1.41 0.235 Nesting and 0.005 0.002 7.63 0.006 NestingFeeding 0.005 0.002 6.71 0.010 Restoration 0.018 0.003 38.91 0.001 Golden Plover CSS & ESS 0.000 0.000 0.08 0.783 ESA 0.000 0.000 0.25 0.618 Feeding 0.002 0.002 0.84 0.359 Grazing 0.007 0.006 1.35 0.245 Hay -0.001 0.001 1.32 0.251 Nesting and Feeding 0.000 0.000 0.08 0.783 Nesting 0.000 0.000 0.07 0.791 Restoration 0.006 0.005 1.56 0.212 Lapwing CSS & ESS 0.008 0.005 2.96 0.085 ESA -0.008 0.007 1.24 0.266 Feeding 0.024 0.005 21.48 0.001 Grazing 0.016 0.009 2.97 0.085 Hay 0.048 0.027 3.22 0.073 Nesting and Feeding 0.008 0.005 2.96 0.085 Nesting 0.010 0.005 4.28 0.039 Restoration 0.036 0.010 14.06 0.001
Table 2.10, continued.
Species AES variable Param Empirical Wald Chi- P eter SE Square estimat e Oystercatcher CSS & ESS 0.004 0.001 7.23 0.007 ESA 0.000 0.002 0.06 0.814 Feeding 0.005 0.001 20.97 0.001 Grazing 0.003 0.002 2.11 0.147 Hay -0.008 0.007 1.22 0.269 Nesting and Feeding 0.004 0.001 7.23 0.007 NestingfFeeFeeding 0.003 0.001 4.47 0.035 Restoration 0.007 0.002 12.46 0.001 Redshank CSS & ESS 0.000 0.001 0.13 0.714 ESA -0.002 0.003 0.63 0.427 Feeding 0.005 0.001 26.11 0.001 Grazing 0.007 0.002 12.3 0.001 Hay 0.001 0.007 0 0.948 Nesting and Feeding 0.000 0.001 0.13 0.714 Nesting 0.000 0.001 0 0.978 Restoration 0.009 0.002 24.56 0.001 Snipe CSS & ESS 0.003 0.001 6.81 0.009 ESA 0.001 0.002 0.36 0.547 Feeding 0.012 0.002 39.17 0.001 Grazing 0.026 0.004 35.71 0.001 Hay -0.004 0.007 0.27 0.602 Nesting and Feeding 0.003 0.001 6.81 0.009 Nesting 0.003 0.001 4.16 0.041 Restoration 0.025 0.004 43.63 0.001
2.2.5 Associations between waders and habitat features
The test results for categorical habitat predictor variables are shown in Tables 2.11 and 2.12 for total adult abundance and numbers of pairs, respectively. The results for the continuous variables tested are, similarly, shown in Tables 2.13 and 2.14. Both total abundance and numbers of pairs were significantly affected by boundary type for all species. Fields with hedges and treelines were avoided, along with those with fences for Snipe, relative to other fields. Snipe were also particularly positively associated with fields with stream boundaries. Variation in numbers of pairs between grazing types was significant only for Lapwing and Redshank, which preferred mixed grazing, followed by sheep grazing alone, while Oystercatcher avoided ungrazed grassland and fields with horses (Table 2.12). There were different patterns for total abundance: all species were significantly affected by grazing type except for Snipe, and the patterns were species-specific (Table 2.11). Curlew avoided fields with horses, Golden Plover preferred fields with cattle or ungrazed fields, Lapwing avoided fields with cattle and mixed livestock, Oystercatcher avoided fields with sheep and horses, and Redshank avoided sheep fields and preferred ungrazed fields, followed by those with horses.
Considering more detailed characteristics of grassland habitat, Curlew and Snipe adult abundance was positively associated with longer swards, whereas Lapwing and Oystercatcher preferred shorter swards (Table 2.11). Numbers of breeding pairs showed the same pattern, although the variation for Oystercatcher was not significant (Table 2.12). Curlew, Lapwing and Snipe all preferred both tussocky and patchy grassland over uniform swards, with the pattern being apparent in respect of both numbers of adults and of pairs, while Golden Plover showed the same pattern in the adult abundance data alone; Redshank preferred the tussocky category in respect of both measures (Table 2.11 and 2.12). Grassland type affected adult abundance for all species except Redshank and numbers of pairs for all species except Oystercatcher (Table 2.11 and 2.12). The Curlew pattern showed a preference for rough grazing over semi-improved grassland, followed by hay and silage, with improved grass being relatively avoided, while Lapwing avoided silage and improved grass fields, Snipe preferred rough grass and hay, followed by semi-improved grassland, Redshank preferred rough grass and hay over the other types, and Oystercatcher avoided hay and silage. Waterlogging was significantly related to abundance in terms of both measures for all species except Golden Plover, with all being more abundant in waterlogged fields (Tables 2.11.and 2.12).
Areas of bare ground and of rushes in fields were associated with Curlew and Lapwing via peaking quadratic relationships, in respect of both total adult abundance and numbers of pairs, and there was a similar pattern for the area of rushes for Snipe (Tables 2.13 and 2.14). These patterns suggest that numbers tend to increase with these variables, but then reach a peak and either level off or decrease with higher areas of the features concerned. Golden Plover showed a linear, positive association between numbers of pairs and areas of rushes (Table 2.14), but a negative quadratic relationship (decreasing, but then levelling off) between adult abundance and areas of bare ground (Table 2.13). Lapwing and Snipe showed linear, positive associations with the area of cover by water, in respect of both abundance measures, while Redshank and Oystercatcher showed the same pattern for adult abundance only (Table 2.13 and 2.14). Oddly, a contrasting, decreasing quadratic relationship was apparent for Oystercatcher pairs, while Redshank (like Curlew and Golden Plover for both measures) gave rise to a non-significant result (Table 2.14).
Table 2.11 Results of tests of preference of total numbers of adults per unit area for habitat characteristics at the field level, within tetrads. The variables tested are defined in Table 2.4. Parameter estimates (on the log scale) show relative effects on counts of each level of each variable, together with the standard error (SE) for the effects. Lower (more negative) values show less preferred habitat types and SEs show the reliability of the parameter estimate values. Score test results and P-values refer to generalized linear model analyses with repeated measures. Significant test results show where one or more of the categories of the variable differ from the others, and the source of this difference is shown by the relative values of the parameter estimates and their SEs: 2×SE provides an approximate 95% confidence interval. Note that data paucity prevented analysis for Golden Plover with respect to grassland type.
Variable Species Categorical levels Estimate Score P (SE) statistic Boundary Curlew Treeline -3.01 (0.10) 63.49 <0.001 type Hedge -3.23 (0.09) Wall -2.52 (0.06) Fence -2.72 (0.12) Stream -2.69 (0.36) Golden Treeline -5.08 (0.82) 13.36 0.010 Plover Hedge -6.10 (0.32) Wall -3.66 (0.28) Fence -4.22 (0.52) Stream -4.36 (0.63) Lapwing Treeline -2.31 (0.13) 70.10 <0.001 Hedge -1.99 (0.10) Wall -1.49 (0.07) Fence -1.79 (0.09) Stream -1.82 (0.39) Oystercatcher Treeline -3.60 (0.14) 29.39 <0.001 Hedge -3.80 (0.21) Wall -3.04 (0.10) Fence -3.11 (0.32) Stream -3.77 (0.57) Redshank Treeline -5.03 (0.36) 42.99 <0.001 Hedge -5.48 (0.19) Wall -4.25 (0.14) Fence -5.33 (0.28) Stream -5.83 (1.03) Snipe Treeline -4.99 (0.21) 43.26 <0.001 Hedge -4.86 (0.18) Wall -3.93 (0.13) Fence -4.34 (0.19) Stream -2.77 (0.50)
Table 2.11, continued.
Variable Species Categorical levels Estimate Score P (SE) statistic Grazing Curlew Cattle -2.80 (0.17) 14.32 0.006 Cattle & Sheep -2.86 (0.22) None -2.62 (0.07) Other -3.25 (0.22) Sheep -2.74 (0.06) Golden Cattle -2.64 (0.07) 12.57 0.014 Plover Cattle & Sheep -5.69 (0.46) None -2.43 (0.68) Other -4.43 (0.47) Sheep -6.16 (0.75) Lapwing Cattle -3.72 (0.34) 10.34 0.035 Cattle & Sheep -3.81 (0.37) None -1.75 (0.13) Other -1.41 (0.21) Sheep -1.80 (0.08) Oystercatcher Cattle -1.72 (0.14) 20.12 <0.001 Cattle & Sheep -1.62 (0.08) None -1.74 (0.09) Other -3.35 (0.17) Sheep -3.02 (0.38) Redshank Cattle -3.42 (0.12) 21.35 <0.001 Cattle & Sheep -3.76 (0.17) None -3.02 (0.11) Other -3.23 (0.13) Sheep -5.12 (0.23) Snipe Cattle -3.91 (0.43) 4.38 0.357 Cattle & Sheep -4.79 (0.16) None -5.25 (0.22) Other -4.34 (0.15) Sheep -4.43 (0.17)
Table 2.11, continued.
Variable Species Categorical levels Estimate Score P (SE) statistic Sward Curlew Short -2.79 (0.07) 7.46 0.024 height Medium -2.64 (0.07) Long -2.48 (0.13) Golden Short -4.12 (0.38) 0.89 0.642 Plover Medium -3.81 (0.37) Long -4.49 (0.81) Lapwing Short -1.60 (0.08) 26.57 <0.001 Medium -1.74 (0.09) Long -2.10 (0.12) Oystercatcher Short -3.13 (0.10) 6.53 0.038 Medium -3.23 (0.13) Long -3.67 (0.28) Redshank Short -4.64 (0.16) 1.69 0.429 Medium -4.43 (0.17) Long -4.62 (0.26) Snipe Short -4.56 (0.15) 19.92 <0.001 Medium -4.01 (0.15) Long -3.56 (0.19) Sward type Curlew Patchy -2.42 (0.12) 57.78 <0.001 Tussocky -2.35 (0.07) Uniform -2.96 (0.06) Golden Patchy -3.41 (0.74) 5.65 0.059 Plover Tussocky -3.30 (0.35) Uniform -4.76 (0.29) Lapwing Patchy -1.62 (0.11) 12.68 0.002 Tussocky -1.55 (0.08) Uniform -1.79 (0.08) Oystercatcher Patchy -3.12 (0.22) 2.08 0.354 Tussocky -3.11 (0.12) Uniform -3.28 (0.11) Redshank Patchy -4.44 (0.21) 5.59 0.061 Tussocky -4.27 (0.17) Uniform -4.73 (0.16) Snipe Patchy -3.34 (0.20) 54.60 <0.001 Tussocky -3.59 (0.12) Uniform -5.02 (0.15)
Table 2.11, continued.
