This post is modified from an article which appeared in the British Ecological Society bulletin in July 2014.
Have you ever wished there was a species list for your field site, or wondered how to get hold of a distribution map for your favourite species? If so, the National Biodiversity Network is here to help.
The UK is renowned as the cradle of natural history. Ever since Gilbert White started observing wildlife at Selborne, British naturalists have been recording faster than anywhere else on the globe. Today around 13% of the occurrence records on the Global Biodiversity Information Facility (GBIF) are from the UK. These records are conveyed to GBIF by the National Biodiversity Network (NBN).
Confusingly, the term NBN is used in three different contexts. The ‘true’ NBN is a confederation of organisations with an interest in the collection, storage and use of biodiversity data. These organisations include data providers such as BTO, statutory bodies such as Natural England. There’s also the NBN Trust, which has a small number of employees including a directorate. The third and most visible part of the NBN is the NBN Gateway, which provides access to over 90 million species records via a range of sophisticated web services, including mapping tools.
Recently, the NBN has been seeking greater interaction with the scientific community. This has prompted considerable debate within the NBN community and among researchers who use NBN data, most recently at a one day workshop hosted by the Natural History Museum. Research scientists have different requirements from other NBN users, such as planners and environmental consultancies, and this presents a number of challenges for both sides. But it also represents a real opportunity for BES members to shape the direction of NBN activities.
The biggest issue with NBN data among research scientists is data quality. Scientists need to know that the data they use is reliable for the questions they hope to ask, and data quality is a complicated and multi-faceted problem. One problem is biased coverage: three quarters of the records on NBN are of birds, Lepidoptera and vascular plants. The coverage of many less charismatic groups is extremely patchy. There’s also a big problem of scale, because records on the NBN Gateway come from a mixed bag of data sources, including local-scale intensive surveys and extensive national schemes, much of it collected by volunteers. The sampling intensity of these data types are vastly different, making it difficult to know how they should be combined. Many of the extensive data were gathered with the primary purpose of producing static maps of species’ distributions at 100 km2 resolution. Increasingly these data are being used to address questions about species’ fine-scale habitat preferences and temporal trends in status, and it’s not clear whether this is really appropriate.
Fortunately, considerable effort is going into quantifying these biases especially in my own recent research. Making NBN data amenable to modelling would help scientists and the statutory organisations such as JNCC that report on the state of the nation’s biodiversity. These developments might one day feed into the way NBN data are presented on the NBN Gateway, such as displaying modelled species distributions alongside the raw data. Another example is that species lists for user-defined sites could include species that are expected to be present, given their distribution elsewhere and the sampling intensity in the reserve, in addition to those actually recorded. This kind of output could be branded somehow to indicate that the modelling conforms to a reputable scientific standard.
There are other ways in which scientists can shape the future of the NBN. Biodiversity data are increasingly being applied to questions that impinge on government policy, where the burden of proof is high and the scrutiny intense. This has led to demands for new records to be collected in ways that are informed by ecological sampling theory, and ecologists can play an important role in this process. Whilst nobody is suggesting that volunteers should be expected to follow Bill Sutherland’s Ecological Census Techniques to the letter, there are many ways in which the recording process could become more systematic. The BTO’s BirdTrack system has led the way in asking recorders to tick a checkbox if the species records constitute a complete list of all the birds that were observed, as opposed to a subset that the recorder found most interesting. In the near future, records submitted by smartphone apps could include metadata about the extent and duration of the survey. Interactions between volunteer and professional ecologists, perhaps assisted by social scientists, will determine which kinds of innovations are adopted, and which are not.
Access to NBN data used to be complicated, but has been streamlined in recent years. Research use of NBN data is now easier still, thanks to the rNBN package. Interactions between NBN and researchers could one day become mutualistic: the open science agenda mandates scientists to make their data available to all, and the NBN has the potential to develop into the data repository of choice for UK research ecologists.
So watch this space. The NBN Trust has recently appointed a new Chief Executive and the strategic priorities are being actively debated. The voice of researchers should be active in shaping that debate.
