Docstoc

Population Viability Analysis _PVA_

Document Sample
Population Viability Analysis _PVA_ Powered By Docstoc
					Population Viability
Analysis (PVA)
PVA
   A systematic examination of interacting factors
    that place a population or species at risk of
    extinction
     How  do we characterize risk?
     How do we examine interacting factors
      (stochasticity,anthroprogenic, genetic, habitat)?
     What are the positives and negatives of this
      approach?
Characterization of Risk
   Probabilities of Extinction/Persistence
     Extinction, management threshold, or quasi-extinction
     Time frame (ex. Probability that spotted owls will
      persist for next 100 years is 0.30)
          Mean time to extinction
   Ex. If probability of persistence for 10 years (P10)
    = 0.8, what is mean probability of persistence?
     Probability   of persistence for 1 year (P1) = 0.81/10 =
      0.9779
     Mean probability of persistence = 1/-ln(0.9779) = 45
      years
How?
 Rules of Thumb
 Count-Based
     Often   assuming a census
   Demographic Models
     Generally   matrix models
Count Based
   Census, Estimates of Abundance/Density
     Determinisitic
          E.g., Exponential/Logistic Models
     Stochastic
          Incorporating process variance
Deterministic vs Stochastic
   Geometric vs arithmetic
    mean
   Stochastic change
       Prediction diverge over
        time
       Ending distribution is
        skewed
       Diverge from arithmetic
        mean
            Center on geometric mean
             for large n
            Still some can go extinct
Probability of Extinction and
Stochasticity
   Define
    μ  = lnλG = mean
      (arithmetic) lnλt-x + …
      lnλ0 / t
          u > 0, λG >1, u < 0, λG
           <1
                < u, higher probability
     σ2= variance of the
      mean lnλG
          > σ2, more peaked
Example: Yellowstone Grizzlies

   Census - Assumed
   Density Independent
   Estimate μ and σ2
    from count data:
Grizzly Extinction Probability
 Calculate visually from stochastic
  projections:
 Or from μ and σ2
Assumptions
   Mean and variance of λ constant
     No density dependence
     No demographic stochasticity
     No environmental trends

 Uncorrelated environmental conditions
 Environmental variation is relative small
 Census is a census
Count Based Generalized
 Density Dependence
 Demographic Stochasticity
 Correlated Environments
 Catastrophes
 Bonanzas
Density Dependence in Small
Population
   Place cap on maximum size
     Population   may be small enough to remain at
      risk of extinction even at maximum size
   Allow higher growth rates at smaller sizes
     AlleleEffects – positive density dependence
     Inbreeding depression
Demographic Stochasticity
 Need variances in rates to incorporate
 Raise quasi-extinction threshold to
  minimize effects
 Magnified effects of environmental
  stochastcity with demographic
  stochasticity
Environmental Autocorrelation
   Positive autocorrelation
     Increasesextinction risk in density-
      independent models
     Complicated for Density-dependent models

   Negative autocorrelation rare or non-
    existent
Catastrophes and Bonanzas
 Rare events that are difficult to incorporate
 But, … can have the greatest effect
Components of Variance
Analysis
   See notes and
   Gould, W. R. & Nichols,
    J. D. (1998). Estimation
    of temporal variability of
    survival in animal
    populations. Ecology
    79:2531–2538

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:3
posted:2/1/2012
language:English
pages:21