inst/ms/discussion-notes.md

In seananderson/metafolio: Metapopulation Simulations for Conserving Salmon Through Portfolio Optimization

discussion notes

from abstract:

Linking our model to real-world salmon conservation

Link simulation and results to real issues for PNW salmon

• salmon management conservation is highly fragmentary - dispersed among hatchery managers, who for example are trucking and releasing smolts all over the downstream catchment into San Fransisco bay - no planning about whether that's a sensible strategy given climate
• Examples of response diversity, specifics that match up with modelling framework
• maybe Doug's paper - genetic diversity is best predictor of response diversity
• Range contraction
• temp. tolerance - spatial contraction of pops - one would predict increased variance and risk - losing response diversity
• Non-random extinction
• really bad or really good, but not random
• e.g. salmon in upper watershed tend to be impacted more by dams
• Climate change and response diversity
• those that seem productive and/or stable now may not be in the future
• by eliminating large spacial areas (e.g. damns) you're creating a higher risk ecological portfolio, if you get lucky you'll be OK, you might do much worse
• @cheung2013 and Mantua papers
• Stock-recruit Fredu paper cod and Benjamin Plonque CJFAS
• @rogers2013 long-term fluctuations
• @thorson2013 - show that spatial variation buffers fluctuations of juvenile salmon for Chinook in Snake River

Our salmon model is relevant to PNW salmonid guidelines and mandates in the US and Canada. In the United States, NOAA consider salmon in Viable Salmonid Populations and Evolutionary Significant Units, which could be considered as portfolios and assets in our model. Viable Salmonid Populations refer to XX, and Evolutionary Significant Units to XX.

Our recommendations policy guidelines for VSPs. In particular, VSP guidelines suggest... Likewise our results suggest...

In Canada, the Department of Fisheries and Oceans Wild Salmon Policy designates management units, and evaluates different levels of management intervention.

focuses on maintaining Conservation Units "protect the genetic and geographic diversity of salmon species" "policy recognizes the importance of habitat protection and rehabilitation" The Wild Salmon Policy recognizes the importance of genetic and geographic diversity as well as the protection and rehabilitation of habitat (REF) WSP will provide [from WSP itself] "Healthy,diverse,and abundant wild salmon populations for future generations" while providing fisheries and accounting for ecosystem and habitat values

or WSP end goal: "Restore and maintain healthy and diverse salmon populations and their habitats for the benefit and enjoyment of the people of Canada in perpetuity"

Criteria

Link to existing guidelines and mandates in Canada and the US

What new do we learn for salmon management and conservation?

• All recommendations for maintaining response diversity from @moore2010 (summarize those briefly?)
• If closer populations are more correlated with each other, and you must prioritize, then focus on a selection that aren't spatially contiguous (as long as straying is maintained) to maximize diversity of response to the environment
• Precautionary management: we assume precautionary abundance should we also be precautionary about response diversity?
• In many ways our results are intuitive, but we don't think about aggregate risk much, but there are many benefits to considering aggregate risk
• another dimension of conservation - or persistence
• not just leave all dams in place... reinvigorate selection for response diversity; double whammy of damns - hatcheries added, further removing response diversity
  • Scale of management; integrated management planning

Prioritizing metapopulation conservation

Highlight how our work complements other simulation-based paths to prioritization

• shiraz - single population - [@scheuerell2006]
• @halupka2003 - salmon prioritization scheme based on adaptive variation...
• @allendorf1997 - ranking scheme based on PVA and extinction consequences
• @haight2008 - scheme based on dispersal probabilities
• @pess2012 - based on Andy's paper
• I have others...

There are a number of specific extensions to our simulation that could be useful.

• We're only looking at one dimension of response diversity - thermal tolerance. All sorts of other kinds - climate, flow, predators and prey, behaviour...
• Metapopulations of other taxa, but also: community ecology, metacommunities
• Extend to environmental filter concept, possibly in addition to response diversity
• On longer timespan - evolution element
• We sought general rules of thumb; could also be used to optimize decisions as part of decision analysis --- working with simulations tailored to real systems to test competing management hypotheses

Our work suggests three broad research priorities.

1. Identifying what drives productivity in different scenarios --- e.g. density, environmental relationship, regimes? [Bring in @vert-pre2013] and therefore identifying management approaches that are robust to our uncertainty
2. Identifying response diversity and spatial asynchrony [e.g. thermal tolerance curves in @eliason2011]
3. Looking for real-life case studies as in situ experiments testing our conclusions; does managing salmon portfolios result in better outcomes?

Conclusions

• reiterate rules of thumb
• reiterate utility of general portfolio simulation and optimization approach

seananderson/metafolio documentation built on Feb. 13, 2024, 5:47 a.m.