Improving species representation in natural capital tools
Improving species representation in natural capital tools
There is an increasing understanding that environmental decision-making needs to take account of natural capital considerations and make assessments of the value of the benefits we gain from ecosystems.
The natural capital valuation tools developed to help with this have tended to overlook or inadequately represent biodiversity considerations since their species models have been restricted to highly recorded species such as birds for which good quality data exists.
In recognition of this limitation, JNCC has created a Biodiversity Module based on a species distribution model framework that supports data from as wide a range of taxa as possible, with improved pre-processing to maximise the use of low resolution observation data. As a result, it is of particular use for lesser-known or poorly recorded species.
The JNCC Biodiversity Module
The JNCC module uses an ensemble modelling approach including regression and machine learning algorithms in order to relate where species have been recorded to associated environmental variables (such as land cover and topography). This makes it possible to make predictions for other data across an entire region or country.
It is provided as open source code, published on GitHub, available for anyone to use with their own data, and at their chosen resolution. It can work as a stand-alone decision support tool or be integrated into natural capital evaluation tool.
Through the module, species presence records are modelled against environmental variables such as climate, elevation and land use. A suite of seven models are tested to evaluate and select the best modelled-estimate of species presence for each species individually. To better account for the random elements within the modelling, the process is repeated and the final map is a mean of the best performing models from each iteration.
Having predicted the probability of the presence of the species in the particular area of interest, the model can be used to explore how species distributions may be affected by an alternate state such as a change in land use or future climate.
NEVO
JNCC has been collaborating with the University of Exeter since 2017 on the use of its Biodiversity Module to provide the biodiversity component of a new natural capital tool known as NEVO (Natural Environment Valuation Online).
NEVO has been developed to provide policy makers and land-use decision-makers in England and Wales with an easy to use, map-based online interface with which to value existing and new land-use options, and to prioritise environmental interventions and investments. This works across multiple land uses and considers hypothetical changes in a wide range of natural capital services such as food production, water quality and quantity, greenhouse gas storage, and recreation. In addition to these, NEVO uses the species distribution framework in the JNCC module to consider the likely impacts of potential changes in land use on biodiversity, focused on 100 species of conservation priority in the UK ranging across mammals, birds, vascular plants, invertebrates, lichens and herptiles.
Other applications
The Biodiversity Module retains its original value as a standalone species distribution model framework. It has also been used in the marine environment, for example for modelling kelp (Laminaria hyperborea) as a provisioning ecosystem service, and on behalf of Defra to consider the biodiversity impacts of policy changes to agricultural subsidies by predicting the species distribution changes of approximately 800 priority species.
Potential future development includes the incorporation of Earth Observation data, in particular to explore if using this in parallel with indicator species may give an indication of habitat condition.
Team: Ecosystem Analysis