About Ecobat

Ecobat provides the most objective tool for assessing bat activity.

Acoustic surveys using bat detectors are widely used to determine species’ presence and to quantify the activity of foraging bats as they are cost-effective, can be automated to run for long time periods, and are non-intrusive. Activity levels are dependent on a number of factors including seasonality, weather conditions and location, with the type of bat detector used during the survey also affecting detection rates.

Using bat passes to assess the relative importance of a site for policymakers therefore requires practitioners to account for how these multiple factors may have influenced the number of bat passes recorded at a site. Although professional opinion is valuable, it can often be based on intuition, is context dependent and can vary considerably between practitioners (Hulme 2014).

It is therefore likely that an assessment of the ecological value of a site (and the impacts of any proposed action) will vary between practitioners based upon their own level of experience and knowledge of the region and/or species.

Overview of Approach
What is a reference range?
What are percentiles?
Can you use percentiles to define areas of 'low', 'moderate' or 'high' activity?
What data do we need and why?

Data Input
Why do you not accept alternative surveying methods?
What happens if I upload multiple detector locations within 1 csv file?

Date Output
Why can I not stratify by habitat type?
Can Ecobat deal with species with low detectability?
Why is there insufficient data in my area to produce a large enough reference range?
How do you account for differences in night length between surveys?
What output do you provide?

Data Security & Costs
Is there a charge for using Ecobat?
Who do you share data with?
What precautions are in place to protect Ecobat data?

Overview of Approach

What is a reference range?
To counteract the inherent subjectivity of bat assessments we are attempting to build a large enough dataset so that bat activity recorded at a focal site can be contextualised against reference levels recorded in the same region, at the same time of year etc. The ‘reference range’ is the stratified dataset by which percentile outputs can be generated.

What are percentiles?
Percentiles provide a numerical indicator of the relative importance of a nights’ worth of bat activity. For example, activity data in the 70th percentile would indicate that the recorded data was in the top 30% of activity for the reference range. 

Can you use percentiles to define areas of 'low', 'moderate' or 'high' activity?
We welcome industry input in suggesting cut-off levels between categories, however, provisionally we suggest the following:

- i) low activity: 0-20th percentiles;
- ii) low to moderate activity: 21st-40th percentiles;
- iii) moderate activity: 41st-60th percentiles;
- iv) moderate to high activity: 61st-80th percentiles; and
- v) high activity: 81st-100th percentiles.

What data do we need and why?
The robustness of output relies on populating the underlying datebase with sufficient data. We currently have more than 50,000 records of nightly bat activity across the UK, however by collating as many nightly records of bat activity as possible this will allow us to provide increased precision to our estimates; we would therefore appreciate any data that you could offer.

Essential
To ensure we have sufficient information to populate our database and to provide you with a robust analysis we require some essential data. These are:
Location: The location where the static detector was placed
Sensitivity: The confidentiality of the dataset (Do not publishBlur records to 10kmPublic)
Species: The species (or genus) of bats recorded
Pass definition: The method used to define a bat pass
Passes per night: The total number of passes recorded per night, per species
Date: The date at the start of the nights surveying
Bat detector make: The make of detector used in the survey 
Bat detector model: The model of detector used in the survey
Additional information on these criteria is available to download here.

Supplementary Data
As the Ecobat database grows we will be able to add additional levels of stratification so that your results become increasingly precise. We recommend that the following supplementary data is also included:
Detector height: The height of the detector used in the survey
Roost proximity: Within 25m of known roost
Roost proximity: Bat activity elevated because of known roost nearby: Detector placed on flight line from/to known roost location. 
Linear features: Detector placed next to linear feature
Linear features: Detector placed <25m of linear feature
Anthropogenic features: Detector placed next to anthropogenic feature
Sunset weather conditions: Temperature (°C)
Sunset weather conditions: Rainfall: 
Sunset weather conditions: Wind speed (mph)
Notes 
Any additional information that you consider important
Additional information on these criteria is available to download here.

Data Input

Why do you not accept alternative surveying methods?
We hope to offer the opportunity to deposit and analyse hourly records of bat activity in the foreseeable future, this will be particularly useful for quantifying bat activity rates during dawn and dusk surveys. The spatial variation inherent within the surveying methodology of transect surveys makes it unsuitable into the reference range framework.   

What happens if I upload multiple detector locations within 1 csv file?
You will receive the standard csv output containing generated percentiles for each night of activity per species at each detector location within your uploaded csv file, and summary output for each detector location within your uploaded csv file.

Data Output

Why can I not stratify by habitat type?
When assessing risk to local bat populations it is pertinent to consider how activity levels vary across all habitat types rather than restricting the analysis to one specific land use. For example, an agricultural development site may be perceived to contain high bat activity relative to surrounding agriculture but in comparison to the surrounding heterogeneous landscape it may contain a relatively low level of bat activity. With the exception of woodland specialists, bats are rarely confined to one habitat and therefore we believe a robust analysis is provided by quantifying bat activity across all habitat types.

Can Ecobat deal with species with low detectability?
Although we provide Ecobat outputs for all species, caution should always be used in interpreting acoustic data from species with low detectability (e.g. Plecotus and Rhinolophus spp.). Although habitat use has been inferred from its echolocation calls for these species, we suggest that additional methods, such as night vision or infrared camera equipment, are additionally used to determine its relative presence within an area.

Why is there insufficient data in my area to produce a large enough reference range?
The effectiveness of Ecobat is dependent on end users uploading their data to make assessments possible. We therefore recommend i) that you check back regularly to see if our database has expanded, ii) upload your own data and encourage everyone in your region to do similarly, iii) expand the geographic filter that you are attempting to stratify by (e.g. from 100km2 to 200km2). 

How do you account for differences in night length between surveys?
Stratifying by both geographical location and date will ensure you are only comparing your data with records containing similar night length. Caution is advised if you decide to analyse your dataset with no geographical filter for species with a UK-wide distribution (e.g. common pipistrelles) as night length varies considerably between northern Scotland and southern England. For species with a relatively restricted range (e.g. serotines) then differences in night length across the range will be less of an issue. 

What output do you provide?
Ecobat will provide a summary of how many nights of surveying effort were classified as either high, high/moderate, moderate, low/moderate, and low activity. It will also provide the median percentile and the 95% confidence interval around the median (where the dataset is insufficient to produce a confidence interval, the range of the percentiles will be generated instead). Ecobat will also list the highest night of activity and which percentile this was on. Finally, Ecobat will list how the reference range was created including how many records of nightly activity the import was contrasted against, and what levels of stratification were applied (i.e. date, location, pass definition).

Data Security & Costs

Is there a charge for using Ecobat?
No, Ecobat is solely intended to be for the benefit of bat conservation.

Who do you share data with?
If your data is marked 'do not publish' then this data will not be shared with anyone. If marked as either 'public' or 'blur to 10km' then this data will be shared with the National Biodiversity Network and Local Record Centres automatically at the spatial resolution that you have selected. 

What precautions are in place to protect Ecobat data?
Data imported to Ecobat are stored on the Indicia data warehouse, hosted by the Centre for Ecology & Hydrology where the security of the server is governed by the NERC security policy; servers are backed up nightly and firewalls are in place to ensure data security. Please contact us if you require any additional information. 

References

Hulme 2014: Bridging the knowing–doing gap: know-who, know-what, know-why, know-how and know-when.