Variable Species Categorical levels Estimate Score P (SE) statistic Grassland Curlew Improved -3.11 (0.08) 61.66 <0.001 type Semi improved -2.59 (0.07) Rough grass -2.23 (0.09) Hay -2.69 (0.17) Silage -2.87 (0.16) Lapwing Improved -1.94 (0.09) 40.52 <0.001 Semi improved -1.53 (0.08) Rough grass -1.66 (0.10) Hay -1.52 (0.14) Silage -1.98 (0.21) Oystercatcher Improved -3.33 (0.11) 9.29 0.054 Semi improved -3.15 (0.13) Rough grass -3.02 (0.21) Hay -3.60 (0.17) Silage -3.52 (0.39) Redshank Improved -4.66 (0.20) 6.29 0.178 Semi improved -4.63 (0.14) Rough grass -4.21 (0.21) Hay -4.18 (0.31) Silage -4.97 (0.29) Snipe Improved -5.18 (0.19) 56.48 <0.001 Semi improved -4.24 (0.13) Rough grass -3.22 (0.14) Hay -3.54 (0.33) Silage -4.97 (0.42) Waterlogged Curlew No -2.76 (0.06) 18.72 <0.001 Yes -2.11 (0.10) Golden No -4.09 (0.30) 1.18 0.278 Plover Yes -3.25 (0.50) Lapwing No -1.78 (0.08) 23.68 <0.001 Yes -1.04 (0.10) Oystercatcher No -3.34 (0.10) 13.76 <0.001 Yes -2.27 (0.17) Redshank No -4.77 (0.15) 16.87 <0.001 Yes -3.28 (0.18) Snipe No -4.55 (0.13) 25.73 <0.001 Yes -2.58 (0.16)
Table 2.12 Results of tests of preference of total numbers of pairs per unit area for habitat characteristics at the field level, within tetrads. The variables tested are defined in Table 2.4. Parameter estimates show relative effects on counts of each level of each variable, together with the standard error (SE) for the effects. Score test results and P-values refer to generalized linear model analyses with repeated measures. Note that data paucity prevented analysis for Golden Plover with respect to grassland type. See Table 2.11 for further notes on the interpretation of these results.
Variable Species Categorical levels Estimate Score P (SE) statistic Boundary Curlew Treeline -3.92 (0.12) 38.99 <0.001 type Hedge -4.10 (0.11) Wall -3.42 (0.07) Fence -3.64 (0.13) Stream -3.17 (0.39) Golden Treeline -9.41 (0.85) 7.50 0.112 Plover Hedge -11.58 (0.49) Wall -6.48 (0.49) Fence -6.89 (0.61) Stream -5.76 (1.01) Lapwing Treeline -3.04 (0.14) 63.43 <0.001 Hedge -2.74 (0.11) Wall -2.21 (0.08) Fence -2.54 (0.10) Stream -2.48 (0.38) Oystercatcher Treeline -4.46 (0.15) 19.16 0.001 Hedge -4.60 (0.22) Wall -3.89 (0.11) Fence -3.86 (0.32) Stream -4.10 (0.65) Redshank Treeline -5.65 (0.43) 37.32 <0.001 Hedge -6.06 (0.20) Wall -4.85 (0.15) Fence -5.91 (0.29) Stream -5.75 (0.83) Snipe Treeline -5.23 (0.23) 40.03 <0.001 Hedge -5.01 (0.19) Wall -4.15 (0.13) Fence -4.57 (0.17) Stream -3.34 (0.50)
Table 2.12, continued.
Variable Species Categorical levels Estimate Score P (SE) statistic Grazing Curlew Cattle -3.68 (0.22) 7.95 0.093 Cattle & Sheep -3.56 (0.27) None -3.62 (0.08) Other -4.09 (0.24) Sheep -3.53 (0.08) Golden Cattle -11.3 (0.65) 7.90 0.095 Plover Cattle & Sheep -11.52 (0.37) None -7.28 (0.42) Other -12.85 (2.47) Sheep -6.44 (0.62) Lapwing Cattle -2.55 (0.13) 17.05 0.002 Cattle & Sheep -2.01 (0.22) None -2.56 (0.09) Other -2.41 (0.17) Sheep -2.33 (0.08) Oystercatcher Cattle -3.96 (0.19) 23.91 <0.001 Cattle & Sheep -3.82 (0.37) None -4.26 (0.12) Other -4.72 (0.21) Sheep -3.86 (0.12) Redshank Cattle -5.93 (0.24) 19.86 0.001 Cattle & Sheep -4.31 (0.41) None -5.50 (0.18) Other -5.83 (0.25) Sheep -4.88 (0.16) Snipe Cattle -4.43 (0.30) 3.44 0.487 Cattle & Sheep -4.17 (0.31) None -4.38 (0.14) Other -4.72 (0.24) Sheep -4.46 (0.14)
Table 2.12, continued.
Variable Species Categorical levels Estimate Score P (SE) statistic Sward Curlew Short -3.67 (0.08) 5.79 0.055 height Medium -3.57 (0.08) Long -3.29 (0.14) Golden Short -6.51 (0.61) 1.17 0.556 Plover Medium -7.46 (0.42) Long -7.64 (0.40) Lapwing Short -2.33 (0.08) 25.82 <0.001 Medium -2.47 (0.10) Long -2.84 (0.13) Oystercatcher Short -3.96 (0.11) 4.48 0.107 Medium -4.08 (0.14) Long -4.48 (0.32) Redshank Short -5.29 (0.17) 1.82 0.402 Medium -5.01 (0.19) Long -5.11 (0.30) Snipe Short -4.8 (0.15) 20.54 <0.001 Medium -4.26 (0.15) Long -3.72 (0.19) Sward type Curlew Patchy -3.38 (0.14) 53.01 <0.001 Tussocky -3.18 (0.07) Uniform -3.88 (0.08) Golden Patchy -7.38 (0.44) 2.77 0.251 Plover Tussocky -6.32 (0.37) Uniform -7.14 (0.78) Lapwing Patchy -2.29 (0.11) 17.72 <0.001 Tussocky -2.24 (0.09) Uniform -2.55 (0.09) Oystercatcher Patchy -4.03 (0.25) 0.92 0.630 Tussocky -3.96 (0.12) Uniform -4.09 (0.12) Redshank Patchy -5.13 (0.27) 6.70 0.035 Tussocky -4.78 (0.19) Uniform -5.38 (0.17) Snipe Patchy -3.59 (0.19) 55.16 <0.001 Tussocky -3.79 (0.12) Uniform -5.29 (0.17)
Table 2.12, continued.
Variable Species Categorical levels Estimate Score P (SE) statistic Grassland Curlew Improved -4.05 (0.10) 55.16 <0.001 type Semi improved -3.55 (0.07)
Rough grass -2.96 (0.1) Hay -3.54 (0.21) Silage -3.90 (0.26) Lapwing Improved -2.66 (0.10) 31.89 <0.001 Semi improved -2.28 (0.08)
Rough grass -2.33 (0.11) Hay -2.23 (0.16) Silage -2.73 (0.24) Oystercatcher Improved -4.19 (0.13) 4.30 0.367 Semi improved -4.01 (0.14)
Rough grass -3.78 (0.20) Hay -4.25 (0.20) Silage -4.21 (0.38) Redshank Improved -5.30 (0.22) 22.94 <0.001 Semi improved -5.31 (0.15)
Rough grass -4.64 (0.22) Hay -4.71 (0.33) Silage -5.82 (0.18) Snipe Improved -5.40 (0.21) 53.03 <0.001 Semi improved -4.51 (0.14) Rough grass -3.41 (0.14) Hay -3.83 (0.31) Silage -5.08 (0.40) Waterlogged Curlew No -3.66 (0.07) 17.08 <0.001 Yes -2.93 (0.12) Golden No -6.86 (0.46) 0.38 0.540 Plover Yes -7.35 (0.73) Lapwing No -2.51 (0.08) 23.12 <0.001 Yes -1.72 (0.11) Oystercatcher No -4.18 (0.10) 11.84 0.001 Yes -3.07 (0.18) Redshank No -5.42 (0.17) 16.83 <0.001 Yes -3.74 (0.18) Snipe No -4.77 (0.14) 27.64 <0.001 Yes -2.82 (0.15)
Table 2.13 Results of tests of preference of total numbers of adults per unit area for continuous habitat variables at the field level, within tetrads. The variables tested are defined in Table 2.4. Linear and quadratic models were fitted, with the more significant being described (or ‘NS’ where neither was significant). Parameter estimates are shown for the linear and, where applicable, squared terms for each variable (multiplied by 100, for presentational purposes), together with their standard errors (SE). Z score test results and P-values refer to the difference from zero of each parameter.
Variable Species Best Linear Squared Linear term Squared term model term (SE) term (SE) Z Score P Z Score P Bare Curlew Quadratic 3.63 (1.65) -0.04 (0.01) 2.12 0.028 -2.85 0.004 ground Golden Plover Quadratic -26.01 0.22 (0.10) area (10.23) -2.54 0.011 2.18 0.029 Lapwing Quadratic 3.31 (1.24) -0.03 (0.01) 2.67 0.008 -2.54 0.011 Oystercatcher NS ------Redshank NS ------Snipe NS ------Rush Curlew Quadratic 3.41 (0.48) -0.03 (0.01) 7.17 <0.001 -5.27 <0.001 area Golden Plover NS ------Lapwing Quadratic 2.5 (0.48) -0.04 (0.01) 5.18 <0.001 -5.03 <0.001 Oystercatcher NS ------Redshank NS ------Snipe Quadratic 5.25 (0.98) -0.03 (0.01) 5.38 <0.001 -2.83 0.005 Water Curlew NS ------coverage Golden Plover NS ------area Lapwing Linear 1.35 (0.52) - 2.57 0.01 - - Oystercatcher Linear 2.84 (1.25) - 2.27 0.023 - - Redshank Linear 2.43 (1.26) - 1.93 0.054 - - Snipe Linear 3.55 (0.74) - 4.78 <0.001 - -
Table 2.14 Results of tests of preference of total numbers of breeding pairs per unit area for continuous habitat variables at the field level, within tetrads. See Table 2.13 for details of table structure.
Variable Species Best Linear Squared Linear term Squared term model term (SE) term (SE) Z Score P Z Score P Bare Curlew Quadratic 5.01 (2.50) -0.05 (0.02) 2.01 0.045 -2.29 0.022 ground Golden Plover NS ------area Lapwing Quadratic 3.73 (1.46) -0.04 (0.01) 2.55 0.011 -2.47 0.014 Oystercatcher NS ------Redshank Quadratic -8.88 (3.28) 0.07 (0.03) -2.71 0.007 2.07 0.039 Snipe NS ------Rush Curlew Quadratic 3.89 (0.50) -0.04 (0.01) 7.72 <0.001 -5.56 <0.001 area Golden Plover Linear 1.66 (0.86) - 1.94 0.052 - - Lapwing Quadratic 2.29 (0.47) -0.03 (0.01) 4.83 <0.001 -4.60 <0.001 Oystercatcher NS ------Redshank NS ------Snipe Quadratic 4.98 (0.99) -0.03 (0.01) 5.03 <0.001 -2.37 0.018 Water Curlew NS ------coverage Golden Plover NS ------area Lapwing Linear 1.26 (0.60) - 2.12 0.034 - - Oystercatcher Quadratic -0.14 (0.08) 8.68 (4.04) 2.15 0.032 -1.82 0.07 Redshank NS ------Snipe Linear 3.30 (0.78) - 4.25 <0.001 - -
3. Discussion
3.1 The importance of in-bye areas for breeding waders in England
Table 2.7 shows that potentially significant numbers of breeding pairs of all the target species were found in in-bye fields, with the exception of Golden Plover. The latter was expected, because most birds will nest on moorland; indeed, the c. 170 pairs estimated by this study as breeding in in-bye is likely to be an over- estimate from observations of pairs of birds foraging together some distance from their nesting site. However, it is noteworthy that sizeable numbers of birds were observed as unpaired, flocking adults (Table 2.7), showing that they use in-bye farmland for foraging, as confirmed by earlier studies (e.g. Whittingham et al. 2000). For the other species considered, 1500-24000 pairs were estimated as breeding in in-bye in England. In order to place these numbers in context, estimates of the all-England populations of each were taken from the most recent collation in the literature (O’Brien 2004), and updated to 2014 assuming that abundance has followed the trends for England estimated by the BTO/JNCC/RSPB Breeding Bird Survey (BBS). This approach was used by Musgrove et al. (2013) to estimate populations for Great Britain and the United Kingdom. The results are shown in Table 3.1: it is estimated that 14% of the Redshank population is found in in-bye, reflecting the primarily coastal (saltmarsh) distribution of this species. The major decline of Redshank in Britain (c. 50%: Malpas et al. 2013) therefore seems to have occurred in in-bye habitats as well as elsewhere. Oystercatcher, while much more common than Redshank inland, has a primarily coastal distribution (on sandy and rocky shores) and only 25% of its population was found in in-bye. Conversely, over 50% of the populations of Curlew, Lapwing and Snipe were found in in in-bye, showing the importance of the habitat for these declining species.
