NSFDEB-NERC: Informing population models with evolutionary theory to infer species' conservation status
Lead Research Organisation:
University of Glasgow
Department Name: College of Medical, Veterinary, Life Sci
Abstract
Natural mortality and environmental resources are intimately related to physiology, body size,
fecundity, and lifespan, all of which play an instrumental role in population dynamics. Yet
mortality and resource limitation are notoriously difficult to measure in wild populations,
hindering our ability to prioritize marine species that are at greatest risk of overexploitation.
Crucially, we lack mechanistic theory linking physiology, life histories and population dynamics.
Our central hypothesis is that evolutionary theory can take the place of missing information
on demographic rates or population trends, and can be used to combine data from similar species
to predict population dynamics. We propose to develop a scientific research program to test
this idea and add to our knowledge of the processes regulating the dynamics of marine populations.
We will use a combination of evolutionary theory and hierarchical Bayesian state-space models
of data to infer and predict the life history and population dynamics of three marine fish
clades with diverse life histories: sharks and rays, tunas, and groupers.
Specifically, we will 1) use state-dependent life history theory to develop evolutionary priors
for demographic rates, including mortality and resource limitation and 2) use state-space
models to impute the population trajectories of related species, given our evolutionary priors.
This will 3) generate and refine new theory for the evolution of sharks and rays, groupers,
and tunas that can ultimately be tested comparatively. Finally, we will 4) engage in species'
assessments, training, and outreach to boost the broader impacts of our work. Our research
will produce theory predicting the demographic rates that are correlated with suites of life
history traits, and then generate more precise posterior estimates of these demographic rates
by fitting a structured population model. This integrative approach will allow us to refine
and validate our results with species that have been assessed, and then to assess the vulnerability
of data-limited and potentially endangered species of sharks and rays, groupers, and tunas.
Along the way, our work will generate new insights about the relationship between life-history
traits of marine species, environmental drivers such as resources and mortality, and resilience
to anthropogenic or environmental perturbations.
Intellectual Merit :
We take a new approach to linking evolutionary theory with ecological data. While previous
work has used evolutionarily derived priors in fishery stock assessments (He et al. 2006;
Mangel et al. 2010), this research will provide a mechanistic framework assessing how stage-specific
mortality and resource limitation determine life history evolution and population dynamics.
The novelty of this approach is that we are not hardwiring our assumptions about life history
trait co-variation into the model. We will test our predictions for how resources and natural
mortality select on life histories by confronting our population dynamics model with real-world
data from wild fishes.
fecundity, and lifespan, all of which play an instrumental role in population dynamics. Yet
mortality and resource limitation are notoriously difficult to measure in wild populations,
hindering our ability to prioritize marine species that are at greatest risk of overexploitation.
Crucially, we lack mechanistic theory linking physiology, life histories and population dynamics.
Our central hypothesis is that evolutionary theory can take the place of missing information
on demographic rates or population trends, and can be used to combine data from similar species
to predict population dynamics. We propose to develop a scientific research program to test
this idea and add to our knowledge of the processes regulating the dynamics of marine populations.
We will use a combination of evolutionary theory and hierarchical Bayesian state-space models
of data to infer and predict the life history and population dynamics of three marine fish
clades with diverse life histories: sharks and rays, tunas, and groupers.
Specifically, we will 1) use state-dependent life history theory to develop evolutionary priors
for demographic rates, including mortality and resource limitation and 2) use state-space
models to impute the population trajectories of related species, given our evolutionary priors.
This will 3) generate and refine new theory for the evolution of sharks and rays, groupers,
and tunas that can ultimately be tested comparatively. Finally, we will 4) engage in species'
assessments, training, and outreach to boost the broader impacts of our work. Our research
will produce theory predicting the demographic rates that are correlated with suites of life
history traits, and then generate more precise posterior estimates of these demographic rates
by fitting a structured population model. This integrative approach will allow us to refine
and validate our results with species that have been assessed, and then to assess the vulnerability
of data-limited and potentially endangered species of sharks and rays, groupers, and tunas.
Along the way, our work will generate new insights about the relationship between life-history
traits of marine species, environmental drivers such as resources and mortality, and resilience
to anthropogenic or environmental perturbations.
