The
demand for petroleum exploration along the North Slope of Alaska has recently
risen. As a result, petroleum developments
have been
proposed for the National Petroleum Reserve, the Arctic National Wildlife
Refuge, and in offshore waters along the Beaufort Sea coast. Accordingly,
petroleum developments are an increasingly important issue for federal and state
resource managers.
Approximately 25 bird species use the waters along the Beaufort Sea coast (including larids, loons, phalaropes, and waterfowl), making these species especially vulnerable to offshore oil and gas activities. The potential effects of petroleum development on these vulnerable bird populations must be objectively determined (National Environmental Policy Act of 1969: 42 U.S.C.).
The goals of this project are to determine the best methods available for modeling the propensity for avian populations to recover from catastrophic events, such as mass mortality resulting from large-scale oil spills, and to provide natural resource professionals with an easy to use computer program for examining the recovery potential of avian populations in the Beaufort Sea region.
Avian
species may be affected by petroleum-related catastrophes in different ways
depending on their life history, and seasonal occurrence of oil spills. Thus,
some populations may experience severe effects, while others may experience
subtle effects if affected at all. Furthermore, life history characteristics
(e.g., clutch size, nesting propensity, age of 1st reproduction,
longevity) of certain species may allow them to recover much quicker from a
mass-mortality event than other species. In addition to these factors,
environmental variability in demographic parameters, population density, and a
population’s age structure can affect population growth and introduce
uncertainty about the likelihood of population recovery regardless of
pre-perturbation population trends.
Matrix population models provide a flexible framework with a sound theoretical basis for modeling the recovery potential of avian populations. Life history characteristics, density-dependence, age structure, and environmental (co)variation of demographic parameters can easily be incorporated into model structure (Caswell 2001), which make matrix models useful for estimating population viability, as well as recovery times and probabilities that require measures of uncertainty (i.e., prediction interval estimation). Additionally, sensitivities of matrix model dynamics can be readily determined, thus simplifying the identification of critical life stages for management or conservation.
While examples of matrix population models abound in the current literature, most require the use of mathematical and computer programming skills that are not readily available to resource managers. Accordingly, we developed the following objectives.
Matlab code
Koons, D.N., R.F. Rockwell, and J.B. Grand. 2005. Population momentum: implications for wildlife management. Journal of Wildlife Management In Press.
Koons, D.N., J.B. Grand, and J.M. Arnold. In Review. Population momentum across vertebrate life history strategies.
Koons, D.N., R.R. Holmes, and J.B. Grand. In Review. General formulas for the sensitivity of population momentum to changes in population vital rates and initial conditions.
1. Beaufort Sea Bird Population and Recovery Modeling Workshop. (link to pdf of workshop agenda)
In conjunction with the USGS Alaska Science Center (ASC), the Alabama Cooperative Fish and Wildlife Research Unit (ALCFWRU) developed a list of target bird species and populations inhabiting the Beaufort Sea region for the modeling exercises. ASC and ALCFWRU hosted an initial workshop that brought together recognized experts in the field of population modeling, life histories of birds inhabiting the Beaufort Sea region, and the impact of oil spills or other catastrophic events on bird populations. The goal of the initial workshop was to identify the best approach(es) (i.e., frameworks) for modeling populations and estimating recovery potential for Beaufort Sea Bird populations. The workshop consisted of a series of invited presentations by a panel of experts followed by facilitated discussions to identify the methods for modeling population recovery following catastrophic perturbations.
Following the initial workshop, researchers from ALCFWRU began an extensive review of the published literature for demographic parameters that could be incorporated into population models for the target species. The annotated literature review was compiled in a searchable database incorporated with a modeling tool: Aves Modeler.
Meanwhile scientists from Auburn University, ALCFWRU, and the American Museum of Natural History began the task of developing methods to simulate and estimate the short and long-term effects of perturbations to population size and structure. This research has resulted in a series of publications that focus on methods for examining short and long-term dynamics of populations using matrix models. The stated objectives of the workshop were to:
Identify the capability of various approaches to population modeling for estimating recovery potential (e.g., matrix vs. individual-based models, Bayesian approaches, recovery times and viability analyses).
