simon linkerobert. l. presseyrobert c. baileyrichard h. norris
the ecology centreuniversity of queensland
australiawww.uq.edu.au/spatialecology
Identifying conservation priorities of catchments using irreplaceability, vulnerability and condition
three key questions in river conservation planning
Conservation valueConservation value
BiodiversityBiodiversity PressurePressure
ConditionConditionVulnerabilityVulnerability
StateState
Conservation valueConservation value
ConditionConditionVulnerabilityVulnerability
three key questions in river conservation planning
consider all three axes for planning
irreplaceability
vulnerability
highlow
high
condition
good
priority: protectionpriority: protection
priority: restorationpriority: restoration
modeled occurrences: probabilities!
assign a probability of occurrence for every taxon in every subcatchment
predictors: GIS
bailey & linke (in prep.) GIS variables predict macro-invertebrate assemblages as well as local habitat
query out for all subbasins: catchment descriptors climate geomorphology/
hypsology vegetation geology
generalized additive models
Environmental factors
30% chance of being at test site
Predicted Biota
70% chance of being at test site
modeling results
400 taxa at genus/species could be predicted successfully at ROC>0.6
0
20
40
60
80
100
50 250
450
650
8501050
1250
1450
1650
1850
Number of predicted occurences
Fre
qu
ency
irreplaceability
run heuristic 1000 times with randomly half of the sites taken out
see which catchments end up selected most often
measures: f(frequency of selection), c(contribution to targets)
irreplaceability
run heuristic 1000 times with randomly half of the sites taken out
see which catchments end up selected most often
measures: f(frequency of selection), c(contribution to targets)
irreplaceability
run heuristic 1000 times with randomly half of the sites taken out
see which catchments end up selected most often
measures: f(frequency of selection), c(contribution to targets)
83%83%
42%42%
13%13%
53%53%
irreplaceability
run heuristic 1000 times with randomly half of the sites taken out
see which catchments end up selected most often
measures: f(frequency of selection), c(contribution to targets)
agriculture
weedsroad density
nutrient load
grazing
forestry
sediment load
urbanization
condition -> stressor gradients
principal components principal components analysis (PCA)analysis (PCA)
condition -> stressor gradients
agriculture
weedsroad density
nutrient load
grazing
forestry
sediment load
urbanization
PC 1PC 1agricultureagriculture
PC 3PC 3forestryforestry
PC 2PC 2urbanurban
PC 1: agriculture (51% explained)
sediment load (0.36)intensive agriculture (0.41)native vegetation (-.42)acidification (0.37)grazing (0.40)forestry (- 0.40)
2 components
If land capability
slope
soils
allows more intensive use than current landuse
vulnerablevulnerable
capability classification(based on Emery (1985))
category 1 – highest capability:low slopes, low erosion and low salinity risk
suitable for cultivation, pasture, forestry
category 3 – low capability:steep slopes, high erosion and potentially high salinity
suitable for national parks
category 2 – medium capability:medium slopes, moderate erosion.
suitable for pasture, forestry
impact classification(after Norris et al. (2001))
cultivationhas a higher impact than
sown pasturehas a higher impact than
native pasturehas a higher/equal impact than
forestryhas a higher impact than
conservation
vulnerability by catchment
already protected -> not vulnerablealready in the highest impact class -> not vulnerable
focus on restoration
high irreplaceability, degraded conditionhigh irreplaceability, degraded condition
candidates for river reserves
high irreplaceability, still good condition, high irreplaceability, still good condition, but high vulnerabilitybut high vulnerability
challenge: integrated catchment planning
consider condition and vulnerability as variables that require cost/effort
priority of action is linked to effort needed targets can be met in multiple ways ->
choose the cheapest/easiest one
proposed framework
present condition
vulnerability
attributes of each catchmentattributes of each catchment
target 1
target 2
target n
subject to condition and vulnerability
aim: to optimize investments in condition and vulnerability so all targets can be met
reservation/’fighting threats’
restoration/improvement
possible types of action
Conditiongood bad
the connected nature of rivers (re-visited)
improvement or degradation ‘travels’ downstream
makes optimisation difficult (yet fun)
investment: restoration
what have I done so far?
adapted the simulated annealing algorithm to include different levels of investment
ran a trial with 3 (ficticious) species, 13 subcatchments, optimized for condition
simulated annealing gives you the optimal investment