Lecture 11: Systemic conservation planning Flashcards
(23 cards)
Lecture outline
1.Intro to Systematic Conservation Planning
2.Background to Site Prioritisation Algorithms (SPAs)
3.Worked example of Site Prioritisation
4.Application of SPAs to planning protected area networks now
5.… and under predictions of future climate change
Site Prioritisation algorithms (SPAs)
Algorithm: a step-by-step procedure for calculations. A finite list of well-defined instructions
Site prioritisation algorithms (SPAs) are designed to select optimal sites in a reserve network based on pre-defined criteria, and within pre-defined constraints.
Two approaches:
- the minimum area approach (or set covering approach)
*requires all species to be represented within a reserve network that has the smallest possible area - the maximum coverage approach
*requires the maximum number of species to be protected in a reserve network that is limited by a cost
There is no simple method for finding an optimal solution
because the addition of an optimal site will depend on the current configuration of the protected area network and the remaining areas outside the network
Approaches to site prioritisation
Richness algorithms (also known as greedy algorithms): start by selecting the sites that hold the greatest diversity available and then add sites that provide the next most diverse site.
Complementarity: introduced by Vane-Wright et al. (1991). Defined by the addition of sites that complement the previously selected sites by adding the greatest number of new species to the network.
The rarity approach: selects sites depending on their irreplaceability in terms of the rarity of the species, or other factors such as endemism, found within that site.
The latter approach reduces the total area of a network as essential sites that hold the only population of a species may hold other species that need not be represented again in the network»_space; reduces redundancy.
However, using a ranking system sequentially to make decisions is flawed.
Prioritising based on rank scores is susceptible to leaving out some of the conservation targets and over-representing others (Pressey 2002).
Goal: to represent each species at least once in the fewest no. of sites
See notes for approach realisation
Optimality: defined by the representation problem used
For minimum area
*The optimal solution will be the network that protects all species with the least area
For maximum coverage
* The optimal solution will protect the greatest proportion of biodiversity whilst restricted by a set cost.
Early algorithms: Selected sites on their intrinsic values:
(i) Terborgh and Winter (1983): chose sites that were identified as hotspots of endemism so as to create a cost-effective reserve network. I.e. which parts of the world have the greatest diversity of animals that occur nowhere else.
This method is likely to create a large overlap in protected species between reserves and, therefore, redundancy within the reserve network
(ii) Kirkpatrick (1983): used an iterative step-wise approach that recalculated the relative diversity of each remaining site after each stage of selection. i.e. a richness algorithm that incorporates the principle of complementarity (I.e. improving additional site selection)
Two approaches to SPAs
Heuristic approach: selective step-by-step approach
* Although they are designed to achieve efficiency, heuristics rarely reach, or can detect if they have reached, an optimal solution.
Simulated annealing: an iterative optimisation process
* Starts with a randomly determined reserve network, and then explores trial alternatives by randomly adding or deleting new sites at each iteration.
Simulated annealing reaches optimal solutions more often than heuristic
However, optimality is theoretical and often not practical in the real world
SPAs and conservation planning
Cowling et al. (2003): compared the choices of conservation planners against those made by systematic reserve selection programs.
*Conservation planners make biased decisions when choosing reserves. Human conservation planners are biased – but some of these biases are important and not accounted for by algorithms
*Biases stem from considerations not important to the algorithms (such as pre-emption of risk), but which can lead to over-representation of some species within a reserve network and under-representation of others.
Remember: reserve selection algorithms are meant to act as a tool for conservation planners & NOT a replacement for them
Incorporating other factors into SPAs
Species presence is not the only factor affecting the optimal choice of reserves.
Other very important factors include:
*Land availability
*Land price
*Area of protected area
*Minimal viable populations for key species (& metapopulation considerations)
*Representation required for individual species
*Other areas already protected
*Connectivity
Systematic conservation planning
1.Target driven
2.Efficient
3.Minimises conflict with other land-users
See: Systematic conservation planning Margules and Pressey (2000) Nature
See example in notes:
Site selection based on rarity
The concept of irreplaceability
*Each planning unit can be given an irreplaceability score based on the extent to which it could be swapped for another.
*Irreplaceable units are always needed to meet the targets.
