DEVELOPMENT OF INDIVIDUALBASED MODELS IN SHOREBIRDS

1
DEVELOPMENT OF INDIVIDUAL-BASED MODELS IN
SHOREBIRDS
2
WASH 1972-75
High Water Mark
Low Water Mark
Intertidal mudflats and sandflats
Proposed fresh-water reservoir
3
HOW WE THOUGHT ABOUT THE ISSUE
Percentage starving over winter
Decreasing area Increasing density Increasing
competition
4
EMPIRICAL APPROACH?
(i) It would be technically very difficult
?
Percentage starving over winter
0 0 0 0 0 0 0 0 0
X X X X
X X
X X
X X
?
Increasing density
5
..and (ii) the function would probably change
after habitat loss
Habitat of above average quality is lost
Percentage starving over winter
Habitat of average quality is lost
Increasing density
6
THE QUESTION HOW TO DESCRIBE THE
DENSITY-DEPENDENT STARVATION FUNCTION - NOT ONLY
AS IT IS AT PRESENT BUT HOW IT WOULD BE IF THE
FEEDING ENVIRONMENT WAS CHANGED BY Habitat loss
Disturbance Shellfishing Sea-level
rise Mitigation measures etc
etc
7




EVENTUAL ANSWER develop and test
individual-based models using oystercatchers
eating mussels on the Exe estuary
8
(No Transcript)
9
MUSSELS
LOCAL ISSUE WOULD HARVESTING MUSSELS AFFECT THE
BIRDS?
10
NO EFFECT EFFECT
Before After
Percentage starving
Increasing harvest/Decreasing food supply
11
DEVELOPMENT OF THE MODEL 1976-1996 Field work on
interference and exploitation competition between
oystercatchers for mussels
12


13


14
(No Transcript)
15
HOW THE MODEL WORKS
MORE DETAILS AT http//www.dorset.ceh.ac.uk/shore
birds/
16
Each bird decides each tide where, when and on
what prey species it is best to feed
http//www.dorset.ceh.ac.uk/shorebirds/
17
Each of the three displaced birds will choose the
next best place in which to feed
PRINCIPLE birds in model use optimality decision
rules ( fitness maximising) to decide how to
respond to a change in their feeding environment
just as real birds do
18
Calibration period for overwinter mortality of
adult mussel-feeding oystercatchers on the Exe

Sept 1976 Mar 1980
19


MORE NATURAL HISTORY WAS NEEDED eg. feeding in
fields over high tide
20
Predicted (retrospective) and observed increase
in mortality 1980 - 1999
Calibration period
21
CONCLUSION Winter mortality was
density-dependent and the model postdicted it
quite well
22
Applying the model to species other than
Oystercatchers some say that they take too long
to parameterise
NO LONGER SO!
The models can usually be built, tested and
applied within the time typically taken to
conduct an EIA i.e. 1 3 years
23

• Bird energetics
• Prey energy content
• Functional responses
• Interference functions
• Food supply, exposure time, weather
• Human activities eg. fishery

MODELS CAN NOW BE BUILT AND TESTED VERY QUICKLY
AS MOST PARAMETERS ARE IN THE LITERATURE
Obtained for the site being modelled Built into
model allometric functions
24

• Applications to other species what
• we need to know
• Bird energetics
• Prey energy content
• Functional response
• Interference function
• Food supply, exposure time and weather

APPLICATION TO OTHER SPECIES - 1
25
Functional responses of oystercatchers eating
mussels
A
26
Predicting the asymptote of the functional
response in shorebirds from 486 spot estimates
of intake rate

• Intake rate (i.e. asymptote) depends on
• Body mass of bird
• Mass of prey
• R2 75.5 (log transformed data)

27
TEST OF PREDICTIONS FOR ASYMPTOTE
y x
Curlew sandpiper Knot Redshank Grey
plover Curlew Oystercatcher
28

• Applications to other species what
• we need to know
• Bird energetics
• Prey energy content
• Functional response
• Interference function
• Food supply, exposure time and weather

APPLICATION TO OTHER SPECIES - 2
29
Interference function for stabbing
oystercatchers intake rate (mgAFDM/s) against
density of birds
30
Predicting interference parameters from
Stillmans state-dependent behavioural model
Prey size
Bird size
Handling time
Running speed
Interference function
31
Test of Stillmans behaviour-based model
cockle-eating oystercatchers on the baie de Somme
32
Predicted and observed interference functions in
cereal-feeding cranes. Predictions from
Stillmans model. Data L. M. Bautista, J. C.
Alonso J. Alonso.
33

• Applications to other species what
• we need to know
• Bird energetics
• Prey energy content
• Functional response
• Interference function
• Food supply, exposure time and weather
• Human activities

APPLICATION TO OTHER SPECIES and SYSTEMS
What do we need to measure?
34
RECENT DEVELOPMENTS

• Risk of being killed by predators
• trade-off between foraging and safety
• Multi-species models
• up to 9 species in one estuary
• Multi-site models
• several estuaries across several countries

