FW662 Lecture 13 – Management 1Lecture 13. Management of populations. Reading: Hilborn, R. and C. J. Walters. Chapter 3. Behavior of exploited populations. Pages 47-103 in Quantitative fisheries stock assessment. Chapman and Hall, New York,New York, USA. 570 pp.Ludwig, D., R. Hilborn, and C. Walters. 1993. Uncertainty, resource exploitation, andconservation: lessons from history. Science 260:17,36.Stacey, P. B. and M. Taper. 1992. Environmental variation and the persistence of smallpopulations. Ecological Applications 2:18-29.Optional:Anderson, D. R. 1985. Constrained optimal exploitation: a quantitative theory. Pages105-116 in S. L. Beasom and S. F. Roberson, Game Harvest Management. McCullough, D. R. 1984. Lessons from the George Reserve, Michigan. Pages 211-242in White-tailed deer ecology and management. Wildlife Management Institute.Walter, C. 1986. Chapter 4. Models of Renewable Resource Systems. Pages 64-128 inAdaptive Management of Renewable Resources. Macmillian, New York, NewYork, USA. 374 pp.Management of a population requires 4 steps (Hilborn and Walters 1992): definition of a goal,development of a model to evaluate management options to achieve the goal,implementation of the management option selected (along with necessary data collectionschemes), and an evaluation procedure to see that the management strategy is working.1. Goal or objective (some objectives may be hidden). Examples are:Threatened and endangered species -- raise the population level to insurepersistence.Minimum viable populationPersistenceTime to extinctionPest control -- lower the population or more reasonably lower the level of damage(same number of coyotes, only the ones present don't like mutton)Commercially important species -- such as halibut, are managed for maximumsustained yield.Game species -- Maximize production of trophy animalsMaximize hunter recreation (generating maximum income)Maximize quality of recreation (which may increase license cost)Minimize game damage payments and/or rancher complaints2. Need a model to be able to develop your management strategies.Once you have defined a goal, you have to have some ability to test variousmanagement strategies (decisions) to see what level will achieve the goal. Anderson (1985), concerning exploitation, argues that to develop an optimalFW662 Lecture 13 – Management 2R(t) ' Ri1 &N(t)KR(t) ' Ri1 &N(t)K2exploitation system, you need a) birth process as a function of density, and b)death process as a function of density and number harvested (a and b constitutedensity-dependent relationships in the population). He also advocates includingthe environmental stochasticity inherent in the natural system in the model. Heargues that deterministic models, or simple models such as the logistic where r =b - d, are totally inadequate. He provides the understatement of the century inpopulation biology: "The availability of adequate data will continue to be aserious limitation." In reality, lack of information is deeper than just some poorlyestimated parameters. Often basic functional relationships are poorly understood,and are not modeled adequately to provide correct system behavior, regardless ofthe parameter values used.Stacey and Taper (1992), concerning population persistence, also argue formodels with a) environmental stochasticity, and b) density dependence, includingthe proper form of the density dependence. In persistence models, as populationdeclines, the compensation for small population size takes the form of increasedbirth rates and decreased death rates (density dependence), and so is a significantfactor in increasing population persistence. Stacey and Taper (1992) tested 2forms of density dependence with their data: logisticand 2-logisticHowever, their data precluded a significant test between these models. Their datadid show significant correlations between adult survival and population size,although the lack of correlation is likely a Type II error, and additional analysesare provided by Middleton and Nisbet (1997). In the following table, 4 variablesare correlated against population size:Variable SampleSizeCorrelation ProbabilityAdult Survival 9 -0.65 0.058Juvenile Surv. 9 -0.30 0.434Reproduction 10 -0.56 0.094FW662 Lecture 13 – Management 3Emigration 9 0.28 0.473Stacey and Taper (1992) also make some grand-stand statements about datarequired to build a model: "it can be exceedingly difficult to provide meaningfulestimates of persistence time because the results depend so largely on theassumptions that are used to create the model." What a revelation! However,they continue: "Furthermore, even if the model is specified correctly, relativelysmall errors in the estimation of the parameters can lead to large errors in thepredictions (see also Goodman 1984)." They explored 2 functions forincorporating density dependence, illustrating a lack of knowledge on functionalforms of a very basic relationship.To summarize, the need for a model is to guide management: explore options,evaluate options quantitatively, and to formulate optimal decision criterion. Evenif the model is built from very poor data, it may be useful in guiding collection ofthe needed data.3. Ready to implement our management scheme. a. Need annual data to maintain an optimal decision (Anderson 1985). Examplesare to estimate population size with rigorous estimation schemes, or maybeestimate reproductive rates via helicopter surveys such as CDOW does with deerand elk, or the USFWS and the Canadian Wildlife Service does with air surveysof pot-hole density and brood density, or maybe estimate annual survival rates viaradio-collared animals or bands, or estimate harvest taken, or maybeimmigration/emigration rates to maintain a metapopulation.An example of a very simple data collection scheme is McCullough et al.'s (1990)linked sex harvest strategy. Hunters are able to shoot animals of either sex, butantlered males are preferred. For a constant population size, the proportion ofmales in the harvest is a function of harvest size:FW662 Lecture 13 – Management 4Incremental increases in harvest are used to seek out a pre-determined percentageof males in the harvest. As the harvest size increases, the percent of females inthe harvest increases. When too many females are taken, the population eventuallydeclines below MSY, and the proportion of females increases even more in theharvest. Typically, try to adjust the harvest so that 40% of the