*Units with low irreplaceability scores are still often needed to meet the targets. Their low score only means that there are many other units that contain the same biodiversity.
see figures in notes:
Area 1 is irreplaceable because it is the only area that can contribute the tortoise, as well as contributing nearly all of the other species (except fish)
*Area 2 contains the toad and butterfly, which don’t add anything (already included in Area 1)
*BUT, Area 2 contains that missing fish so it could be included in portfolio
…as could Area 3 – again, the bug doesn’t count but this Area has the missing fish
*Thus, Area 1 is irreplaceable because it’s the only Area with mouse and tortoise, and also includes some other stuff
*Area 2 and Area 3 are interchangeable because they can both contribute the missing fish
*One of Area 2 or 3 must be included, so they still have value it’s just an equal value
Maximum irreplacability here is 1 species / 2 sites = score of 0.5 for each site
Systematic conservation planning aims
Aims also to minimise conflict with other land-users e.g. agriculture
Each planning unit can be given a value based on its value for other land-uses (e.g. for agriculture or development).
Software identifies conservation landscapes that meet the targets, whilst minimising the conflict with other uses/user groups.
We must consider cost associated with how much land would be worth if converted to agriculture (see notes)
Targeted protected area expansion
In 2010 in Nagoya (Japan), the Conference of the Parties (COP) adopted 20 headline targets:
Aichi Biodiversity Targets:
*By 2020, at least 17% of terrestrial and inland water, and 10 % of coastal and marine areas conserved through protected areas
In 2022 in Kunming-Montreal COP, 23 new targets:
*By 2030, at least 30% of terrestrial, inland water, marine and coastal areas, conserved through protected areas
*Protected area expansion to meet 30% target is a Site Prioritisation problem
The current status of the PA network
Aichi Target 11 was partially met:
BUT current network biased towards cheap sites – wrong prioritisation algorithm!
Not targeted towards species in most need of protection
E.g. not considering risk, diversity, irreplaceability, ecosystem services…
Resulting in ‘paper’ protected areas which don’t exist in reality
Or
‘Protection’ of areas too remote to be at threat e.g. summit areas
30% terrestrial mammals, amphibians, and birds have < 10% of their habitat within PAs
^ increase in protected area is not providing the benefits expected due to inaccurate allocation
*Most of these are threatened species at high risk of extinction
*91% of threatened species have insufficient representation within PAs
Yiwen, Senior et al. (2023) Science Advances
Targeted protected area expansion
*Cost of PA expansion is minimised and benefit maximised by using systematic conservation planning (Venter et al. 2014, PloS Biology)
*Many ways to prioritise biodiversity
*Cost is not shared equitably
See figure from Yiwen, Senior et al (2023) Science Advances
*Some of the poorest countries could make the biggest contributions to habitat protection (biggest circles) but due to low GDP they cannot make this investment
*Though there are some wealthy countries that can make a big difference e.g. USA
Protected area expansion under climate change
ways to mitigate impacts of climate change strategically through PA network expansion:
1.Buffer zones around reserves
2.Corridors between reserves
3.Stepping stone reserves
4.Protect key stable sites (climate ‘refugia’)
(see prev. lecture)
“We can summarise the essence of what needs to be done in four words: more, bigger, better and joined” Lawton et al. (2010) Making Space for Nature: a review of England’s wildlife sites and ecological network. Report to Defra.