Table 3.1 National population estimates (numbers of pairs) for the target species. Figures collated by O’Brien (2004) for England for the late 1990s have been updated using the smoothed BBS trend for England only and the ratio of the index value for 2015 to that from 1998, producing an estimate for 2014 by simple multiplication. The mean numbers of breeding pairs estimated from the current survey (Table 2.7) were then used to estimate a proportion of this population that is found in in-bye habitats.
Species Population Ratio Estimated Percentage c. 1998 2015:1998 population 2014 in in-bye from BBS Curlew 39816 0.721 28722 53 Lapwing 62651 0.817 51176 54 Oystercatcher 19191 1.43 27444 25 Redshank 21421 0.566 12132 14 Snipe 6990 1.28 8945 53
Curlew are also commonly found in moorland, with small numbers also breeding in lowland wet grassland and heathland, but in-bye farmland is clearly an important habitat for them (53% of the population). Similarly, Lapwing are common in lowland arable and pastoral farmland, including coastal grazing marshes, but 54% of the population is found in in-bye farmland. The in-bye area is certainly important for breeding Snipe, along with other wet grassland habitat, but the method used in the present study, as well as those used in the BBS and in some of the data sets considered by O’Brien (2004), are not targeted for optimum detection of this species. Ideally, Snipe should be surveyed using bespoke methods near dawn and/or dusk, so it is likely that the precise figures presented in Tables 2.5 and 3.1 are subject to unknown levels of bias
and the importance of in-bye suggested by the 53% figure in Table 3.1 should be treated with caution. The present survey is likely to have underestimated Snipe numbers; the previous surveys collated by O’Brien (2004) included both some bespoke Snipe surveys and some more general survey coverage, so are also likely to have underestimated numbers, but probably to a lesser extent than the surveys described here.
3.2 Associations of breeding waders with AES management of in-bye land
At the tetrad scale, there were significant, positive associations between several of the target wader species, especially for Curlew and for grazing and restoration, among the management types (Table 2.8). There were no such effects for Golden Plover but, with the majority of pairs of this species nesting on open moorland and not in in-bye, factors elsewhere are likely to have been important for them, weakening any associations with the habitat on which this study focused. For example, AES-managed habitats may be selected for foraging by Golden Plover if all else is equal, but distance from nesting sites in moorland may be a more significant influence, so no association with AES areas per se is apparent.
The only negative association found was for Curlew and ESA management (Table 2.8). This may show that this scheme has failed, with the birds now avoiding managed areas. This, together with the lack of positive effects for other species, may contradict the findings of Dallimer et al. (2010), who found that overall abundance of the upland bird community and birds of conservation concern was higher in fields that were surrounded by more ESA management in a more restricted study area in northern England. However, Dallimer et al. also reported that the patterns were heterogeneous among species, and the relationships for Curlew and Lapwing were less clear in species-specific analyses. Nevertheless, the negative associations with ESAs found in this study may not reflect real negative effects of the management, because such patterns could also arise from an uneven distribution of ESAs with respect to bird distributions, with this scheme having been replaced and updated with new schemes in areas with larger populations of target species, but remaining in place in areas where populations are less healthy.
The patterns found with respect to the selection of habitat areas within tetrads largely mirrored the tetrad- scale results, with the addition of negative Golden Plover associations with CSS & ES and nesting habitat (Table 2.9). The latter may show some avoidance of managed areas, but probably has limited significance because the species does not nest in in-bye land, and this pattern may be more likely to reflect a spurious association with ease of access from core, moorland areas, for example.
The positive associations found suggest that management in the grazing, restoration and, to a lesser extent, feeding categories, at the tetrad level, together with the nesting category within tetrads, could be widely beneficial across species, whereas nesting and hay category options are not selected (Tables 2.6, 2.7. and 2.8). That these patterns are only detectable at the whole scheme level for Curlew shows how management effects are option-specific and that such detailed assessments are necessary to inform about AES effects (see also Baker et al. 2012). Nevertheless, they suggest that some of the option types are succeeding in benefiting breeding waders in in-bye land. However, it is important to note that the associations detected do not prove causation: the waders may be selecting managed land, or alternatively the management may have been targeted successfully at important areas for waders. Moreover, even if managed areas are actively selected by birds, they may not have a positive influence at the population level because the results simply show a redistribution, rather than a demographically driven increase in total abundance over time. To test this possibility, long-term comparisons of population growth rates or changes in abundance are required, such as could be provided in due course by a repeat of the present survey, or (data availability permitting) by analyses of long-term survey data (e.g. Baker et al. 2012).
3.3 Relationships between waders and in-bye field habitats
The field survey results revealed a range of patterns of association between the wader species considered and the characteristics of in-bye fields. Many of these are already well-established, such as the avoidance of woody field boundaries by all species and various associations with wetter or waterlogged ground, but the ability of the survey to detect these relationships provides support for its power to detect key relationships with AES management. Noteworthy results include the contrasting patterns with respect to grazing type between species, as well as between abundance measured in terms of pairs (more likely to be breeding locations) or total adult abundance (probably dominated by feeding birds). The interspecific differences underline the variation in habitat preferences between species and, hence, the importance of multiple habitats, even within in-bye, to support multiple species. This is apparent from the relationships with both grazing types and grassland types. Similarly, differences between pair and total abundance results may show the value of different habitats for nesting and feeding. In particular, it is interesting that there were few clear preferences for ungrazed fields, or those grazed with cattle (as opposed to sheep), which are likely to feature higher habitat heterogeneity and availability of both food and nest sites. However, it is important to note that the results can only inform about the range of habitats that were available in the data and a category like ‘ungrazed grass’ could encompass a wide range of habitats, from plant-species-rich meadows to a temporarily empty improved pasture, depending on how surveyors interpreted the categories (and on how closely they followed the survey protocol).
The preferences for vegetation characteristics found largely followed expectations, for example with all species preferring more heterogeneous sward structures in terms of one or both abundance measures. Similarly, the contrasting sward height preferences of Curlew and Snipe versus Lapwing and Oystercatcher probably reflect their respective feeding ecologies. The relationships with the areas of bare ground, rushes and open water again reflect predictable associations with key habitat features such as ground moisture and vegetation heterogeneity. It is possible that the failure to find such relationships for all species reflects genuinely novel patterns in some cases, but sample sizes were limited and/or the survey data less than ideal for Redshank, Golden Plover and Snipe (as noted above, Golden Plover feed, but do not nest, in in-bye land and Snipe surveys ideally require bespoke methods), for example. In addition, the range of habitats covered within the survey data for each species is uncertain, as discussed above. It would therefore be unwise to place too much weight on individual results here, especially ones that involve a failure to find a significant result. Thus, it is possible that the positive association between Oystercatcher total counts and water cover reflects a benefit for feeding, whereas the negative one with breeding pairs reflects an unsuitability of the same fields for nesting, presumably due to lack of nesting cover or risk of flooding. However, further, more direct evidence should be sought before such patterns are used to derive management prescriptions for the species.
3.4 Implications for future field survey design
The new field survey conducted in this project was broadly successful, but fell short of the intended survey coverage of 600 tetrads and provides some important lessons for future survey projects in similar habitats or regions.
First, the definition of the target habitat needs to be clear to allow an efficient selection of survey areas and organization of fieldwork in practice. Here, ‘in-bye farmland’ was defined operationally using a definition of the moorland line provided by Natural England. This was mostly appropriate, but in some local areas habitat parcels were found to be misclassified, and the occurrence of moorland habitat at low altitudes in Northumberland caused further uncertainties of definition. It is also possible that areas under ‘in-bye’
change over time, most likely through abandonment of former pasture; such changes should ideally be identified in advance, or allowance made for them to be dealt with incorporated within project plans and budgets.
Future surveys using the present data set as a baseline will need to consider whether to revise survey area coverage. Clearly, there are good reasons to use the same survey areas for the strongest comparisons, but it should be straightforward to subset from the data collected here to facilitate data matching with reduced numbers of fields that might be of interest in subsequent years. GIS data from this project would be available to contribute to this process.
Second, achieving sufficient statistical power to detect effects of interest is critical, but the sample size required for this varies with the research question and it is impossible to identify required sample sizes without first knowing biological effect sizes. Further, the certainty with which questions can be answered depends on variability in bird numbers, sample sizes and the effect sizes that are of interest. This study attempted to maximise power by covering as much of the area of in-bye land as possible, with a focus on areas with larger local populations. The combination of professional and volunteer surveyors helped to maximize coverage. This approach worked reasonably well in high-density areas, but it proved difficult to achieve even low levels of coverage in regions where zero counts were likely, because resources did not permit the use of professionals everywhere. This increases the chance that unknown local populations were not covered adequately, but is unlikely to be a serious problem in the current context because the main target species are conspicuous, popular and generally well-reported in birdwatching circles. . Therefore, it is unlikely that any significant local populations went unreported during the Atlas project, or that areas with such populations were included in the low abundance strata here, causing bias. Biases due to failures to detect significant local populations are, therefore, unlikely. It should be noted, however, that this problem might well be more significant with volunteer surveys of other taxonomic groups, regardless of sample stratification.
Third, with any survey focusing on regions with low human population density, it can be difficult to recruit volunteer surveyors. This was recognized a priori and solutions put in place, most notably to use a significant amount of professional effort. Such effort is likely to be required in subsequent surveys in upland or northern regions, and where target species are rare or unlikely to be recorded. In addition, however, for this survey, a protocol was produced that was sufficiently flexible to allow different levels of field effort in tetrads, whether due to variable access or surveyor commitment, whilst still recording useful data. The intention was to make the survey widely accessible to observers with different amounts of available time and not to put them off if tetrads contained very large areas of target land for coverage. However, this inevitably meant the data-recording protocol was more complex than would have been ideal and this may in itself have deterred some surveyors. It also meant that some surveyors failed to submit data after reporting that they had actually conducted the survey, while others apparently surveyed the wrong areas (non-in-bye land), presumably because they did not read the instructions, so their data could not be used. The balance between enhancing data quality and maximization volunteer participation (at the cost of the quality of individual records) is a difficult one and the right decision may depend on the importance of local evidence and relationships, as opposed to gross patterns, because the latter will be less influenced by noise in the data.