Intellectual Merit :
We take a new approach to linking evolutionary theory with ecological data. While previous
work has used evolutionarily derived priors in fishery stock assessments (He et al. 2006;
Mangel et al. 2010), this research will provide a mechanistic framework assessing how stage-specific
mortality and resource limitation determine life history evolution and population dynamics.
The novelty of this approach is that we are not hardwiring our assumptions about life history
trait co-variation into the model. We will test our predictions for how resources and natural
mortality select on life histories by confronting our population dynamics model with real-world
data from wild fishes.
Planned Impact
Conservation impact
Biologists and managers around the globe urgently need a method to infer
vulnerability and recovery potential of marine species. Working with the Rutgers Center for
Ocean Education Excellence: Networked Ocean World (COSEE NOW), we will develop an
educational media package introducing our combined theoretical and statistical approach. We
will simultaneously design a web-based tool with a graphical user-interface, using Shiny and R.
Our intention is to design the tool so that anyone can estimate vulnerability for data-limited
species, given life history information, taxonomy, and some representation of data-rich species
in their clade. The staff of COSEE NOW have experience with educational media (i.e. Digital
Storytelling). Therefore we will rely on the COSEE NOW staff's experience in the development
of ocean education aids to ensure these products are effective. We have requested support for
a data technician to ensure the implementation of these plans.
Training and outreach in quantitative marine biology
As this project is quite technical, we are requesting dedicated support for only one postdoc
(from NERC), and expect to do much of the work ourselves. However, NSF support will allow
the senior personnel to maintain our ongoing teaching and outreach activities, and the research
activity is likely to generate spin-off projects for students in each of our research groups.
Quantitative skills are fundamental to our research and to the future of our scientific fields.
We are committed to training the next generation of students. For example, Mangel and
Matthiopoulos have both written textbooks on mathematical biology. Kindsvater and Mangel will
continue teaching informal workshops to undergraduate and graduate students as well as
fisheries scientists. We have previously posted these online (as webinars) and propose to
continue. (Examples can be found at http://www.soe.ucsc.edu/~msmangel/LQFS Links.docx).
Teaching these workshops, as well as formal courses, offers an opportunity to engage and
mentor undergrads and grad students in STEM. In the United States, about 60% of college
students who enroll in a STEM field switch to a non-STEM field or drop out (President's Council
of Advisors on Science and Technology, 2012). That figure is roughly 80% for those from
minority groups and for women. We have a strong record of successfully recruiting and training
undergraduate and graduate students from groups traditionally under-represented in
quantitative disciplines, including many women and Latinos. Many of these students have
continued to pursue higher education and careers in STEM (see Mangel's Biosketch and
Kindsvater's Results of Prior NSF support). We accomplished this by inviting students in our
courses or workshops to work on independent projects. This gives students a chance to be
involved in the primary research, and to interact informally with us. Engaging with students
outside of class can make them more comfortable in class and encourages them to talk with us
candidly about the challenges and rewards of a career in STEM. We have found this type of
mentorship to be highly effective.
We will also continue our involvement in formal mentoring opportunities, as well as in public
outreach in marine biodiversity and evolution. We will continue to participate in these
opportunities locally (e.g. with COSEE NOW) and at meetings (e.g. the SSE Diversity Program).
We also will continue to engage in public outreach opportunities. For example, we are frequently
invited to participate in the development of science media products (documentaries, popular
science news) and to give talks at museums, government agencies, and non-profit
organizations.
Biologists and managers around the globe urgently need a method to infer
vulnerability and recovery potential of marine species. Working with the Rutgers Center for
Ocean Education Excellence: Networked Ocean World (COSEE NOW), we will develop an
educational media package introducing our combined theoretical and statistical approach. We
will simultaneously design a web-based tool with a graphical user-interface, using Shiny and R.
Our intention is to design the tool so that anyone can estimate vulnerability for data-limited
species, given life history information, taxonomy, and some representation of data-rich species
in their clade. The staff of COSEE NOW have experience with educational media (i.e. Digital
Storytelling). Therefore we will rely on the COSEE NOW staff's experience in the development
of ocean education aids to ensure these products are effective. We have requested support for
a data technician to ensure the implementation of these plans.
Training and outreach in quantitative marine biology
As this project is quite technical, we are requesting dedicated support for only one postdoc
(from NERC), and expect to do much of the work ourselves. However, NSF support will allow
the senior personnel to maintain our ongoing teaching and outreach activities, and the research
activity is likely to generate spin-off projects for students in each of our research groups.
