Population Sampling

1
Population Sampling

• ESM 211

2
Common failures in monitoring programs (Elzinga
et al. 2002)

• Technical problems
• Poor design leads to inconclusive results
• Use of multiple or unreliable observers
  complicates interpretation of results
• Data are lost (including inability to interpret
  datasheets)
• Data are not analyzed
• Natural system fluctuation obscures change caused
  by management
• Institutional problems
• Premature termination of monitoring
• Inadequate resources to collect or analyze data
• Managers refuse to use monitoring data to make
  decisions
• Failure to place monitoring within a management
  framework leads to perception (or reality) that
  the data are irrelevant

3
Population vs. Sample
True Population
Sample
4
Confidence intervals

• Measure of precision
• 95 CI commonly used
• 95 chance that the true mean is in the CI
  (Bayesian)
• If the sampling process is repeated many times,
  the the CI will cover the true mean 95 of the
  time (Frequentist)

5
Accuracy Precision Bias
Imprecise without bias
Precise with bias
Bias Does the mean of the estimates converge
to the true value? Precision How variable are
the estimates?
6
Pseudo-Replication

• A term often used in experimental studies, but
  also occurs when sampling natural populations,
• Replicate samples are the smallest units to which
  treatments are independently applied,
• When there is pseudo-replication, there is
  dependency among the replicated samples (beyond
  the treatments!),

7
Hypothetical Population


































































































































8
Clearly defined target population

• Biological population of interest
• Must be well defined in
• Spatial extent
• County, State, Region
• Habitat extent Streams, rivers, ponds...
• Temporal extent Season, month,

• Sampled population should be representative of
  target population
• Encompass spatial range
• Encompass habitat range
• Encompass temporal scale of population processes
• Sampled population should account for spatial and
  temporal aggregation of target population.

9
Random Sampling

• Good statistical properties
• Can be time-consuming to locate sites
• May miss some habitats
• Many sites may be empty

10

• Select qualitatively different habitats of
  interest strata
• Sample randomly within each stratum
• Effort may vary among strata
• Mean is weighted mean of the individual strata
  (weighted by stratum area)
• CI calculation is moderately complex see
  Greenwood pp. 104-105.

Stratified Random Sampling

• Guarantees coverage of all habitat types
• Allows focused effort where it is most valuable
  high density, variability, area

11

• Biased if sampling grid matches environmental
  periodicity
• Urban environments
• Estimated confidence intervals may be too small

Systematic sampling


















































































































12
Sampling schemes to avoid

• Haphazard neither random nor systematic e.g.
  the first bush seen in a field
• Accessibility sites that are easy to get to
• Judgment sites that are deemed typicalOnly
  advantage of these is lowered cost of locating
  sites

13
Sample Methods

• Direct Counts
• Trap, Quadrat, transect
• Assesses density, CPUE
• Mark-Recapture
• Plotless

14
DIRECT COUNTS
TRAP

• Station that captures or records passage of
  mobile animal
• Live traps, sticky traps, nets, automatic camera
• For baited trap, need to know area of attraction
  of bait

• Count all individuals in sample unit
• Difficult when organisms abundant (aerial surveys
  of wildebeest herds)
• Can leads to undercounts if some individuals are
  cryptic
• Can leads to over-counts if the same individuals
  are counted twice or more

15
Quadrat

• Useful for small sessile organisms

http//simp.ucsc.edu/Sites/Images/quadrats/pp1mqua
d.jpg
16
Transect

• Long line sampled continuously or at regular
  intervals
• Belt transect long narrow quadrat
• Line intercept count all individuals touched by
  line
• Line transect count everything seen, correcting
  for distance

17
Line transect sampling

• For each individual sighted, record
• Distance from observer
• Angle from transect line
• Calculate distance from transect line

di zi sin ?i
transect
?i
zi
18
Sighting probability declines with distance

• Fit a function (g(x) probability of
  sighting at distance x) to data on number seen at
  various distances

Wallaroo in Queensland
McCallum (2000)
19
Estimating density with line transect

• Integral of g(x) gives the average probability of
  sighting over the entire width of the transect
• It is also the effective strip width (ESW) of the
  transect
• Density is D n / (2LESW)
• n number sighted
• L transect length

20
POPULATION INDEX

• Number observed or captured for a known effort
• Catch per unit effort (CPUE)
• captured per 100 trap-nights
• seen by observer in 30 seconds
• Often assumed to be proportional to abundance,
  but usually has power relation
• Recalibration needed when technology or observer
  changes

21
MARK-RECAPTURE

• Capture and mark known of individuals
• 2nd round of captures soon after
• Time for mixing, but not mortality
• Fraction of marked individuals in recapture
  sample is estimate of the proportion of
  population marked in first capture

22
Lincoln-Peterson index
Mark
Recapture
23
Marking methods

• Paint or dye
• Color band
• birds
• Unique markings
• Large mammals keep photo record
• Toe clipping
• Reptiles, amphibians, rodents

(NPS 2000)
(Sutherland 1996)
24
Plotless

• Select random individual or location and measure
  distance to nearest neighbor
• Good for trees, shrubs
• Sutherland pp. 60-62

(Greenwood 1996)
25
Resources

• Buckland, S.T., Anderson, D.R., Burnham, K.P. and
  Laake, J.L. 1993. Distance Sampling Estimating
  Abundance of Biological Populations. Chapman and
  Hall, London, reprinted 1999 by RUWPA, University
  of St. Andrews, Scotland. 446pp. Available online
  at http//www.colostate.edu/depts/coopunit/distanc
  ebook/download.html.
• Elzinga, C.L., D.W. Salzer, J.W. Willoughby, and
  J.P. Gibbs. 2002. Monitoring plant and animal
  populations. Blackwell Science, Malden, NY.
• Hayek, L.-A. C., and M. A. Buzas. 1997. Surveying
  Natural Populations. Columbia University Press,
  New York.
• Henderson,P.A. 2003. Practical Methods in Ecology
  Blackwell, Oxford.
• Hilborn, R., and M. Mangel. 1997. The Ecological
  Detective Confronting Models with Data.
  Princeton University Press, Princeton, NJ.
• McCallum, H. 2000. Population Parameters
  Estimation for Ecological Models. Blackwell,
  Oxford.
• New, T. R. 1998. Invertebrate Surveys for
  Conservation. Oxford University Press, Oxford.

26
Resources
National Park Service (NPS). 2000. Glacier Bay
National Park and Preserve Humpback Whales.
Online document at http//www.nps.gov/glba/learn/
preserve/projects/whale/index.htm Schmitt, R. J.,
and C. W. Osenberg, eds. 1996. Detecting
Ecological Impacts Concepts and Applications in
Coastal Habitats. Academic Press, San
Diego. Sutherland, W. J., ed. 1996. Ecological
Census Techniques A Handbook. Cambridge
University Press, Cambridge, UK. Thompson, W. L.,
G. C. White, and C. Gowan. 1998. Monitoring
Vertebrate Populations. Academic Press, San
Diego. Young, L. J., and J. H. Young. 1998.
Statistical Ecology A Population Perspective.
Kluwer Academic Publishers, Boston.