A further approach to increasing survey coverage here involved attempts to facilitate the participation of visiting surveyors in covering the target areas, in addition to the default reliance on local residents. However, this was unsuccessful and offers of support for visitors from other parts of the country were not taken up. This was probably because most potential volunteers who might have been interested are people with limited free time during the bird breeding season, because they are already committed to survey work or personal projects elsewhere. More specific targeting of known, individual surveyors, such as those who have retired from full-time employment, might be a more successful tactic if a similar approach were to be taken in future surveys.
3.5 Patterns revealed by historical data
3.5.1 Value and potential of the integration of disparate data sets
It is a key conclusion of this project that the historical data that provided the strongest inference in this study were those from bird atlas projects and the BBS, i.e. structured data collection. This shows the value of structured survey data, which is simply not provided by data that does not come from consistent, standardized sampling. The other historical data considered produced some results that seem reliable because they are consistent with the Atlas patterns, but relied heavily on the contribution of BBS data. There was probably some benefit from the integration of other historical data improving the BBS trend, but the BBS data had to be degraded to some extent for integration with the other data available, so the benefits are not clear-cut.
In effect, regardless of survey standardization for local purposes, many of the historical data sets considered here provided only unstructured information when collated at the national scale and across decades. To combine data sets collected in different ways, it is necessary to reduce them to the lowest common denominator, which means a loss of structure (e.g. the benefits of standardized surveys). High-quality individual surveys, such as those using standardized methods on grid squares, can also be difficult to combine with data from other sources because of habitat specificity. For example, survey coverage with respect to habitat may not be controlled or recorded, so data from non-relevant habitats cannot be removed; this might be the case for records from in-bye within upland Brown & Shepherd 1km squares.
However, if simple pooling of data prior to analysis does not lead to strong models, where parallel, standardized local studies are available, even with different field approaches, another possible modelling method would be to combine parameter estimates using meta-analysis methods, although this would not work for long-term analyses if all the available data sets were purely spatial or short-term. In addition, spatio-temporal biases will affect all analyses of such data and need to be considered as caveats or corrected for, if data to support the correction are available. If data are assumed to be unstructured, new analytical methods are in development that will increase the quality of inference possible (Isaac & Pocock 2015). In particular, these approaches attempt to deal with the under-recording of zeroes or absences in unstructured records. Bayesian methods offer one approach to deal with this issue, using independent knowledge to assign implied zeroes to certain locations, but there would be elements of circularity in such approaches in many cases, because data and prior knowledge may not be truly independent. Related methods in development in the BTO for the integration of BirdTrack data with structured monitoring effort could prove useful for integrating datasets in the context of the present study, but have yet to be tested. Overall, however, structured data will always be more valuable for inference, and the existence of multiple data sets over a long time-series, as was the case here, does not necessarily mean that they can produce a high-quality integrated data set, regardless of the equality of the individual data sets that are combined.
3.5.2 Inference gained from constructed trends
The novel population trends based on presence-absence analyses and the 105 data sets that were collated revealed long-term declines in Curlew, Lapwing and Redshank, but increases in Oystercatcher and a noisy pattern, but overall stability, for Snipe (Figure 1.1). These patterns are very similar to those reported for the
abundance of these species across all terrestrial habitats in England by the BBS (Robinson et al. 2016). In practice, this probably mostly reveals the domination of the trends by the contribution of BBS data, although it suggests (a) that the relevant additional data sets for in-bye did not obscure or confound the patterns shown by the BBS for the wider landscape in England, and (b) that the BBS trends seem broadly representative of the populations of these species in in-bye areas (or at least that there is no evidence to the contrary). Nevertheless, BBS sample sizes in purely in-bye areas remain low, severely restricting any possible inference about relationships specific to this habitat.
3.5.3 Evidence for effects of AES management
The constructed data sets allowed analyses equivalent to those used with Atlas data to be conducted, but failed to reveal any clearly different insights or conclusions. The atlas analyses revealed several significant relationships between AES management and the target species, despite presence-absence analyses providing only coarse measures of population change. Overall, ESA management was associated with negative changes (local extinction or lack of colonization), whereas ES/CSS management was associated with positive effects. The negative ESA effects in respect of colonization prevailed for all species except Redshank, while ES and CSS showed significantly positive associations for Curlew, Lapwing and Snipe. Local extinction was apparently promoted by ESAs for Curlew and Lapwing, but restricted by ES/CSS for Curlew. However, it is important to note that all these results are correlative and that the associations do not prove either positive or negative effects of the different AES schemes, for which repeat or long-term monitoring with counterfactuals would be required.
These patterns suggest that ESAs have had little benefit for the target species, and may even have had negative effects on them. This occurred whether areas of wader-relevant ESA options alone were considered, or total ESA area, were used, suggesting that the patterns do not reflect the effects of otherwise targeted ESA management. ES and CSS effects were mostly similar whether data from before 2005 were included or not, so the restricted spatial targeting and management in CSS did not obscure the patterns due to the more sophisticated and widespread management that was brought in with ES in 2005.
The positive associations with ES and CSS management suggest the schemes have been successful in supporting the populations of target species to some extent, although the continuing declines of three of them suggest that, at best, the schemes have slowed the rates of decline. The negative effects of ESAs may show that inappropriate management has actually contributed to the declines, directly or indirectly, but ESAs may also have apparently negative effects because of a combination of insufficiently effective management and targeting towards poorer areas. A general caveat to these analyses is that they were only possible in a presence-absence framework, because count data were not collected for the 1988-91 Atlas (Gibbons et al. 1993). Presence-absence is a blunt instrument for detecting the effects of management or environmental change, so a range of more subtle effects on abundance could have been missed. However, a key advantage of the approach is that it considered a long-term change. Effect time lags are frequently a concern in studies of AES impacts, but the 2007-11 Atlas followed several years after most of the AES management considered here had occurred, so lagged effects should have had time to operate, although this is not the case for new management brought in after the revisions to ES in 2010, for example.
3.6 General conclusions
This study has confirmed that in-bye farmland is important for several wader species. Populations in this habitat have been falling, but there is evidence that conservation action via AESs may be having positive effects.
The results have shown that ESAs are associated with negative effects across species in respect of several different analyses, but ES/CSS management appears to have been much more positive. This study does not allow firm conclusions about the causality of patterns of change, but at the very least, these patterns suggest that management has been targeted effectively, even if impacts on population trends cannot be generated.
It would be valuable to repeat the survey conducted in this study in around five years to investigate the influence of management on population changes. In the meantime, further analyses of BBS data may be fruitful, but this is strongly dependent upon sample sizes in and around upland areas, which are low and are likely to restrict inference in practice for the foreseeable future. Developing analyses of other, independently collected data sets is unlikely to be of significant value for national assessments because the methods used have been too variable, but repeats of regional surveys using the same methods may well be useful to assess management impacts at more local scales.
Acknowledgements
This project was funded by Natural England. We are grateful to Allan Drewitt, Phil Grice, Andy Cooke and Stephen Duncan of Natural England for their advice and assistance throughout this project. Simon Gillings (BTO) provided invaluable advice for analysing Atlas data, Dawn Balmer (BTO) advised on survey design and volunteer liaison, Dario Massimino, Sarah Harris and David Noble helped with BBS trends, and Dave Turvey and Mark Hammond (BTO) designed and operated the online data entry system. Jean Roberts, John Loder, Nick Unwin, Julian Hudson, Patrick Woods and Tim Chamberlain conducted surveys on behalf of RSPB. We also thank the various landowners for granting survey access, and NE, RSPB and national park staff for helping to organize it. Finally, we are indebted to the volunteers who surveyed in-bye areas and the BTO Regional Representatives who coordinated their efforts.
References
Baker, D. J., Freeman, S. N., Grice, P. V., & Siriwardena, G. M. (2012). Landscape‐scale responses of birds to agri‐environment management: a test of the English Environmental Stewardship scheme. Journal of Applied Ecology, 49: 871-882.
Balmer, D. E., Gillings, S., Caffrey, B., Swann, R. L., Downie, I. S., & Fuller, R. J. (2013). Bird Atlas 2007-11: the breeding and wintering birds of Britain and Ireland. Thetford: BTO.
Brown, A. F., & Shepherd, K. B. (1993). A method for censusing upland breeding waders. Bird Study, 40: 189- 195.
Dallimer, M., Gaston, K. J., Skinner, A. M., Hanley, N., Acs, S., & Armsworth, P. R. (2010). Field-level bird abundances are enhanced by landscape-scale agri-environment scheme uptake. Biology Letters, 6: 643-646.
Gibbons, D.W., Reid, J.B. & Chapman, R.A. 1993. The New Atlas of Breeding Birds in Britain and Ireland: 1988–1991. T. & A.D. Poyser., London.
Hayhow DB, Bond AL, Eaton MA, Grice PV, Hall C, Hall J, Harris SJ, Hearn RD, Holt CA, Noble DG, Stroud DA and Wotton S (2015) The state of the UK’s birds 2015. RSPB, BTO, WWT, JNCC, NE, NIEA, NRW and SNH, Sandy, Bedfordshire.
Isaac, N. J., & Pocock, M. J. (2015). Bias and information in biological records. Biological Journal of the Linnean Society, 115(3), 522-531.
Malpas, L.R., Smart, J., Drewitt, A.L., Sharps, E. & Garbutt, A. 2013. Continued Declines of Redshank Tringa totanus breeding on saltmarsh in Great Britain: is there a solution to this conservation problem? Bird Study 60:370-383
Musgrove, A., Aebischer, N., Eaton, M., Hearn, R., Newson, S., Noble, D., Parsons, M., Risely, K. & Stroud, D. (2013). Population estimates of birds in Great Britain and the United Kingdom. British Birds, 106, 64-100.
O’Brien, M. (2004). Estimating the number of farmland waders breeding in the United Kingdom. International Wader Studies 14 (O. Thorup, Ed.) : 135–139.
O’Brien, M., & Wilson, J. D. (2011). Population changes of breeding waders on farmland in relation to agri- environment management. Bird Study, 58(4), 399-408.
Pan, W. (2001). Akaike's information criterion in generalized estimating equations. Biometrics, 57(1), 120- 125.
Robinson, R.A., Leech, D.I., Massimino, D., Woodward, I., Eglington, S.M., Marchant, J.H., Sullivan, M.J.P., Barimore, C., Dadam, D., Hammond, M.J., Harris, S.J., Noble, D.G., Walker, R.H. & Baillie, S.R. (2016) BirdTrends 2016: trends in numbers, breeding success and survival for UK breeding birds. Research Report 691. BTO, Thetford. www.bto.org/birdtrends
Smart, J., Wotton, S. R., Dillon, I. A., Cooke, A. I., Diack, I., Drewitt, A. L., Grice, P.V. & Gregory, R. D. (2014). Synergies between site protection and agri‐environment schemes for the conservation of waders on lowland wet grasslands. Ibis, 156(3), 576-590.
Whittingham, M.J., Percival, S.M. & Brown, A.F. 2000. Time budgets and foraging of breeding golden plover Pluvialis apricaria. Journal of Applied Ecology. 37:632-646.