Identify additional considerations for the modeling exercise (e.g., density-dependence, suppression of demographic rates).
Contrast various approaches to model interpretation.
Provide a list of recommendations to guide the development of models and modeling tools for Beaufort Sea bird populations.
2. Understanding the Impacts of Catastrophes on Marine Bird Populations: An Introduction to Aves Modeler. (link to pdf of workshop agenda)
This workshop included a half-day introduction to population modeling, a hands-on opportunity for biologists from universities, state and federal agencies, and private industry to use the developing software tool and provide direction on the interface and capabilities of the database and modeling tool. The stated objectives of the workshop were:
To introduce, or refresh, individual’s knowledge of matrix population models and their use in management and conservation.
To present and demonstrate Aves Modeler using a series of case studies and hands-on examples.
To provide a forum for feedback and questions regarding Aves Modeler.
To present case studies which introduce alternative methods of modeling and parameterization, and to receive feedback on the methodology and type of output that would be appropriate to achieve the goals of users.
3. Understanding the Impacts of Catastrophes on Marine Bird Populations: Using Aves Modeler for Population Level Analyses. (link to pdf of workshop agenda)
This information-transfer workshop introduced the near-final version of Aves Modeler, and included:
A mini-course in matrix population modeling, deterministic and stochastic models, perturbation analysis, and model selection.
An introduction and hands on demonstration of the improved Aves Modeler database.
An introduction and hands on demonstration of the improved Aves Modeler population modeling software.
An opportunity to develop a population model with your own data using Aves Modeler and receive assistance from our staff.
Modeling the Effect of Perturbations to Population Structure
In
management and conservation of wildlife, agency and stakeholder goals are often
centered on the population, its size, and changes in size over time. Population
structure (i.e., the distribution of population size across age, stage, or size
classes) can have strong effects on both population growth rate and future
size. If population structure is perturbed, the proportion of fertile
breeding adults in the population is changed and it takes some time for it to
get back to the pre-perturbation level. As population structure changes, it
causes short-term fluctuations in population size and growth rate, collectively
known as “transient dynamics”. Yet, most population models assume a stable
population structure, and suggested management prescriptions resulting from
these models are dependent on this assumption. The consequences of assuming a
stable population structure are not well known.
Recent empirical evidence has shown that population structure is rarely stable in nature (Bierzychudek 1999, Clutton-Brock and Coulson 2002). Furthermore, environmental catastrophes like large-scale oil spills have great potential to perturb population structure. Thus, transient dynamics and their effect on long-term population size may be very important to consider in population recovery and viability modeling.
We
have developed new tools to study transient dynamics in greater detail to
illuminate their effect on populations of a variety of species. In a series of
theoretical studies, we have found transient dynamics to be both highly erratic
and much different than the commonly modeled stable (i.e., asymptotic)
dynamics. Notably, we found that short-term transient dynamics also have the
potential to set the population on a completely different long-term trajectory
known as population momentum. Across species, long lifespan of slow reproducing
species increases the chances for variability in somatic and reproductive
investment across age classes (Charlesworth 1994). For this reason, transient
dynamics of slow reproducing species were very responsive to our theoretical
changes in population structure when compared to short-lived, fast reproducing
species. Related to this, long-lived species also experienced greater
magnitudes of population momentum in our experiments than short-lived species.
In effect, population momentum could push populations far past an environmental carrying capacity or even to extinction, depending on the direction of momentum. Relative to predictions from asymptotic population modeling, population momentum could shorten or lengthen the time it takes for a population to go extinct, recover from a perturbation, or explode to levels that become a nuisance. We suggest that resource managers place a strong emphasis on estimation of population structure and determine how it might be affected by anthropogenic activities. Such studies will help reduce uncertainty in decision-making and the likelihood of deleterious actions in the future.
James B. Grand, Unit Leader
Alabama Cooperative Fish and Wildlife Research Unit
108 M. White Smith Hall
Auburn University
Auburn, AL 36849
Phone: 334-844-9237
Fax: 334-844-1084
Email: grandjb@auburn.edu
Principal Investigators and Software Engineers
James B. Grand
Jennifer M. Arnold
David N. Koons
Nitin Yogi
Dirk V. Derksen