Have you ever wished there was a species list for your field site, or wondered how to get hold of a distribution map for your favourite species? If so, the National Biodiversity Network is here to help.
The UK is renowned as the cradle of natural history. Ever since Gilbert White started observing wildlife at Selborne, British naturalists have been recording faster than anywhere else on the globe. Today around 13% of the occurrence records on the Global Biodiversity Information Facility (GBIF) are from the UK. These records are conveyed to GBIF by the National Biodiversity Network (NBN).
Confusingly, the term NBN is used in three different contexts. The ‘true’ NBN is a confederation of organisations with an interest in the collection, storage and use of biodiversity data. These organisations include data providers such as BTO, statutory bodies such as Natural England. There’s also the NBN Trust, which has a small number of employees including a directorate. The third and most visible part of the NBN is the NBN Gateway, which provides access to over 90 million species records via a range of sophisticated web services, including mapping tools.
Recently, the NBN has been seeking greater interaction with the scientific community. This has prompted considerable debate within the NBN community and among researchers who use NBN data, most recently at a one day workshop hosted by the Natural History Museum. Research scientists have different requirements from other NBN users, such as planners and environmental consultancies, and this presents a number of challenges for both sides. But it also represents a real opportunity for BES members to shape the direction of NBN activities.
The biggest issue with NBN data among research scientists is data quality. Scientists need to know that the data they use is reliable for the questions they hope to ask, and data quality is a complicated and multi-faceted problem. One problem is biased coverage: three quarters of the records on NBN are of birds, Lepidoptera and vascular plants. The coverage of many less charismatic groups is extremely patchy. There’s also a big problem of scale, because records on the NBN Gateway come from a mixed bag of data sources, including local-scale intensive surveys and extensive national schemes, much of it collected by volunteers. The sampling intensity of these data types are vastly different, making it difficult to know how they should be combined. Many of the extensive data were gathered with the primary purpose of producing static maps of species’ distributions at 100 km2 resolution. Increasingly these data are being used to address questions about species’ fine-scale habitat preferences and temporal trends in status, and it’s not clear whether this is really appropriate.
Fortunately, considerable effort is going into quantifying these biases especially in my own recent research. Making NBN data amenable to modelling would help scientists and the statutory organisations such as JNCC that report on the state of the nation’s biodiversity. These developments might one day feed into the way NBN data are presented on the NBN Gateway, such as displaying modelled species distributions alongside the raw data. Another example is that species lists for user-defined sites could include species that are expected to be present, given their distribution elsewhere and the sampling intensity in the reserve, in addition to those actually recorded. This kind of output could be branded somehow to indicate that the modelling conforms to a reputable scientific standard.
There are other ways in which scientists can shape the future of the NBN. Biodiversity data are increasingly being applied to questions that impinge on government policy, where the burden of proof is high and the scrutiny intense. This has led to demands for new records to be collected in ways that are informed by ecological sampling theory, and ecologists can play an important role in this process. Whilst nobody is suggesting that volunteers should be expected to follow Bill Sutherland’s Ecological Census Techniques to the letter, there are many ways in which the recording process could become more systematic. The BTO’s BirdTrack system has led the way in asking recorders to tick a checkbox if the species records constitute a complete list of all the birds that were observed, as opposed to a subset that the recorder found most interesting. In the near future, records submitted by smartphone apps could include metadata about the extent and duration of the survey. Interactions between volunteer and professional ecologists, perhaps assisted by social scientists, will determine which kinds of innovations are adopted, and which are not.
Access to NBN data used to be complicated, but has been streamlined in recent years. Research use of NBN data is now easier still, thanks to the rNBN package. Interactions between NBN and researchers could one day become mutualistic: the open science agenda mandates scientists to make their data available to all, and the NBN has the potential to develop into the data repository of choice for UK research ecologists.
So watch this space. The NBN Trust has recently appointed a new Chief Executive and the strategic priorities are being actively debated. The voice of researchers should be active in shaping that debate.