Appendix 1. Summary of data resources available for evaluating historical wader populations in in-bye areas
Dataset Start End Survey Bird Data Survey Estimated Data Notes Year Year Extent Resolution Unit Surveyed used Defined Area (ha) 1979, Breeding Bird Survey Squares - 1979 Present Yes 100m Plot ? N Outside In- Dartmoor Environmental Baseline Data bye 1981, RSPB Moorland Survey - Peak 1981 1981 No 100m Plot 29500 Y District 1981-83, Upland Breeding Bird Surveys 1981 1981 No 1km 1km c.26000 Y In County Durham & North Yorkshire NCC 1991, Inventory Of Areas Of 1981 1981 Yes Plot Plot 13817 Y Ornithological Importance In The North Pennines (Wader Data) 1991, Areas Of Ornithological 1981 Present Yes 100m Plot 1182 Y Importance In The North Pennines (Summary Data) Survey Sites 1983, RSPB Moorland Survey - North 1982 1982 No 100m Plot ? Y York Moors 1982, RSPB Moorland Survey - Trawden 1982 1982 No 1km 1km 2000 Y
1981, RSPB Moorland Survey - Peak 1982 1982 No 1km 1km 100 Y District 1982, Breeding Waders Of Wet 1982 1982 Yes Plot Plot 7806 Y Meadows Survey In England And Wales 1981-83, Upland Breeding Bird Surveys 1982 1982 No 1km 1km 200 Y In County Durham & North Yorkshire NCC 1982 Bwwm Survey - Non Systematic 1982 1982 Yes 1km 1km ? Y Records Of Other Species 1991, Inventory Of Areas Of 1982 1982 Yes Plot Plot 4475 Y Ornithological Importance In The North Pennines (Wader Data) 1982/3 Lizard Breeding Bird Survey 1982 1983 No 2km 2km ? Y (Tetrad Data) 1982 & 2002 Breeding Waders Of Wet 1982 1983 Yes 1km 1km ? Y Meadows Summary Data (Bto, RSPB, EN) 1983, RSPB Moorland Survey - North 1983 1983 No 1km 1km 2500 Y York Moors 1981-83, Upland Breeding Bird Surveys 1983 1983 No 1km 1km ? Y In County Durham & North Yorkshire NCC 1991, Inventory Of Areas Of 1983 1983 Yes Plot Plot 3159 Y Ornithological Importance In The North Pennines (Wader Data) 1980-Present, Repeat Upland Bird 1983 Present Yes 100m Plot 399 N Outside in- Survey Sites, RSPB bye 1983-1984, North Yorkshire Upland Bird 1983 1983 Yes Plot Plot 621 Y Survey (Rubs), Nature Conservancy Council 1991, Inventory Of Areas Of 1984 1984 Yes Plot Plot 13817 Y Ornithological Importance In The North Pennines (Wader Data) 1983-1984, North Yorkshire Upland Bird 1984 1984 No 100m Plot ? Y Survey (Rubs), Nature Conservancy Council 1985, Upland Bird Survey, North York 1985 1985 No 1km 1km ? Y Moors National Park, RSPB, NCC 1985, North Staffordshire Moorland 1985 1985 Yes 1km 1km ? Y Breeding Bird Survey, RSPB 1996 North Staffordshire Moorland 1985 1985 Yes Plot Plot ? Y Breeding Bird Survey, RSPB Dataset Start End Survey Bird Data Survey Estimated Year Year Extent Resolution Unit Surveyed Defined Area (ha) 1980-Present, Repeat Upland Bird 1985 Present Yes 100m Plot 414 N Survey Sites, RSPB
1985-1986, Lake District Upland Bird 1985 1985 No 100m Plot ? Y Survey (Rubs), Lake District National Park Authority 1985-1986, Lake District Upland Bird 1986 1986 No 100m Plot ? Y Survey (Rubs), Lake District National Park Authority 1986 Breeding Bird Survey Sites - 1986 Present Yes 100m Plot ? N Outside In- Dartmoor Environmental Baseline Data bye 1991, Inventory Of Areas Of 1987 1987 Yes Plot Plot 13817 Y Ornithological Importance In The North Pennines (Wader Data) 1988, Survey Maps Of North Yorkshire 1988 1988 No 1km 1km ? Y Moors Used By D Taylor Nos. 1 - 17 (Check With Nero For Use) 1991, Inventory Of Areas Of 1988 1988 Yes Plot Plot 13817 Y Ornithological Importance In The North Pennines (Wader Data) 1988 Moorland Bird Surveys In The 1988 1988 Yes Plot Plot 507 Y Northern Pennines, NCC 1980-Present, Repeat Upland Bird 1988 Present Yes 100m Plot 591 N Survey Sites, RSPB 1988, North Pennines Upland Breeding 1988 1988 Yes Plot Plot 591 Y Bird Survey (Rubs), Nature Conservancy Council 1988 Moorland Bird Surveys In The 1988 Present Yes 100m Plot 598 Y Northern Pennines, NCC Survey Sites 1989, Breeding Waders Of Wet 1989 1989 Yes Plot Plot 4698 Y Meadows Survey In England And Wales Bto RSPB 1991, Inventory Of Areas Of 1989 1989 Yes Plot Plot 4475 Y Ornithological Importance In The North Pennines (Wader Data) 1996, Breeding Waders In The West 1990 1990 Yes Plot Plot 3300 N Plot area Pennine Moors too large Pre-1989, North England Records (Check 1990 1990 No 1km 1km ? Y With Nero For Use) 1990, Breeding Birds Of The South 1990 1990 No 1km 1km ? Y Pennine Moors, JNCC 1991, Inventory Of Areas Of 1990 1990 Yes Plot Plot 13817 Y Ornithological Importance In The North Pennines (Wader Data) 1980-Present, Repeat Upland Bird 1990 Present Yes 100m Plot 403 N Survey Sites, RSPB 1990, South Pennines Upland Bird 1990 1990 No 100m Plot ? Y Survey (Rubs), English Nature 1993 Onwards Breeding Corncrake 1990 Present Yes 100m Plot 3721486 N Annual Monitoring - Survey Areas 1996, Breeding Waders In The West 1991 1991 Yes Plot Plot 3300 N Plot area Pennine Moors too large 1991, Belmont Study Area 1991 1991 Yes Plot Plot 3300 Summary N data over large area 1991, Inventory Of Areas Of 1991 1991 Yes Plot Plot 13817 Y Ornithological Importance In The North Pennines (Wader Data) 1992, North Staffordshire Moors Upland 1992 1992 Yes 1km 1km 22600 Y Breeding Bird Survey, RSPB 1992, Northumberland Np Breeding 1992 1992 No 100m Plot ? Y GP & D only GoldEN Plover And Dunlin 1992, Breeding Success Of Curlew 1992 1992 No 1km 1km ? Y CU only Dataset Start End Survey Bird Data Survey Estimated Year Year Extent Resolution Unit Surveyed Defined Area (ha) 1996, Breeding Waders In The West 1992 1992 Yes Plot Plot 3300 N Plot area Pennine Moors too large 1985, North Staffordshire Moorland 1992 1992 Yes Plot Plot ? Y Breeding Bird Survey, RSPB
1992 Onwards, Dartmoor Breeding 1992 1992 No 100m Plot ? y Wader Data (Confidential)
1992 Breeding Bird Survey Squares - 1992 Present Yes 100m Plot ? N Outside In- Dartmoor Environmental Baseline Data bye From 1992 Report 1993, Forest Of Bowland Survey 1993 1993 No 100m Plot ? Y (Confidential) 1993, Northern Lowland Breeding 1993 1993 No 1km 1km 11700 Y Waders - Random Sites 1993 Survey Of Breeding Waders In 1993 1993 No 1km 1km 13600 Y Baldersdale And Lunedale, Co. Durham, England 1996, Breeding Waders In The West 1993 1993 Yes Plot Plot 3300 N Plot area Pennine Moors too large 1993, Lowland Waders (Breeding) In 1993 1993 Yes 1km 1km ? Y Northern England - Key Site Summaries 1993,Survey Of Breeding Wader, Forest 1993 1993 No 1km 1km ? Y Of Bowland (Check With Nwro For Use) 1992 Onwards, Dartmoor Breeding 1993 1993 No 100m Plot ? N Outside In- Wader Data (Confidential) bye Bodmin Moor Winter Surveys 1993-1995 1993 1993 No 100m Plot ? N Non- & 2002-2004 breeding 1999 And 1993, Breeding Waders In 1993 1993 No 1km 1km ? Y Baldersdale And Lunedale, Co. Durham, England 1994 Survey Of Upland Breeding Birds In 1994 1994 No 1km 1km 5200 Y The Northumberland Np : Otterburn Training Area 1994, Breeding Season In The Forest Of 1994 1994 No 100m Plot ? Y Bowland: Birds Of Prey (Check With Nwro For Use) 1994 - Birds In Greater Manchester, 1994 1994 Yes Plot Plot ? Y County Report (Includes Some Confidential Records) 1994 Survey Of Breeding Waders In 1994 1994 Yes Plot Plot ? Y South Tynedale And West Allendale (Check With Nero For Use) 1996, Breeding Waders In The West 1994 1994 Yes Plot Plot 3300 N Plot area Pennine Moors too large 1992 Onwards, Dartmoor Breeding 1994 1994 No 100m Plot ? Y Wader Data (Confidential) Bodmin Moor Winter Surveys 1993-1995 1994 1994 No 100m Plot ? N Non- & 2002-2004 breeding season Nro Regional Records - Site Data 1994 Present Yes 100m Plot 3300 N
1994 Survey Of Breeding Waders In 1994 Present Yes 100m Plot ? Y South Tynedale And West Allendale - Survey Sites 1994-1995, North Pennines Upland 1994 1995 No 100m Plot ? N Waders 1993-1995, Greater Manchester Report 1995 1995 Yes Plot Plot ? Y Dervived 1995, Forest Of Bowland 1995 1995 No 100m Plot ? Y
1995 Breeding Waders Survey On Low 1995 1995 Yes Plot Plot ? Y Intensity Agricultural Land In South Northumberland Dataset Start End Survey Bird Data Survey Estimated Year Year Extent Resolution Unit Surveyed Defined Area (ha) 1995 Breeding Birds On Low Intensity 1995 1995 Yes Plot Plot 713 Y Agricultural Land In The Branchend Farm Area, Northumberland 1992 Onwards, Dartmoor Breeding 1995 1995 No 100m Plot ? Y Wader Data (Confidential) Bodmin Moor Winter Surveys 1993-1995 1995 1995 No 100m Plot ? N Non- & 2002-2004 breeding 1995 Breeding Birds On Low Intensity 1995 Present Yes 100m Plot 713 Y Agricultural Land In The Branchend Farm Area - Survey Site 1995 Breeding Waders Survey On Low 1995 Present Yes 100m Plot ? Y Intensity Agricultural Land In South Northumberland Survey Sites
1996 North Staffordshire Moorland 1996 1996 Yes 1km 1km ? Y Breeding Bird Survey, RSPB 1996 Fellhouse Fell Breeding Wader 1996 1996 No 1km 1km ? Y Survey 1996 Greenridge Breeding Wader Survey 1996 1996 No 1km 1km ? Y (Check With Nero For Use) 1996 Coan Wood Breeding Wader 1996 1996 No 1km 1km ? Y Survey (Check With Nero For Use) 1996 Wellhope Breeding Wader Survey 1996 1996 No 1km 1km ? Y
1996 Wooley Breeding Wader Survey 1996 1996 No 1km 1km ? Y