35
SOME EXAMPLES OF RECENT APPLICATIONS IN SHOREBIRDS
Oystercatchers Shellfishing, disturbance 8
estuaries 4 countries Other shorebirds
Disturbance, habitat loss, 12 species bait-dig
ging, Spartina 10 estuaries encroachment,
hunting 4 countries sea-level rise,
monitoring estuary quality, mitigation Wildfow
l Hunting, farming, wind 4 species farms
gt25 estuaries 4 countries
Blue - examples used here
36
1. OYSTERCATCHERS and SHELLFISHING
How much shellfish should we leave after
shellfish harvesting to ensure that the birds
fitness is not reduced?
37
EXE oystercatchers eating mussels
Mortality
Fail to reach target body mass
With disturbance
Range 1976-1999



No disturbance


0 55 110
165 kg/bird
0 55 110
165 kg/bird
Prey biomass/bird left over after shellfish
harvesting
38
CRITICAL THRESHOLD or ECOLOGICAL FOOD
REQUIREMENT (food/bird)
EXE ESTUARY
Critical threshold 61kgAFDM/bird
Percentage starving
39
 
Ecological requirement
Multiples of
(threshold) kgAFDM/bird Exe
Mussel 61 9.13
7.74 Bangor Mussel
50 9.62
6.42 Burry Cockle 44
9.27 5.58 Wash
Cockle 20 7.93
2.52 Somme Cockle 33
6.56 5.03 Taking
into account wastage of shellfish flesh (stolen
from the birds winter loss of flesh from the
shellfish themselves)
Critical threshold or Ecological
requirement (E) Physiological requirement
(P) kgAFDM/bird kgAFDM/bird
E/P
40
CONCLUSION ecological requirements are 3-8
times larger than physiological requirements
Because of interference and individual variation
in efficiency, just leaving enough shellfish for
oystercatchers after harvesting is not enough to
ensure they survive in good condition - as
Dutch experience confirms
41
2. DISTURBANCE people and raptors
42
SOMME DISTURBANCE IN OYSTERCATCHERS
ACTUAL AMOUNT 1994-95 1996-97 1997-98 1997-98
including raptors
Severe winter
PERCENTAGE STARVING
Mild winter
DISTURBANCES PER HOUR
43
POLICY ADVICE IN SEVERE WINTERS, DO NOT ALLOW
OYSTERCATCHERS TO BE DISTURBED BUT IN MILD
WINTERS, THEY CAN BE DISTURBED but only up to
about ONE disturbance/hour - including raptors
44
3. SPREAD OF THE GRASS Spartina ON TO UPSHORE
MUDFLATS
45
SOMME Spartina spreads downshore at 20-40m
per year into the feeding areas of dunlin
High Water mark
46
Dunlin with Spartina spreading at 40m/year
Mortality over winter
POLICY ADVICE get rid of it!
0ha 100ha
200ha 0 years 5
years 10 years
47
4. HABITAT LOSS and MITIGATION for it
48
PORT DEVELOPMENT SEINE ESTUARY
PORT (105 ha of mudflats)
English Channel
R. Seine
High Water Mark
MITIGATION Convert reed beds into 50 or 100 ha
of mudflats
49
DUNLIN in SEINE ESTUARY Port development and
proposed mitigation
Before Port
After Port
50ha mitigation 100ha mitigation
5 0
15
Failing to reach 75 body mass
Mortality
0
Scenario
Scenario
50
POLICY ADVICE The mitigation is needed but
should be 100ha if it is to fully effective
51
CONFIDENCE IN MODEL PREDICTIONS?

• TESTS of model predictions
• Distribution and prey choice
• Impact on food supply
• Hours feeding per daylight tide
• 4 Mortality of oystercatchers

52
1. DISTRIBUTION and PREY CHOICE proportion of
birds eating cockles or mussels in the Burry Inlet
Observed Predicted



Percentage of birds
53
2. IMPACT ON FOOD Mussel consumption by
oystercatchers on the Exe
Observed
Predicted
Exclosure Survey
Percent
54
3. HOURS FEEDING oystercatcher (open symbols),
little stint, sanderling, dunlin, curlew. Exe
estuary, Burry Inlet, Bangor flats, Seine
estuary Cadiz bay.
55
4. MORTALITY (from autumn to spring)
(se) (se) N
Predicted Observed Wash shellfish abundant
4 0 1.4 (0.3) Wash
shellfish scarce 3 15.9 (4.9) 16.8
(6.4) Exe low bird density 3 2.3 (0.4)
1.7 (0.3) Exe high bird density 3
3.8 (0.1) 3.5 (1.0) Burry Inlet cockles and
mussels 1 0 ? Bangor
mussel beds 1 0
? Number of years Assuming upshore intake
rates same as on Exe
56
(No Transcript)
57
FIND OUT MORE ON
http//www.dorset.ceh.ac.uk/shorebirds/
Co-workers Richard Stillman Richard
Caldow Sarah Durell Andy
West
58
EVEN A SMALL DECREASE IN MORTALITY COULD BE VERY
IMPORTANT!!
59
..population size is very sensitive to mortality
rate, irrespective of the density-dependence on
the breeding grounds
Weak density dependence in summer
Equilibrium population size
Strong density dependence in summer
Oystercatcher range
60
The model had to be able to predict FITNESS

• Mortality rate over the winter
• BUT ALSO
• 2 Fat reserves in spring

PRINCIPLE If the FITNESS of the birds in the
non-breeding season is maintained, the quality of
the estuary for the birds is being maintained