Severn Estuary

51.585172, -2.660959

 

Study A

Study A

Birds

Grey plover Pluvialis squatarola

Golden plover Pluvialis apricaria

Lapwing Vanellus vanellus

Ringed plover Charadrius hiaticula

Oystercatcher Haematopus ostralegus

Curlew Numenius arquata

Black-tailed gotwit Limosa limosa

Knot Calidris canutus

Redshank Tringa tetanus

Snipe Gallinago gallinago

Turnstone Arenaria interpres

Modelling

MORPH

Abstract

Reproduced with permission under an Open Government Licence.https://www.nationalarchives.gov.uk/doc/open-government-licence/version/3/

This annex reports on individual-based modelling studies of waterbirds carried out by Bournemouth University (BU) in order to inform the Waterbirds Topic paper. Reports on bird surveys and habitat association modelling, undertaken by the British Trust for Ornithology (BTO), are reported in separate annexes. To help inform the Waterbirds Topic paper, this annex presents the results of modelling work aiming to assess the primary effect of loss of intertidal habitat and changes to the nature of intertidal habitat within the Severn Estuary on the numbers of the non-breeding waterbird receptors.

This project used the MORPH individual-based model, the latest in a series of such models developed for coastal birds. MORPH predicts how environmental change affects foraging animal populations. MORPH has been applied and tested for several systems, and has been shown to accurately predict the impact of environmental change. When parameterised for the Severn Estuary, MORPH followed the foraging decisions (i.e. patch and prey choice) of each individual bird, as they attempt to meet their daily energy requirements. The model predicted the distribution of birds between patches and the percentage of birds that survived to the end of winter.

For compatibility with the Habitat Association modelling Waterbirds Annex, the key prediction was percentage change in supported population size resulting from a tidal power option. This was calculated by initially predicting the population size of each species that could be supported in the baseline case (i.e. the population size at which survival equals 90%). The population size of each species that could be supported with each option was then predicted in the same way, from which the percentage change from baseline could be calculated.

In the baseline simulations, for all species except curlew and oystercatcher, the maximum population size that could be supported was greater than the population size during 2008-2009. This implies that surplus food was available for these species, and that the duration of exposure was long enough to allow these species to meet their daily energy requirements throughout the winter. There was a general tendency for the larger bird species to be either be  supported at current population sizes (Curlew and Oystercatcher), or to be very close to carrying capacity (Black-tailed Godwit). These species have higher energy requirements and require larger prey items (which are generally lacking on the Severn) to meet their daily requirements.

The model was tested by comparing the observed and predicted mean overwinter distribution of each species across the estuary. Although the observed data were used to determine the initial distribution of the birds, model individuals moved to different feeding patches if they were unable to meet their daily energy requirements by feeding in their initial patch. In the baseline simulations, with the exception of Oystercatcher, Curlew and Black-tailed Godwit, the observed and predicted numbers of each species on each patch were virtually identical. This means that most model individuals of most species were able to obtain their overwinter energy requirements by feeding in the same locations as the real individuals. In the case of Oystercatcher, Curlew and Black-tailed Godwit, food resources were more limiting and so individuals were less able to meet their requirements by feeding in the same locations as the real individuals. Due to the uncertainties associated with the predictions for Oystercatcher and Curlew (low predicted survival and poor distribution predictions) the predictions for these species are not discussed in the main Waterbirds Topic paper.

Model parameters were derived from the following sources: (i) The outputs of the Hydrology and Geomorphology (e.g. tidal exposure of patches), Marine Ecology (e.g. habitat type) and Waterbird (e.g. bird numbers) topics; (ii) A previous intertidal invertebrate survey of the Severn Estuary (numerical density of prey); (iii) Data collected on other sites (e.g. energy content of prey); (iv)General relationships thought to apply across sites (e.g. feeding rate of waterbirds). Even though all parameters were derived from measurements (either on the Severn, other sites, or across a range of species) the confidence that can be placed on these (as descriptions of the real system) will vary.

Generally more confidence can be placed on parameters that are measured on the Severn itself as close as possible to the present day. The developed model is the best representation of the current system achievable with the data currently available. Further Severn-estuary specific studies, particularly of prey abundance and biomass, will be required to increase confidence that the model more closely describes the current system.

The following two sets of simulations were run, based on different assumptions of how tidal power options would effect invertebrate productivity. Predictions were made both for the short-term (i.e. ignoring long-term mud deposition or erosion) and the long-term (i.e. including long-term mud deposition or erosion).

Severn Estuary invertebrate simulations:These simulations assumed that mudflats outside and within tidal power options (i.e. upstream of a barrage or within a lagoon) were occupied by the invertebrate species community recorded on the Severn.

Non-Severn Estuary invertebrate simulations:These simulations assumed that mudflats within tidal power options (i.e. upstream of a barrage or within a lagoon) were occupied by the invertebrate species community recorded on other south-western UK estuaries (increasing

prey biomass relative to the Severn Estuary invertebrate simulations). Mudflats outside of tidal power options were still assumed to be occupied by the invertebrate species community recorded on the Severn.

For all species except curlew and oystercatcher, the maximum population size that could be supported was greater than the population size present during 2008-2009. This implies that surplus food is available for these species, and that the duration of intertidal exposure was long enough to allow these species to meet their daily energy requirements throughout the winter.

There was a general tendency for the larger bird species to be either be unsupported at current population sizes (curlew and oystercatcher), or to be very close to carrying capacity (black-tailed godwit). These species have higher energy requirements and require larger prey items (which are generally lacking on the Severn) to meet their daily requirements.