Quantitative skills are fundamental to our research and to the future of our scientific fields.
We are committed to training the next generation of students. For example, Mangel and
Matthiopoulos have both written textbooks on mathematical biology. Kindsvater and Mangel will
continue teaching informal workshops to undergraduate and graduate students as well as
fisheries scientists. We have previously posted these online (as webinars) and propose to
continue. (Examples can be found at http://www.soe.ucsc.edu/~msmangel/LQFS Links.docx).
Teaching these workshops, as well as formal courses, offers an opportunity to engage and
mentor undergrads and grad students in STEM. In the United States, about 60% of college
students who enroll in a STEM field switch to a non-STEM field or drop out (President's Council
of Advisors on Science and Technology, 2012). That figure is roughly 80% for those from
minority groups and for women. We have a strong record of successfully recruiting and training
undergraduate and graduate students from groups traditionally under-represented in
quantitative disciplines, including many women and Latinos. Many of these students have
continued to pursue higher education and careers in STEM (see Mangel's Biosketch and
Kindsvater's Results of Prior NSF support). We accomplished this by inviting students in our
courses or workshops to work on independent projects. This gives students a chance to be
involved in the primary research, and to interact informally with us. Engaging with students
outside of class can make them more comfortable in class and encourages them to talk with us
candidly about the challenges and rewards of a career in STEM. We have found this type of
mentorship to be highly effective.
We will also continue our involvement in formal mentoring opportunities, as well as in public
outreach in marine biodiversity and evolution. We will continue to participate in these
opportunities locally (e.g. with COSEE NOW) and at meetings (e.g. the SSE Diversity Program).
We also will continue to engage in public outreach opportunities. For example, we are frequently
invited to participate in the development of science media products (documentaries, popular
science news) and to give talks at museums, government agencies, and non-profit
organizations.
Organisations
People |
ORCID iD |
| Jason Matthiopoulos (Principal Investigator) |
Publications
Horswill C
(2021)
Improving assessments of data-limited populations using life-history theory
in Journal of Applied Ecology
Horswill C
(2019)
Global reconstruction of life-history strategies: A case study using tunas
in Journal of Applied Ecology
Kindsvater HK
(2018)
Overcoming the Data Crisis in Biodiversity Conservation.
in Trends in ecology & evolution
| Description | 1. We have mapped out methods in which classical population and life history theory can be incorporated as sophisticated joint priors into statistical models fitted to sparse data on demographic rates. 2. We have shown for three taxa (scombrids, shark & rays) how multispecies statistical models can be used to impute missing information on life-history traits, hence enhancing models of stock assessment and population viability. 3. We have extended these techniques to other marine taxa (seabirds). 4. We have developed models linking life history evolution with size spectrum theory in marine fish communities. 5. We are developing models linking imperfect fishing data from multiple fisheries in multiple countries to the underlying population dynamics of multiple fish species. |
| Exploitation Route | Incorporation of results into stock assessments and population viability analyses. We will be doing this work as part of the ongoing project. |
| Sectors | Agriculture Food and Drink |
| Title | Improving assessments of data-limited populations using life-history theory |
| Description | 1. Predicting how populations may respond to climate change and anthropogenic pressures requires detailed knowledge of demographic traits, such as survival and reproduction. However, the availability of these data varies greatly across space and taxa. Therefore, it is common practice to conduct population assessments by filling in missing values from surrogate species or other populations of the same species. Using these independent surrogate values concurrently with observed data neglects the life-history trade-offs that connect the different aspects of a population's demography. Consequently, this approach introduces biases that could ultimately lead to erroneous management decisions. 2. We use a Bayesian hierarchical framework to combine fragmented multi-population data with established life-history theory and reconstruct population-specific demographic data across a substantial part of a species breeding range. We apply our analysis to a long-lived colonial species, the black-legged kittiwake Rissa tridactyla, that is classified as globally Vulnerable and is highly threatened by increasing anthropogenic pressures, including offshore renewable energy development. We then use a projection analysis to examine how the reconstructed demographic parameters may improve population assessments, compared to models that combine observed data with independent surrogate values. 