Surveys of Breeding Birds – Survey Methodology
Prior to starting the survey, record the date, start time, temperature (°C), wind (Beaufort scale 0–12), cloud cover (Okta 0–9) and rain (as a brief description). The end time, and weather (if there has been a substantial change during the survey), should also be recorded.
Where possible, surveys should be carried out in good conditions, avoiding heavy rain, strong winds (Beaufort force >6) and any scenario where visibility/detection is negatively affected (e.g. fog, excessive disturbance on site etc.).
Generally, surveys should be carried out between dawn and mid-morning (10-11 am, with some regional variation). Consideration should be given to species which are active earlier or later in the morning making sure that both are covered by the survey
- This allows for clear differentiation between birds using the site and the immediate surroundings, and those singing farther away. In addition, certain species are not as detectable during the dawn chorus as they are later in the morning (i.e. Willow Warbler). It should also be noted however, that some species sing strongly at (or shortly before/after) dawn or dusk and should be considered on a specific basis (e.g. Black Redstart on industrial sites).
Consideration should be given to whether at least one survey should be carried out in the evening, during the last few hours of the day, and extending beyond dusk for at least one hour (also see supplementary survey methods).
- Certain species call into the dusk and after dark. These include several common species (e.g. European Robin) and some which can be difficult to detect during the day (e.g. Grasshopper Warbler, Nightingale, Nightjar and several owl species). Dusk survey visits also provide a good opportunity to identify roost sites on site.
Surveys should be carried out at intervals of about one month. However, it may be beneficial for surveys targeted at certain species to prioritise particular months. For example,
- Nightingales stop singing after egg laying, so early or later visits may be more valuable in recording this species. Likewise, Skylark detectability drops off after crops are harvested in late summer. Also note that several common species including Dunnock are at their detection peak in March-April, and therefore may be missed during the core breeding season in ?????.
Any alterations to the survey length or timings should be clearly recorded and appropriate justification provided in any reporting.
A survey transect should be walked at a slow, ambling pace, stopping to scan priority habitat/features where appropriate.
- Priority habitat/features could include (but are not limited to) trees, dense hedgerows or berry bearing shrubs, reed bed, or small bodies of water – generally any area associated with skulking and shy species with low detection (e.g. Dunnock during the summer months)
One or more survey transects should cover the entirety of the site to a minimum distance of 50 m dependent on terrain and habitats present (see supplementary survey methods), but without excessive disturbance of the habitats on site.
- This is currently considered best practice but will be constrained by the topography of the site itself. Additional consideration should be given to any hazards identified in the site-specific risk assessment. Under no circumstances should observers enter into a hazardous situation in order to complete a survey. Instead, surveys of inaccessible or hazardous habitats should be undertaken from a vantage point (see point counts) or other safe distance.
- Care should also be taken to avoid causing damage to arable crops of disturbance to livestock present within a survey area. In these situations, it is acceptable for observers to keep to field margins or tram lines but consider a vantage point (see point counts), or supplementary survey, from a reasonable distance.
Surrounding habitat (outside of the redline boundary) should also be surveyed to a reasonable distance – particularly land immediately surrounding the site itself and, any nearby, potentially valuable habitat which is likely to be impacted by development/disturbance (including but not restricted to waterbodies and ancient woodland).
Where there is a requirement to split a site into multiple survey areas, due to the overall size of the site, there are two primary approaches that can be considered, splitting or sampling.
Large sites can be split into more appropriate sized areas which can then be surveyed either by multiple teams on the same day, or by a single team over the course of several days. While this will provide good coverage, splitting a site has the potential to artificially inflate the number of recorded territories, as there may be overlaps between survey areas. The potential for such duplication should be considered at both the survey design and data analysis stages, and accounted for in any reporting.
A site may be randomly sub-sampled by dividing it into a grid and randomly selecting an appropriate number of squares to survey. Such an approach should only be undertaken within areas of uniform habitat. Where a site consists of multiple habitat types then stratified random sampling will be required, whereby a separate grid is created for each habitat type, and a random selection of squares across each habitat grid should be selected. Accounting for such variations in habitat composition in the sampling methodology will maximise the chances of collecting data that is representative of the entire site.
Point counts are when observations or bird calls are recorded from a fixed position overlooking a survey area. Habitats which are inaccessible habitat (e.g. wetland or reed beds) or where access is not permitted should be scanned for a set time from a high and clear vantage point (see SNH methodology, appendix 1).
During point counts an observer stands in one location for a set period of time (usually 5-10 mins, but some evidence has shown that two hours is best for short-eared owl surveys etc.), recording all birds seen and plotting them accordingly.
This method is particularly suitable for conspicuous birds in woody or scrubby habitats. It is suitable for extensive areas but does not allow detailed territory mapping. Point counts may be more suitable in difficult terrain than transect surveys. Point count locations can be placed on a regular grid, thus ensuring unbiased coverage of habitats, or can be targeted to ensure all habitats are represented in the sampling.
- Detailed guidance on conducting point counts/vantage point surveys is outside the scope of this guidance document, instead see SNH guidance listed in appendix 1. However, point counts should be incorporated into bird survey transects in order to survey inaccessible, sensitive or hazardous habitats. Point counts are also useful to establish nest site location and therefore should also be considered for specific species surveys (also see, Schedule 1 species).
Vantage point surveys should be considered for sites where development plans include any structure which may pose a collision risk to birds (e.g. the erection or modification of pylons, or tall structures) (see SNH methodology, appendix 1).