The B3 Cardiff-Western barrage had the largest negative effect on the bird populations. This barrage reduced mudflat area and reduced the duration of exposure of most patches. The total food resource for birds was therefore reduced as was the time birds had to feed. Relatively large amounts of longterm erosion were predicted for the B3 Cardiff-Western barrage and this further reduced the population sizes of birds that were predicted to be supported. Assuming non-Severn Estuary invertebrate communities increased the number of birds supported by this option, but it was still predicted to have a more negative effect on bird populations than other options.

The B4 Shoots and B5 Beachley barrages and the L2 Welsh Grounds lagoon had very similar negative effects on the size of bird populations supported. These options either had no effect or decreased the area of mud within the patches they enclosed, and reduced the duration of exposure within the patches they enclosed. These options had a double negative effect on the birds as they both reduced the total amount of food available and reduced the time available for feeding. Relatively small amounts of erosion were predicted for these options and so short and long-term predictions were similar. Assuming non-Severn Estuary invertebrate communities had little effect on the population sizes of birds predicted to be supported with these options.

The L3 Bridgwater Bay lagoon decreased the mud area but increased the duration of exposure within the patch it enclosed. Despite this, when assuming Severn estuary invertebrate communities, this option had a slightly more negative effect than the B4, B5 and L2 options. This implies that the size of the bird population that can be supported is more sensitive to variation in mud area than it is to variation in the duration of exposure. Relatively small amounts of erosion were predicted for this option and so short and long-term predictions were similar. Assuming non-Severn Estuary invertebrate communities increased the population sizes of birds predicted to be support, and in these simulations the L3 Bridgwater Bay lagoon was predicted to support larger population sizes of birds than were predicted in the baseline simulations.

Three species, Dunlin Calidris alpina, Curlew Numenius arquata and Redshank Tringa totanus, are named in the Special Protection Area (SPA) designation for the Severn Estuary. If the Severn Estuary is to remain in favourable condition from an SPA perspective, any tidal power options should still allow at least these population sizes of these species to be supported. The baseline simulations predicted that this was only the case for Redshank. For Dunlin and Curlew, the food supplies included in the baseline model were predicted not to be sufficient to support the population sizes of these species present at the time of SPA designation. Most tidal power options reduced the population size that could be supported. In these cases, the prediction was that the food supplies included in the model, in combination with the tidal power options were not sufficient to support the population sizes of Dunlin and Curlew. The food supplies were sufficient to support the SPA population size of Redshank in all cases. The L3 Bridgwater Bay lagoon in combination with non-Severn estuary invertebrates increased the population sizes of all of these species. In this case, the prediction was that the population sizes of Dunlin, Curlew and Redshank present at the time of SPA designation could be supported by the food supplies included in the model.

Funding and Collaboration

Parsons Brinckerhoff Ltd, Black and Veatch Limited, Department for Energy and Climate Change

 

Study B

Study B

Birds

Grey plover Pluvialis squatarola

Ringed plover Charadrius hiaticula

Oystercatcher Haematopus ostralegus

Curlew Numenius arquata

Black-tailed gotwit Limosa limosa

Dunlin Calidris alpina

Knot Calidris canutus

Redshank Tringa totanus

Turnstone Arenaria interpres

Modelling

MORPH

Abstract

With the pressures that today’s ecosystems are being placed under, from both environmental change and anthropogenic developments, the speed at which management decisions need to be made has increased. Coastal development means that estuaries are particularly affected and their characteristic species, like wading birds (Charadrii), are now experiencing worldwide declines. In such situations there is a need for predictive ecology to understand in advance how species might react to future changes.

This thesis looks into how we can use individual-based models (IBM) to make accurate predictions of how wading birds are affected by environmental change. Starting with previously validated models I show the importance of measuring size of invertebrates though an IBM investigation into regime shifts and wading birds responses. The models show that by altering their diet preferences, birds adapt to regime shifts in their prey but that this maintenance of population size masks the true changes in the system and limits the use of waders as direct bio-indicators of ecosystem health. Using the current literature, an analysis on empirical responses of wader populations to environmental change revealed the lack of comparability between studies and the scarcity of studies on small scale events.

Data from literature and fieldwork was used to develop a comparable suite of individual-based models for five UK estuaries with up to eleven wading bird species. These models were validated using current BTO Wetland Bird Surveys data to increase confidence in final results. Using these new models, investigations of population thresholds and environmental change were carried out. Increases to current populations revealed that several estuaries are no longer able to support the number of birds around the time of Special Protection Area designation. This, alongside higher populations currently seen since the years of designation, indicates the need for re-assessment of SPA species numbers. When looking at the impacts of two types of environmental change, habitat loss and sea-level rise, certain species declined predictably across sites whilst the individual make up of each estuary had particular impacts on some waders more than others. The work of this thesis further indicates the great potential of using individual-based models to predict the effects of a wide range of environmental changes. With the new models and a quicker and systematic way of developing IBMs for additional areas, we can aid the conservation and management of estuarine systems for wading birds.

Funding and Collaboration

Bournemouth University and HR Wallingford
 

 

Related Paper(s):

Study A
Bournemouth University, 2010. Severn tidal power – Sea topic paper. Waterbirds. Annex 3 – Waterbird Individual based modelling. Poole, Dorset, Bournemouth University.

 

Study B
Bowgen, K.M., 2016. Predicting the effect of environmental change on wading birds: insights from individual-based models. PhD thesis, Bournemouth University in collaboration with HR Wallingford.