Integration with species distribution models
Future distribution tells us:
*Risk – which species have little future habitat or can’t reach it
*Routes – how will species reach future habitat
*Refugia – which sites are important for retaining & gaining species
> > where new PAs are needed to act as:
(1) buffer zones
(2) corridors
(3) stepping stones
or (4) to protect climate refugia
SDMs (Species distribution models) can be used to identify where species’ distributions are going to be in the future. And can be used in systematic conservation
See Mi et al (2023) Science Communications
^ A recent example where SDMs were used to identify regions that currently have high species richness and are climate robust (blue cells)
*They identify as gaps the places where those blue cells fall outside of PA network
*This layer could then be easily combined with other factors, such as a cost layer
Limitations of Species Distribution Models (SDMs): imperfect inputs and outputs
SDMs are powerful, but all models have limitations
‘all models are wrong but some are useful’
Imperfect inputs:
1.Species distribution data:
*Errors of omission – species not recorded where it does exist
*Errors of commission – species recorded where it does not exist
*Spatial resolution matters
*Presence/absence vs. Abundance - not just where they occur but where they occur at the highest density most of the time
2.Environmental data:
*Errors (e.g. habitat classification)
*Spatial resolution matters
*Temporal resolution matters
Imperfect outputs
*Outputs can vary depending on model parameters
^ E.g. type of model, occupancy threshold, lots of arbitrary decisions
*Outputs are rasters (gridded data), the world isn’t
^ Spatial mismatch with real world conservation
Limitations of Species Distribution Models (SDMs): Issues with statistics
*Insufficient data to parameterise the model
*Missing key explanatory variable(s) e.g. other species, habitat quality, human pressure, snow cover/presence for example is a factor that changes distribution throughout the year
*Violation of model assumptions (not all stats appropriate for all data)
Assumption that distribution is in equilibrium with environment
(realised niche = fundamental niche) this is not usually the actual case
E.g. suitable habitat hasn’t yet been colonised
E.g. suitable habitat might encompass conditions that don’t yet exist
This results in extrapolation beyond bounds of the input data e.g. climate change prediction data
See: Trew et al. (in review) https://www.researchsquare.com/article/rs-3272916/v1
^ Appearance of novel climates in the tropics predicted according to climate change
These climates will never before have been seen in modern times and we do not know how they will respond
mitigating climate change is not just about species range shifts
Mitigating climate change is not just about species range shifts
We focus on range shifts because they are easy to model and observe
^ it can be modelled from fossil records
The other aspects identified by Scheffers are hard to detect from fossils so have been less focused on despite their importance:
Genetics:
*Adaptive evolution to heat stress in small organisms with short generations
*Limited evidence for adaptive evolution in higher level vertebrates and trees
*Increased hybridization
Physiology:
*Increases in coral disease
*Extensive dissolution of carbonate-shelled
organisms and corals
*Direct mortality from heat stress
*Changes in life history traits and migration timing
Morphology:
*Decreases in body size and changes in shape
*Changes in colour and brightness
Phenology
*Changes in spawning times of marine and freshwater fish
*Earlier budding and flowering in
plants and earlier growing season
*Early and later migration in birds
*Increased asynchrony
Dynamics:
*Changes in recruitment and age structure
*Changes in abundance of reef-building corals, plants, mammals, and birds
*Changes in sex ratio
Distribution:
*Latitude and altitude range shifts
*Range expansion and contraction
*Loss of habitat
Interspecific relationships:
*Tropicalization of temperate
ecosystems
*Borealisation of Arctic ecosystems
*New competitive interactions among species
*Desynchronization among dependent species
Productivity:
Changes in net primary productivity on land
Changes in phytoplankton biomass in marine and freshwater ecosystems
Info from: Scheffers et al 2016
The broad footprint of climate change from genes to biomes to people.
Science. DOI:10.1126/science.aaf7671
at:https://www.science.org/doi/10.1126/science.aaf7671
Conservation prioritisation: A Durham perspective
How to most efficiently protect key biodiversity across the university estate?
Which are the key species?
Where are the key areas?
Where is conservation most cost-effective?
11.2 computer workshop SPA modelling
The workshop will guide you through the process of creating and running a series of SPAs to try to optimise protection of key wildlife across the land that Durham University own and manage.
In the workshop we will be using a tool in Excel to run through protecting parcels of land based on some very simple starting objectives. This will give you some experience of developing a basic SPA.
RSAs and BAP species in the UK
See figure from: Franco et al. (2009) Surrogacy and persistence in reserve selection: landscape prioritization for multiple taxa in Britain. Journal of Applied Ecology, 46, 82–91.
Using the area-prioritization algorithm ZONATION to identified the locations where UK Biodiversity Action Plan (BAP) species of mammals, birds, herptiles, butterflies and plants occur in concentrated populations with high connectivity
Biodiversity Action Plan (BAP – high priority) species of one taxonomic group were not particularly good surrogates (indicators) for BAP species of other taxonomic groups. Hence, maintaining population concentrations of one taxonomic group did not guarantee doing likewise for other taxa.
Species with narrow geographic ranges were most effective at predicting conservation success for other species, probably because they represent the range of environmental conditions required by other species
Replacing underperforming protected areas
- Inefficiency Of PA system Widely acknowledged
- Conservation performance of PA systems have been shown potentially to be radically improved by replacing some inefficient current PAs with much more efficient new ones.
Cost effectiveness can be measured as the
benefit divided by the cost
See: R.A. Fuller et al. (2010) Nature, 466, 365-367
