From point samples to continuous maps; remote sensing and biodiversity

A recent publication to come from some of the CLCR’s researchers has got me thinking on how biodiversity can be mapped and predicted across the Welland Catchment.

This post just captures some of my initial thoughts so that I can come back to them later. Bush et al. (2017) discuss an approach to converting point samples of species to continuous maps of alpha diversity (amongst other variables) using an approach they term CEOBE (connecting earth observation to biodiversity and ecosystems).

Several things from this study jump out at me that might be appropriate to applying similar techniques to the Welland Catchment. My original plan is to model habitat quality and threat using the InVEST toolkit. Here the model is paramaterised based on:

  • the suitability of a land cover / habitat to an individual or group of species (based on knowledge of that group, but still subjective)
  • the distance from / potential risk of possible threats (again not an objective value)

While I can see this output having a use, it remains subjective.

What would be great is a continuous map of alpha diversity (say farmland specialist passerines) across the whole of the Welland Catchment based on spatially explicit RS and point based data. Potential inputs into the creation of this map might be:

  • 1km gridded data from the BTO’s annual breeding bird survey; providing the point based species samples
  • High resolution RS data on NDVI from DEFRA
  • Gridded temperature and precipitation data
  • High resolution land cover data from processed Sentinel-2 imagery

Modern statistical techniques allow site by species matrix biodiversity data to be combined with site by environmental variable matrix data (perhaps taken from remotely sensed data-sets) (Bush et al. 2017). These techniques include:

  • joint species distribution models
  • community-occupancy detection models
  • generalised dissimilarity models

A baseline map of alpha diversity across the catchment would be an interesting baseline and could have useful applications to the project. Something for me to consider in more depth later…



Bush, A., Sollmann, R., Wilting, A., Bohmann, K., Cole, B., Balzter, H., Martius, C., Zlinszky, A., Calvignac-Spencer, S., Cobbold, C.A. and Dawson, T.P., 2017. Connecting Earth observation to high-throughput biodiversity data. Nature Ecology & Evolution.


Birds as ecological surrogates

In a broad sense, ecological surrogacy involves the assessment of one or more components of an environment or its biota, with the assumption that variation in the surrogate reflects change in another important, but difficult-to-measure attribute” – Westgate et al. (2017)

I am obsessed with birds, they brighten every single day of my life and give me an enormous amount of pleasure. For me, the intrinsic value of this taxonomic group is enough to justify continued efforts to monitor and conserve their populations. However are they of value to wider conservation and ecological studies?

One of the analyses I aim to perform as part of my PhD is an assessment of the habitat quality of the Welland Catchment using the InVEST module for habitat quality and risk. The model parameters need to be calibrated with a specific taxonomic group in mind to achieve the best results (Sharp et al. 2016) and naturally I am hoping to model habitat quality for passerine songbirds. While it is not an explicit aim of my research to scale these results to the wider ecological health of the catchment, it would be useful if the habitat quality for birds could be used to infer quality and threats for other taxa. Fortunately there is evidence in the literature to justify my choice.

 “Excellent knowledge exists of bird ecology and behaviour and they have great public resonance and so are good at raising awareness of biodiversity issues” – Eglington, Noble and Fuller, (2012)

Birds are the most extensively monitored taxonomic group within Europe, whilst worldwide they comprise 75% of biodiversity atlases (Eglington, Noble and Fuller, 2012). Up until 2013, bird population indexes formed part of DEFRA’s annual sustainability indicators (DEFRA, 2013).

Yesterday, the British Trust for Ornithology published its annual ‘State of the UK’s Birds report’, an example of the kind of long term monitoring that birds receive . The report has tracked population changes in the UK’s bird species since 1999 and identifies those of conservation concern (Red, Amber or Green). A brief explanation of how the levels of concern are determined can be found here.

The report is based, in part, on data from breeding bird surveys, carried out by volunteers twice a year in the spring and early summer. My Dad and I conduct the surveys for two sites, it is a great way for anyone with an interest in ecology and conservation to help contribute to long term data (more information on getting involved is here). This year’s report was the product of surveys from 2,600 volunteers, of whom a large proportion have monitored their sites for many years.

There were mixed headlines from the report, with an increase in the number of species now included on the red list. Worryingly, a quarter of all species assessed are now found on the list of species of highest concern (BTO, 2017). However 13 species moved from the amber to green list based on improvements in their status (BTO, 2017). The Tree Sparrow, Passer montanus, was the focal species of my MSc dissertation and I was pleased to see that their population is continuing to recover following near devastation at the end of the 20th century.

The theory behind their power as bio-indicators is that as birds utilise a wide variety of habitats and can be found at or near the top of their food chains they make useful general indicators of the state of wildlife (Eglington, Noble and Fuller, 2012). But doubts remain whether the accuracy of birds in this role is a benefit or hindrance to conservation. Two recently published studies address this and provide some support for my choice of taxa for the intended habitat quality assessment.