1996 Pennine Grouse Moor Survey 1996 1996 No 1km 1km ? Y (Check With Con Sci Shq For Use) 1996, North York Moors Breeding Wader 1996 1996 Yes Plot Plot 60+ Y Survey, RSPB And The North York Moors National Park 1992 Onwards, Dartmoor Breeding 1996 1996 No 100m Plot ? Y Wader Data (Confidential) 1996 Fellhouse Fell Breeding Wader 1996 Present Yes 100m Plot 1344 Y Survey: Survey Boundary 1996 Wellhope Breeding Wader Survey - 1996 Present Yes 100m Plot 297 Y Survey Boundary 1996 Wooley Breeding Wader Survey - 1996 Present Yes 100m Plot ? Y Survey Boundary 1996 Coan Wood Breeding Wader 1996 Present Yes 100m Plot 953 Y Survey (Check With Nero For Use) Survey Sites 1996 Greenridge Breeding Wader Survey 1996 Present Yes 100m Plot 468 Y (Check With Nero For Use) Survey Sites 1996, North York Moors Breeding Wader 1996 Present Yes 100m Plot ? Y Survey, Survey Sites 1997, Lancashire & Merseyside (Forest 1997 1997 Yes 2km 2km ? Y Of Bowland) Breeding Bird Survey 1992 Onwards, Dartmoor Breeding 1997 1997 No 100m Plot ? Y Wader Data (Confidential) North Staffordshire Breeding Bird Survey 1997 Present Yes 100m Plot ? N
1997, Breeding Wader Survey Of 1997 1997 No 100m Plot ? Y Marginal/In-Bye Land And OpEN Moorland In Nidderdale AONB 1998, Forest Of Bowland General 1998 1998 Yes 2km 2km 4400 Y
1998 Survey Of Breeding Waders On 1998 1998 No 1km 1km ? Y Agricultural Land In The Pennine Dales Esa 1992 Onwards, Dartmoor Breeding 1998 1998 No 100m Plot ? Y Wader Data (Confidential) Dataset Start End Survey Bird Data Survey Estimated Year Year Extent Resolution Unit Surveyed Defined Area (ha) 1999-2000 Upland Grazing Study- 1999 1999 Yes Plot Plot 6869 Y Breeding Season Summary, South Scotland & North Pennines Black Mountains Breeding Bird Survey 1999 Present Yes 100m Plot 18343 Y 1999 1999, Bodmin Moor Breeding Birds - 1999 1999 No 1km 1km ? Y Wader And Non-Wader 1km Squares 1999, Bodmin Moor Breeding Bird 1999 1999 No 1km 1km ? Y Survey (Bbs), RSPB 1992 Onwards, Dartmoor Breeding 1999 1999 No 100m Plot ? Y Wader Data (Confidential) 1999 Black Mountains Breeding Bird 1999 1999 No 1km 1km ? Y Survey Upland Grazing Study - Sites Data 1999 Present Yes 100m Plot ? Y
1999-2003, Upland Grazing Study - Raw 1999 1999 Yes Plot Plot ? Y Data, Wales, Pennines, South Scotland 1999-2003, Upland Grazing Study - 1999 1999 Yes Plot Plot ? Y Summary Data, Wales, Pennines, South
Scotland 1999, Bodmin Moor Breeding Bird 1999 Present Yes 100m Plot ? Y Survey (Bbs), RSPB Survey Sites Geltsdale- Upland Plots 1-11- Survey 1999 Present Yes 100m Plot ? Y Areas 1999 - 2014 1999 And 1993, Breeding Waders In 1999 1999 No 1km 1km ? Y Baldersdale And Lunedale, Co. Durham, England 1999-2000 Upland Grazing Study- 2000 2000 Yes Plot Plot 6869 Y Breeding Season Summary, South Scotland & North Pennines 2000 North York Moors Breeding Wader 2000 2000 Yes Plot Plot 60+ Y Survey, RSPB, EN And North York Moors National Park 2000, Yorkshire Dales National Park 2000 2000 Yes Plot Plot ? Y Upland Breeding Wader Survey 1992 Onwards, Dartmoor Breeding 2000 2000 No 100m Plot ? Y Wader Data (Confidential) 2000, North Pennines Repeat Upland 2000 2000 Yes Plot Plot 591 Y Bird Survey (Rubs), RSPB 2000, South Pennines Repeat Upland 2000 2000 No 100m Plot ? Y Bird Survey (Rubs), RSPB 2000, Yorkshire Dales National Park 2000 Present Yes 100m Plot ? Y Upland Breeding Wader Survey, Survey Sites English Seabird Monitoring Project South 2000 Present Yes 100m Plot 428 N West England 2006-2009 - Survey Sites 1992 Onwards, Dartmoor Breeding 2002 2002 No 100m Plot ? Y Wader Data (Confidential) 2002, Breeding Waders Of Wet 2002 2002 No 100m Plot ? Y Meadows Survey - Raw Data Bto, RSPB, En 2002, Breeding Waders Of Wet 2002 2002 No 100m Plot ? Y Meadows Survey - Site Totals Bto, RSPB, En 1982 & 2002 Breeding Waders Of Wet 2002 2002 Yes 1km 1km ? Y Meadows Summary Data (Bto, RSPB, En) Bodmin Moor Winter Surveys 1993-1995 2002 2002 No 100m Plot ? N Non- & 2002-2004 breeding 2002, Lake District Repeat Upland Bird 2002 2002 No 100m Plot ? Y Survey (Rubs), RSPB 2002, North Yorkshire Repeat Upland 2002 2002 No 100m Plot ? Y Bird Survey (Rubs), RSPB Dataset Start End Survey Bird Data Survey Estimated Year Year Extent Resolution Unit Surveyed Defined Area (ha) 1999-2003, Upland Grazing Study - Raw 2002 2002 Yes Plot Plot ? Y Data, Wales, Pennines, South Scotland 1999-2003, Upland Grazing Study - 2002 2002 Yes Plot Plot ? Y Summary Data, Wales, Pennines, South Scotland 2002, Peak District Upland Waders 2002 2002 No 100m Plot ? Y
2003, Zone One - Survey Of Moorland 2003 2003 No 100m Plot ? Y Fringe Wading Birds & Twite Around The South Pennine Moors 2003, Zone Two- Survey Of Moorland 2003 2003 No 100m Plot ? Y Fringe Wading Birds & Twite Around The South Pennine Moors 2003, Zone Three - Survey Of Moorland 2003 2003 No 100m Plot ? Y Fringe Wading Birds & Twite Around The South Pennine Moors 2003, North York Moors Wader Survey 2003 2003 Yes Plot Plot 7478 Y (Check With Nero For Use) Bodmin Moor Winter Surveys 1993-1995 2003 2003 No 100m Plot ? N Non- & 2002-2004 breeding 1999-2003, Upland Grazing Study - Raw 2003 2003 Yes Plot Plot ? Y Data, Wales, Pennines, South Scotland 1999-2003, Upland Grazing Study - 2003 2003 Yes Plot Plot ? Y Summary Data, Wales, Pennines, South Scotland
2004-2005, North Pennines Breeding 2005 2005 Yes 100m Plot ? Y Wader Monitoring 2005, West Penwith Breeding Bird 2005 2005 No 1km 1km ? Y Survey (RSPB) 2005-2012 Curlew Breeding Data From 2005 2005 No 1km 1km ? Y Operation Wader Project CU only 2006-2007, North Pennines Spa Survey, 2006 2006 No 100m Plot ? Y Natural England 2004-2010, Lapwing Recovery Project - 2006 2006 Yes 100m Plot ? Y Raw Data, Bowland (RSPB, Ne) 2004-2010, Lapwing Recovery Project - 2006 2006 Yes 100m Plot ? Y Field Level Summary Data, Bowland (RSPB, Ne) 2004-2010, Lapwing Recovery Project - 2006 2006 Yes Plot Plot ? Y Site Level Summary Data, Bowland (RSPB, Ne) 2006-2007, North Pennines Spa Survey, 2007 2007 No 100m Plot ? Y Natural England 2004-2010, Lapwing Recovery Project - 2007 2007 Yes 100m Plot ? Y Raw Data, Bowland (RSPB, Ne) 2004-2010, Lapwing Recovery Project - 2007 2007 Yes 100m Plot ? Y Field Level Summary Data, Bowland (RSPB, Ne) 2004-2010, Lapwing Recovery Project - 2007 2007 Yes Plot Plot ? Y Site Level Summary Data, Bowland (RSPB, Ne) 2005-2012 Curlew Breeding Data From 2007 2007 No 1km 1km ? Y CU only Operation Wader Project 2007-2010 Pennines Lapwing Trial 2007 2007 Yes Plot Plot ? Y Management Project, Farm Level Summary Data 2008 Dartmoor Breeding Snipe Survey 2008 2008 No 100m Plot ? Y (Dartmoor Wader Project) 2008, Bodmin Moor Breeding Bird 2008 2008 No 1km 1km ? Y Survey (Bbs) 2008, Bodmin Moor Wader/Non-Wader 2008 2008 No 1km 1km ? Y 1km Squares Dataset Start End Survey Bird Data Survey Estimated Year Year Extent Resolution Unit Surveyed Defined Area (ha) 2004-2010, Lapwing Recovery Project - 2008 2008 Yes 100m Plot ? Y Raw Data, Bowland (RSPB, Ne) 2004-2010, Lapwing Recovery Project - 2008 2008 Yes 100m Plot ? Y Field Level Summary Data, Bowland (RSPB, Ne) 2004-2010, Lapwing Recovery Project - 2008 2008 Yes Plot Plot ? Y Site Level Summary Data, Bowland (RSPB, Ne) 2008-2010 South Pennine Twite 2008 Present Yes 100m Plot ? N Recovery Project Monitoring (Site Boundaries Only), NE & RSPB 2005-2012 Curlew Breeding Data From 2008 2008 No 1km 1km ? Y CU only Operation Wader Project 2007-2010 Pennines Lapwing Trial 2008 2008 Yes Plot Plot ? Y Management Project, Farm Level Summary Data 2009 National Cirl Bunting Survey - 2009 Present Yes 100m Plot 400 N Surveyed Tetrads (RSPB) Breeding Bird Survey, Visit Data 1985 2015 No 1km 1km ? Y (BTO/JNCC/RSPB) 2004-2010, Lapwing Recovery Project - 2009 2009 Yes 100m Plot ? Y Raw Data, Bowland (RSPB, Ne) 2004-2010, Lapwing Recovery Project - 2009 2009 Yes 100m Plot ? Y Field Level Summary Data, Bowland (RSPB, Ne) 2004-2010, Lapwing Recovery Project - 2009 2009 Yes Plot Plot ? Y Site Level Summary Data, Bowland (RSPB, Ne) 2009-2010, Curlew Declines 2009 2009 Yes Plot Plot 38634 Y Investigation, Summary Data, South Pennines And South Scotland
2005-2012 Curlew Breeding Data From 2009 2009 No 1km 1km ? Y Operation Wader Project CU only 2007-2010 Pennines Lapwing Trial 2009 2009 Yes Plot Plot ? Y Management Project, Farm Level Summary Data 2010 Upland Breeding Bird Survey - 2010 2010 No 100m Plot ? Y Eastern Moors And Sheffield City Council Moors, Peak District Breeding Bird Survey, Visit Data 2010 2010 No 1km 1km ? Y (Bto/JNCC/RSPB) 2004-2010, Lapwing Recovery Project - 2010 2010 Yes 100m Plot ? Y Raw Data, Bowland (RSPB, Ne) 2004-2010, Lapwing Recovery Project - 2010 2010 Yes 100m Plot ? Y Field Level Summary Data, Bowland (RSPB, Ne) 2004-2010, Lapwing Recovery Project - 2010 2010 Yes Plot Plot ? Y Site Level Summary Data, Bowland (RSPB, Ne) 2009-2010, Investigating The Causes Of 2010 2010 Yes Plot Plot ? Y Curlew Declines, Raw Data, South Pennines And South Scotland 2009-2010, Curlew Declines 2010 2010 Yes Plot Plot 38634 Y Investigation, Summary Data, South Pennines And South Scotland 2005-2012 Curlew Breeding Data From 2010 2010 No 1km 1km ? Y Operation Wader Project CU only 2007-2010 Pennines Lapwing Trial 2010 2010 Yes Plot Plot ? Y Management Project, Farm Level L. only Summary Data 2005-2012 Curlew Breeding Data From 2011 2011 No 1km 1km ? Y Operation Wader Project CU only 2012-2014 Bodmin Moor Mires Project 2012 2012 No 100m Plot ? Y
Dataset Start End Survey Bird Data Survey Estimated Year Year Extent Resolution Unit Surveyed Defined Area (ha) 2012, Bowland Upland Waders 2012 2012 No 100m Plot ?