3. Demographic parameters reconstructed using a hierarchical framework can be utilised in a range of population modelling approaches. They can also be used as reference estimates to assess whether independent surrogate values are likely to over or underestimate missing demographic parameters. We show that surrogate values from independent sources are often used to fill in missing parameters that have large potential demographic impact, and that resulting biases are driven in unpredictable directions thus precluding assessments from being consistently precautionary. 4. Synthesis and applications. Our study dramatically increases the spatial coverage of population-specific demographic data for black-legged kittiwakes. The reconstructed demographic parameters presented can also be used immediately to reduce uncertainty in the consenting process for offshore wind development in the United Kingdom and Ireland. More broadly, we show that the reconstruction approach used here provides a new avenue for improving evidence-based management and policy action for animal and plant populations with fragmented and error prone demographic data. 22-Mar-2021 |
| Type Of Material | Database/Collection of data |
| Year Produced | 2021 |
| Provided To Others? | Yes |
| URL | https://datadryad.org/stash/dataset/doi:10.5061/dryad.qnk98sfg0 |
| Title | Reconstructing mean rates of adult survival and fecundity from pooled taxonomic data |
| Description | 1. Predicting how populations may respond to climate change and anthropogenic pressures requires detailed knowledge of demographic traits, such as survival and reproduction. However, the availability of these data varies greatly across space and taxa. Therefore, it is common practice to conduct population assessments by filling in missing values from surrogate species or other populations of the same species. Using these independent surrogate values concurrently with observed data neglects the life-history trade-offs that connect the different aspects of a population's demography. Consequently, this approach introduces biases that could ultimately lead to erroneous management decisions. 2. We use a Bayesian hierarchical framework to combine fragmented multi-population data with established life-history theory and reconstruct population-specific demographic data across a substantial part of a species breeding range. We apply our analysis to a long-lived colonial species, the black-legged kittiwake Rissa tridactyla, that is classified as globally Vulnerable and is highly threatened by increasing anthropogenic pressures, including offshore renewable energy development. We then use a projection analysis to examine how the reconstructed demographic parameters may improve population assessments, compared to models that combine observed data with independent surrogate values. 3. Demographic parameters reconstructed using a hierarchical framework can be utilised in a range of population modelling approaches. They can also be used as reference estimates to assess whether independent surrogate values are likely to over or underestimate missing demographic parameters. We show that surrogate values from independent sources are often used to fill in missing parameters that have large potential demographic impact, and that resulting biases are driven in unpredictable directions thus precluding assessments from being consistently precautionary. 4. Synthesis and applications. Our study dramatically increases the spatial coverage of population-specific demographic data for black-legged kittiwakes. The reconstructed demographic parameters presented can also be used immediately to reduce uncertainty in the consenting process for offshore wind development in the United Kingdom and Ireland. More broadly, we show that the reconstruction approach used here provides a new avenue for improving evidence-based management and policy action for animal and plant populations with fragmented and error prone demographic data. 22-Mar-2021 |
| Type Of Technology | Software |
| Year Produced | 2021 |
| Open Source License? | Yes |
| Impact | Contribution to parameterisation of models for Population Viability Analysis |
| URL | https://datadryad.org/stash/dataset/doi:10.5061/dryad.qnk98sfg0 |
| Description | Networked presentation to SNH about the use of state-space models for conservation |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | Regional |
| Primary Audience | Professional Practitioners |
| Results and Impact | Briefing talk and training workshop delivered to members of Scottish Natural Heritage, via video link. |
| Year(s) Of Engagement Activity | 2019 |
| Description | ScotMER Workshop on State-Space modelling for seabird conservation |
| Form Of Engagement Activity | A formal working group, expert panel or dialogue |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Professional Practitioners |
| Results and Impact | Led by the Scottish Government, the ScotMER programme is improving the scientific evidence to understand the environmental and socioeconomic implications of offshore renewable developments. **Access the full programme here** This event will present recent research on seabirds, marine mammals, fish, benthic and socioeconomics , in relation to offshore renewable developments, as well as provide an update on the Scottish Government's commitment to a Blue Economy Action Plan and the Sectoral Marine Plan for Offshore Wind. We will also hold a session to get views on the key evidence gaps for new floating wind technology. |
| Year(s) Of Engagement Activity | 2020 |
| URL | https://hopin.com/events/scotmer |