In a large meta-analysis, investigating 145 measures of effect size from the literature, Eglington, Noble and Fuller (2012) demonstrated that, on average, bird species richness explained 19% of the species richness of other taxa. An ecological explanation for this lack of strength is that birds have repeatedly been shown to respond to ecological resources on a landscape scale (e.g. Hardman et al. 2016) whereas plants and invertebrates respond to resources at much finer scales (Eglington, Noble and Fuller. 2012). For mammals, who respond to ecological resources at comparable spatial scales to birds, the strength of the relationship was greater.

While the average strength of the general relationship indicates that birds do not make good indicators of total species richness, the study highlights other aspects of birds as bio-indicators that do suggest utility. Effect sizes were shown to be much larger in studies that were based on heterogeneous, patchy landscapes, such as agricultural areas, when compared to studies in more homogenous landscapes such as grasslands (Eglington, Noble and Fuller. 2012). The Welland Catchment fits the bill for patchy and heterogeneous habitats, so inferences about other taxa from birds may potentially be more relevant for my landscape of study.

Westgate et al. (2017) also performed a meta-analysis examining the optimal taxa in terms of species richness and composition for use as ecological surrogates for other taxa. This study takes a different approach to Eglington, Noble and Fuller (2012), focusing on an analysis of patterns of pairwise, cross taxon congruence. Their key finding was that

birds and vascular plants outperform a range of alternative taxa as surrogates for the richness and composition of unmeasured taxonomic groups” – Westgate et al. (2017)

Results were shown to be highly dependent on the target of surrogacy and also the metric used for analysis but birds featured in many of the most powerful surrogate pairs for other taxa. Of all studies analysed, birds and mammals were shown to be optimal taxa for representing the composition of unobserved vertebrate taxa, possibly due to the ecological similarities mentioned earlier.

The most complex statistical analysis carried out by Westgate et al. (2017), involving the inclusion of study sample size and spatial variables, showed that birds were the optimal taxa for representing both the richness and composition of other vertebrates. For an assessment of richness and composition for all taxa, a combination of birds and plants was shown to be the most powerful surrogate.


Validation of these findings

Both studies included sensitivity testing of the results of their analyses to validate their utility. For Eglington, Noble and Fuller (2012) this took the form of a vote counting method which, although it does not include information on the strength of correlation, indicates that “significantly more studies have reported positive relationships between species richness in birds and that of other taxa” (Eglington, Noble and Fuller 2012) than have not (keeping in mind the ‘file drawer-effect of meta analyses (Eglington, Noble and Fuller 2012)).

Westgate et al. (2017) showed that the optimal surrogate identified by their analysis always had higher monitoring power than randomly selected surrogates, especially when specific taxonomic targets are identified over assessing ‘total’ biodiversity.



Eglington, Noble and Fuller (2012) state that birds will continue to be used as bio-indicators or ecological surrogates due to the wealth of knowledge on their ecology, their public resonance and the relative ease with which they can be assessed. While both studies discussed in this post highlight the potential of birds as indicators it is vital to be critical and explicit in terms of what taxa or component of a region’s ecology they are acting as surrogates for. This is because although birds often outperform other taxa as surrogates for diversity and composition, especially when specific taxa are targeted, they still account for relatively small variances in total diversity.




BTO (2017), ‘State of the UK’s Birds 2016’, British Trust for Ornithology, available at: (Accessed: 12/04/2017)

DEFRA (2013) ‘Sustainable development indicators’ Department for Environment Food & Rural Affairs, 2013

Eglington, S.M., Noble, D.G. and Fuller, R.J. (2012) ‘A meta-analysis of spatial relationships in species richness across taxa: birds as indicators of wider biodiversity in temperate regions’, Journal for Nature Conservation, 20(5), pp. 301-309

Hardman, C.J., Harrison, D.P.G., Shaw, P.J., Nevard, T.D., Hughes, B., Potts, S.G., Norris, K. and Marini, L. (2016) ‘Supporting local diversity of habitats and species on farmland: a comparison of three wildlife‐friendly schemes’, Journal of Applied Ecology, 53(1), pp. 171-180

Sharp, R., Tallis, H.T., Ricketts, T., Guerry, A.D., Wood, S.A., Chaplin-Kramer, R., Nelson, E., Ennaanay, D., Wolny, S., Olwero, N., Vigerstol, K., Pennington, D., Mendoza, G., Aukema, J., Foster, J., Forrest, J., Cameron, D., Arkema, K., Lonsdorf, E., Kennedy, C., Verutes, G., Kim, C.K., Guannel, G., Papenfus, M., Toft, J., Marsik, M., Bernhardt, J., Griffin, R., Glowinski, K., Chaumont, N., Perelman, A., Lacayo, M. Mandle, L., Hamel, P., Vogl, A.L., Rogers, L., Bierbower, W., Denu, D., and Douglass, J. 2016. InVEST +VERSION+ User’s Guide. The Natural Capital Project, Stanford University, University of Minnesota, The Nature Conservancy, and World Wildlife Fund

Westgate, M. J., Tulloch, A. I. T., Barton, P. S., Pierson, J. C. and Lindenmayer, D. B. (2017), Optimal taxonomic groups for biodiversity assessment: a meta-analytic approach. Ecography, 40: 539–548