2012-2014 Bodmin Moor Mires Project 2013 2013 No 1km 1km ? Y
2012-2014 Bodmin Moor Mires Project 2014 2014 No 100m Plot ? Y
2014 Exmoor Breeding Bird Survey Raw 2014 2014 No 100m Plot ? Y Data Geltsdale Farmland Wader Surveys 2015 2015 No 100m Plot ? Y
2015 Curlew Trial Management Project 2015 2015 No 100m Plot ? Y
Geltsdale - Raw Data
Appendix 2. Survey documentation
Preliminary overview of survey documentation and information
Downloads (from survey Resources webpage):
- Survey and recording Methods (pages 2-4) - Tetrad visit details & bird recording codes (page 5) - Field coverage and Habitat and bird recording codes (pages 6) - Map of the survey tetrad showing fields (see example page 7) - Surveyor Letter of introduction – TWIMC – still to be prepared - Tetrad visit form (page 10) - Field and coverage recording form (page 11) - Bird recording form (page 8-9) Resources (from survey webpage):
- Sward type examples - Examples of habitat, grassland type and rush appearance - Guide to crop types - Examples of boundary types - Example tetrad with labelled ‘in-bye’ fields (page 7) -
Survey Methods - Breeding Waders of English Upland Farmland 2016
Background Many species of breeding waders are in decline, including those found on moorland and farmland. For this survey we will focus on the upland farmland component, just below the moorland line, which is termed ‘In- bye’ and is known to hold substantial numbers of breeding waders, especially Curlew and Lapwing. We have defined all farmland up to 1km from the edge of unenclosed moorland as ‘In-bye’, and this forms our core survey area. The principal conservation measures in ‘In-bye’ farmland are agri-environment schemes (AES), which include management aiming to maintain or to increase breeding wader numbers. This survey aims to collect data on current population levels in ‘In-bye’ land, both in and out of AES management, and a baseline against which to measure future change. Details of survey locations and background information, including detailed survey methods, and bird and habitat recording forms, is available from: www.bto.org/BWEUF
Survey design Tetrads (2km x 2km squares) will be used as the survey units. Our aim is to cover 700 core tetrads, but up to 300 additional tetrads will be available. Each tetrad contains a minimum of 80 hectares of potentially suitable ‘In-bye’ habitat and typically the maximum will not be more than half the tetrad. This survey is funded by Defra, so only covers England, and survey sites are only found in or near the uplands. However, we are keen to involve surveyors from non-upland areas and welcome expeditions – for further details of how we can help you to take part, please contact the survey organiser: Greg Conway (Email: [email protected] Tel.: 01842 750050).
What to do First, visit the survey page www.bto.org /BWEUF to view the available survey tetrads. Next login, or register, if you have not previously taken part in a BTO online survey, to request your chosen survey tetrad(s). Then the BTO Regional Organiser will confirm the tetrad(s) that you have been allocated, which will be displayed when you next login to the survey. Download and print a detailed map showing all the ‘In-bye’ (priority coverage), other field and moorland areas (optional coverage), each uniquely numbered. If your tetrad has a large number of small fields and the labels are not all displayed, or you would just prefer a larger-scale field map, there is an option to print off separate maps for each of the four 1km squares.
Secondly, form the Resources page on the website (www.bto.org /BWEUF/Resources), print a copies of survey instructions along with a sufficient number of copies of the recording forms (Tetrad Visit, Habitat and Bird Recording), to cover all the fields in the square. Ideally, we would like bird and habitat data for all labelled ‘In-bye’ fields, but we recognize that this will be difficult where fields are small and numerous, so please cover as many as possible. Also ensure that you do cover the same fields on both visits.
Visits Two survey visits are required: Visit: 1: 1st April to 31st May & Visit 2: 1st June to 15th July, with at least two weeks between visits. Survey visits should start no earlier than 30 minutes after sunrise and aim to be finished before midday. Surveys should be undertaken in fine conditions and wind below Beaufort Force 4 wherever possible, with good visibility.
It is strongly recommended that you make a recce visit prior to the first survey visit in order to contact landowners and arrange access to fields. Where possible, we ask that you make contact with the landowner to ask for permission to enter the fields. A letter of introduction can be downloaded from the Resources page, which can be shown to landowners to explain the purpose of the survey and why access to the fields is required. Where agri-environment schemes are present, we will aim to inform the landowner about the survey and help to arrange access. However, where it has not been possible to contact the landowner, check to see if the fields can be viewed from public rights of way or viewed from a vantage point using a telescope.
Survey coverage Within the tetrad, we have indicated, using grey shading, the priority ‘In-bye’ fields requiring coverage, access or visibility permitting. In addition, if you see or hear any of the target species in the other fields outside the shaded area, or in the unenclosed moorland, please also record them. Details of habitat and bird recording methods are given below, but please note that effective wader surveys in fields with tussocky or long vegetation may require access within c. 100m of each point in the field. Note that prolonged disturbance of breeding birds should be avoided, but a short survey visit is unlikely to pose any significant problem. However, if you find yourself in an area with a number of active nests, please retreat carefully, retracing your footsteps, to minimize the potential risk to the nest contents. It is important to maintain good public relations with landowners and ensure that the survey has minimal impact.
Also, it is very important that we know which parts of the nominated ‘In-bye’ area you have been able to access, so if you cannot access, or view, all parts of a field, please record how much of the field you have covered. Additionally, we need to know whether the field was entered or viewed from outside the boundary, as the method of coverage is required so that we can interpret the count results. Overall, we need you to tell us how much of the indicated ‘in-bye’ areas you have been able to survey adequately, which will partly be a question of judgement, based on topography and vegetation height. If there is good visibility over a field and you can confirm absence or count the birds present with confidence from the boundary, there is no need to enter the field.
Field Coverage and Habitat recording The habitat and boundary features associated with each field are essential for assessing both the suitability for, and use by, breeding waders so, if at all possible, we need to know habitat information for each field, even if no waders are observed or bird surveys were not possible.
All fields containing potentially suitable habitat within the ‘In-bye’ have been shown, shaded grey, and uniquely numbered. For fields outside the 1km ‘In-bye’ buffer, these have labelled with numbers preceded by an X (e.g. X1, X2). Areas of moorland are labelled with numbers preceded by an M (e.g. M1, M2). These habitat categories are derived from satellite and mapping data, and may not be 100% correct, so if the indicated field contains unsuitable wader breeding habitat (e.g. woodland, scrub, urban, etc.), please record it as ‘Other unsuitable’.
On the first visit, please record habitat and land-use details using all applicable items on the ‘Field Coverage and Habitat Form’ for all the numbered ‘in-bye’ fields, unless they could not be viewed even from a distance, and for all non-target ‘X’ fields and ‘M’ moorland areas where birds have been recorded. There is no need to repeat this on the second visit, unless there have been obvious changes (such as ploughing, crop emergence, mowing, flooding, etc.). For all ‘In-bye’ fields, please record coverage types as: Complete or Partial (record the percentage of field surveyed) and for Not covered, indicate whether this was because Access denied or Not viewable (not viewed).
Field boundary changes If you find that field boundaries have changed such that a boundary has been removed please still record the two areas as separate fields. However, if there have been more substantial changes, so that the boundaries no longer resemble those on the map, please mark the boundary changes on a printed copy of the map and label the ‘new’ field by re-allocating the original field numbers, or adding new numbers if needed. Then send the amended maps to the Survey Organiser after the first visit, so the field boundaries can be changed online before you make the second visit.
Bird recording The Key species to record are the waders: Curlew, Lapwing, Oystercatcher, Redshank and Snipe. There is also a list of Required species to record, which include raptors and corvids (Carrion Crow, Magpie, Raven, Buzzard, Hen Harrier, Kestrel, Merlin, Peregrine and Sparrowhawk) to help interpret predation risk and breeding productivity and gamebirds (Black Grouse, Grey Partridge, Pheasant, and Red-legged Partridge) to inform about game keeping intensity. Also the ‘In-bye’ is an important habitat for a number of other bird species of conservation value; we are particularly interested in records of the following species: Cuckoo, Golden Plover, Linnet, Meadow Pipit, Reed Bunting, Ring Ouzel, Skylark, Stonechat, Twite, Wheatear, Whinchat and Yellow Wagtail. Optionally, any other bird species may be recorded. Please download as many copies of the bird recording form as required, for use in the field.
Please make the following counts for all Key and Required species, and optionally any other species of interest observed, for each field covered (either fully or partially):
1) Total adults (all species) using the field, including birds that may have moved from other fields.
2) Total chicks & young (Key species only). 3) Estimated pairs (all species), based on the number of adults apparently holding territory (excluding feeding flocks), displaying/territorial males and active nests. NOTE When making the Estimate of pairs, DO NOT include birds that are known, or strongly suspected, to have moved between fields and have been counted already. Thus, counts of birds in each field should reflect the use of the field that you observe and may therefore include double-counting across fields, but the Estimated pairs should add up to the total number of pairs present in the tetrad. Counts of birds should also, therefore, include birds that are not nesting in the field, such as feeding flocks, or in flight (excluding high flying commuting species, such as gulls, etc), including either hunting/foraging or song/display flight .
For more information, please visit the Resources page (www.bto.org/BWEUF/Resources), which contains a full list of survey instructions, recording forms and other useful information, that can be viewed and downloaded, or contact the Survey Organiser: Greg Conway, BTO, The Nunnery, Thetford, Norfolk, IP24 2PU Tel.: 01842 750050 Email: [email protected]
Summary field recording codes and information
Tetrad visit details and codes
Use the Tetrad Visit Form to record the dates and start and finish times for both survey visits.
Mammalian predators Record details of mammalian predators for the whole tetrad on each visit, including: Badger Fox Weasel Cat Hedgehog Polecat/Ferret Dog Stoat
For each visit, enter a total count from all sightings within the tetrad, or tick if present in 2016, based on the following: fresh signs (droppings or footprints), reliable reports or seen on additional visits.
Game keeper and predator control activity Please provide counts of all signs of game keeper/predator control activity observed within the survey tetrad on each visit, including the following: Pheasant pens Game bird feeders Larsen traps Crow traps Stoat traps Dead corvids - hung on fence lines Dead moles - hung on fence lines
The presence of game keeping activities should be recorded if they are known to be present during 2016, based upon: reliable reports from locals or landowners observed on additional visits
Bird recording details
For each field covered (either fully or partially), please record all Key and Required species, plus any other species of interest. Counts consist of the following: Total adults = all adults seen or heard (displaying/singing) in each field (all species) Total young = all chicks or juveniles seen (only required for waders). Estimated pairs = best estimate of the actual number of pairs (excluding mobile birds already counted in other fields and feeding flocks), based on the number of: adults apparently holding territory (2 adults/single adult = 1 pair), displaying/singing males (male=1 pair) or active nests (nest = 1 pair). Key species: Curlew Lapwing Oystercatcher Redshank Snipe
Required species: Buzzard Kestrel Peregrine Merlin Hen Harrier Sparrowhawk Carrion Crow Magpie Raven Black Grouse Cuckoo Golden Plover Pheasant Grey Partridge Red-legged Partridge Linnet Meadow Pipit Reed Bunting Ring Ouzel Skylark Stonechat Twite Wheatear Whinchat Yellow Wagtail
Field Coverage and Habitat and Recording code
First print off Habitat Recording Forms and write in all the field numbers, in order (e.g. 1 to 87), for the ’In-bye’ fields, plus take some extra forms for any additional fields or moorland areas you might also cover.
Coverage - For each field within the ‘In-bye’’ area (shaded and numbered) we need to know whether it was covered or not and if coverage was achieved by walking the field or just viewing it.
Coverage type: -Tick whether coverage was Complete or Partial (give % covered). Coverage methods: -Tick whether the field was Walked inside the boundary or Viewed from outside the boundary. Not covered: - Tick whether it was Access denied or Not viewable (not viewed or not visited)
Habitat type: Grassland (enclosed) – below the moorland line, meadows, grass and rush pastures enclosed by fences or walls. Crop – including Cereal, Root, Maize and tilled/unplanted fields. Moorland – above the moorland line, typically unenclosed (containing heather moor, blanket bog, Molinia or rough grassland). Other unsuitable – includes scrub, woodland, urban and other land uses.
Enclosed Grassland type (See examples on resource page): Improved - with or without livestock and typically a uniform height sward, green in colour. Semi-improved - with or without livestock, green in colour, often with areas of tussocks/rushes Rough grass - usually enclosed grazed land (not moorland) with an uneven, tussocky, rank green-brown sward; any grazing at low density. Hay Meadow - Tall uneven (or mown) sward with many wildflowers, light green in colour. Silage - Tall uniform thick (or mown) sward, with few, if any, wildflowers, dark green in colour.
Crop type: Cereal - Winter/spring cereal (uniform in height and colour, typically bluish green) Root / Other - including Maize, fodder crops, ploughed or emerging crop/re- seeded
Sward height (use majority, grass only – excluding rushes): Short - 0 to 5 cm Medium - 5 to 15 cm Long - 15+ cm Not determined - not known
Sward structure (grass only – excluding rushes): Uniform - nearly all the same height Tussocky - majority short with minority tall Patchy - majority tall with minority short
Grazing (only record livestock if observed): Cattle Sheep Horse/Pony Other No grazing
Field details: Rush/juncus cover - estimate to nearest 5%. Standing water - estimate to nearest 5%. Waterlogged - Tick if Yes, otherwise leave blank Mown (within 4 weeks) - Tick if Yes, otherwise leave blank Bare ground - estimate to nearest 5%.
Boundary type: (tick all that make up more than 20% of the total field boundary): Treeline Hedge Stonewall Fence River/Stream
‘Example’ Tetrad Map - Breeding Waders of English Upland Farmland
Please aim to cover all, or as many as possible, of the ‘In-bye’ fields (shaded grey and numbered), and record habitat details as far as possible, even if you are unable to conduct a bird survey.
Optionally, cover any fields outside the ‘In-bye’ (labelled X1, X2, X3, etc), if the habitat appears suitable for breeding waders.
There is no need to cover the moorland areas (labelled M1, M2, etc), if present, however please do record any waders or bird species observed.
X7
X5 X6
X1 X3 X8 X4 X2
Bird Recording Form - Breeding Waders of English Upland Farmland
Tetrad: ______Date:____/____/2015 Visit:____
For each field covered (either fully or partially), please record all Key and Required species, plus any other species of interest. Counts should include feeding flocks and flying birds, either hunting/foraging or singing/displaying over fields (excluding high flying commuting birds e.g. gulls), consisting of the following: - Total adults = all adults seen or heard (displaying/singing) in each field (all species) - Total young = all chicks or juveniles seen (only required for waders). - Estimated pairs = best estimate of the number of pairs, excluding mobile birds already counted in other fields and feeding flocks, based on the number of: A) adults apparently holding territory (1 or 2 adults= 1 pair), B) displaying/singing males (male=1 pair) or C) active nests (nest = 1 pair). Key species: Curlew Lapwing Oystercatcher Redshank Snipe
Required species: Buzzard Kestrel Peregrine Merlin Hen Harrier Sparrowhawk Carrion Crow Magpie Raven Black Grouse Cuckoo Golden Plover Pheasant Grey Partridge Red-legged Partridge Linnet Meadow Pipit Reed Bunting Ring Ouzel Skylark Stonechat Twite Wheatear Whinchat Yellow Wagtail
Field Species Total Total Estimated Field Species Total Total Estimated no. Adults Young pairs no. Adults Young pairs eg Curlew 2 1 1 eg Wheatear 5 - 3 14 7
Tetrad Visit Recording Form - Breeding Waders of English Upland Farmland 2016
Tetrad: ______Visit 1 date: ___/___/2015 Start time: ___:____ Finish time: ___:____
Visit 2 date: ___/___/2015 Start time: ___:____ Finish time: ___:____
Mammalian predators Please provide a Count of all mammalian predators observed within the survey tetrad on each visit. Tick as Present if predators are known to have occurred during 2016, based upon*: A) fresh signs (droppings or footprints), B) reliable reports or C) seen on additional visits.
Visit 1 Visit 2 Count Present* Count Present* Badger Cat Dog Fox Hedgehog Stoat Weasel Polecat/Ferret
Game keeper and predator control activity: Please provide counts of all signs of game keeper/predator control activity observed within the tetrad on each visit. Tick as Present, if game keeping activities are known to have occurred during 2016, based upon*: A) reliable reports or B) observed on additional visits
Visit 1 Visit 2 Count Present* Count Present* Pheasant pens Game bird feeders Larsen traps Crow traps Stoat traps
Dead corvids Dead moles
Habitat Recording Form - Breeding Waders of English Upland Farmland Tetrad: ______Date:___/___/2016 Visit:___
Please record coverage and habitat details for all ‘In-bye’ fields (labelled 1,2,3.…), even if no waders were found, plus any optional additional fields ( X1, X2,…) or moorland areas (M1, M2,..), if present . Tick all items that apply to each field and enter percentage cover, if relevant, for columns shaded dark grey (see example row at top of table.)
Crop Other Coverage All fields/habitat Enclosed Grassland Type Habitats Not Boundary type Grazing Sward height Type Method Field features Grassland type Sward type covered (all >20% of total length) (livestock present) (majority)
Complete Partial % covered inside Walked outside Viewed viewed Not Tree line Bank Hedge / / Stream River Rush/Juncus (%) cover Water Waterlogged weeks) (< 4 Mown (%) Bare ground Cattle Sheep HorsePony / Other grazing No Improved Semi grass Rough Meadow Hay Silage (0 Short Medium(5 (15+ cm) Long determined Not Uniform Tussocky Patchy Cereal / Other Root Moorland unsuitable Other
Access denied Stonewall Fence
-
improved
-
5cm)
-
15cm)
(%)
Field
no. eg. 9 20 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Appendix 3. Breakdown of population estimates for each species by government office region. Note that the totals do not sum to precisely the same totals as shown in Table 2.7 because of rounding errors and sampling variation.
Species GOR Numbers of pairs Numbers of adults Mean Median 5th 95th Mean Median 5th 95th percentile percentile percentile percentile Curlew East Midlands 569.3 530.0 198.8 1001.1 1317.0 1273.6 500.5 2399.9 North East 4420.0 4343.7 2183.4 6872.2 10487. 10369.6 5489.9 15887.0 8 North West 2294.0 2266.8 1153.7 3509.2 7074.0 6777.0 2811.6 13319.8 South West 23.3 21.3 1.5 64.7 24.5 21.4 1.7 69.4 West Midlands 221.6 212.7 91.7 384.8 882.7 840.8 223.8 1957.0 Yorkshire and The Humber 7194.0 7061.1 4565.0 10237.3 17340. 17127.5 11800.0 23671.0 6 Golden Plover East Midlands 0.0 0.0 0.0 0.0 120.5 88.6 0.0 362.2 North East 91.7 88.9 38.2 156.8 4805.7 4382.6 355.2 12060.0 North West 10.7 10.2 0.0 29.6 154.9 143.4 12.9 396.2 South West 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 West Midlands 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Yorkshire and The Humber 138.8 125.5 3.3 391.3 2452.2 2246.3 323.9 5709.3 Lapwing East Midlands 232.3 219.3 87.5 402.6 795.1 715.2 214.0 1572.6 North East 10118.1 9834.0 5219.6 16214.9 29118. 28397.7 13167.0 47752.6 7 North West 4707.8 4498.6 1503.7 8802.8 14040. 13119.3 3924.0 29238.1 8 South West 43.3 39.4 0.0 127.6 87.8 81.9 0.0 263.6 West Midlands 149.3 147.6 86.1 238.4 563.8 565.9 236.4 1080.1 Yorkshire and The Humber 13441.2 13207.4 8019.8 19544.8 34003. 33373.2 21580.9 48868.7 5
Appendix 3, continued.
Species GOR Numbers of pairs Numbers of adults Mean Median 5th 95th Mean Median 5th 95th percentile percentile percentile percentil e Oystercatcher East Midlands 4.8 4.8 4.8 4.8 6.0 6.0 6.0 6.0 North East 2255.3 2167.0 989.8 3822.0 5506.6 5345.1 2331.4 9285.7 North West 1232.5 1196.1 373.2 2340.8 5056.6 4694.1 875.6 11576.0 South West 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 West Midlands 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Yorkshire and The Humber 3540.6 3407.1 1663.9 5979.6 9051.1 8750.0 4303.4 14982.3 Redshank East Midlands 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 North East 1165.5 1095.5 453.4 2264.1 2189.6 2070.5 856.1 4052.5 North West 141.6 133.9 24.5 282.2 283.2 270.8 62.7 567.0 South West 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 West Midlands 0.0 0.0 0.0 0.0 126.8 110.9 0.0 390.7 Yorkshire and The Humber 768.4 717.1 182.4 1667.2 1531.8 1418.2 344.6 3247.1 Snipe East Midlands 28.3 25.7 3.6 60.6 27.3 24.8 3.6 59.6 North East 3125.0 2995.7 1281.7 5464.7 4066.2 3943.2 1753.6 6876.4 North West 393.5 373.2 79.8 845.4 550.6 522.7 122.0 1177.8 South West 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 West Midlands 37.4 36.6 16.1 67.7 37.9 36.5 16.1 71.1 Yorkshire and The Humber 1619.9 1522.6 431.1 3359.8 2084.3 1961.9 653.3 4177.4