Avoiding Mistakes in Population
Modeling
The following list includes some of the mistakes modelers make in
developing ecological models at the population or metapopulation levels, for
example for population viability
analysis. It grew out of my experiences in reviewing and editing papers on
modeling, answering questions from RAMAS users, as well as my own mistakes.
My goal is mostly to help users of
RAMAS Metapop and
RAMAS GIS avoid these common
mistakes. Although the list focuses on models developed using these programs,
most (if not all) of these issues are relevant for models developed in any
program or using any modeling platform.
Note that whether a model feature is a mistake or not depends on the
context (the species, other model components and features, etc.). Thus, for
example, modeling only females is a mistake
only under some circumstances. Such circumstances are discussed in the links
below.
I welcome any feedback, comments, and suggestions, including mundane
things like misspellings and broken links, and (more importantly) suggestions
about additional types of mistakes, references, approaches, etc.
General mistakes
- Invalid model assumptions
- Model too complex
- Model too simple
- Internal inconsistency (Not paying
attention to error messages)
Modeling Demographic
Structure
- No demographic structure (scalar
models for age-structured populations)
- Bias in fecundity estimation
- Fecundity vs. maternity
- Not incorporating proportion breeding
- Sex ratio
- Fecundity of first age class or stage
- Bias in survival rate estimation
- Uncertainty in survival rate
estimation
- Using survivorship instead of survival
rate
- Survival rates in the diagonal
- Ignoring constraints
- Too many (or too few) age classes or
stages
- Modeling only females
- Ignoring genetics
Modeling Density
Dependence
- Using (the wrong type of) density
dependence
- Not using density dependence
- Bias in Rmax estimation
- Underestimating maximum growth rate
(Rmax)
- Overestimating maximum growth rate (Rmax)
- Incorrectly modeling impact under density
dependence
- Toxicity vs. harvest under density
dependence
Adding Variability
(Stochasticity)
- Ignoring variability
- Not using demographic
stochasticity
- Environmental stochasticity
distribution
- Ignoring correlations
- Too few replications
- Random fluctuations vs. regular
oscillations
- Using catastrophes instead of environmental
stochasticity
- Overestimating catastrophe impact
- Not incorporating delayed effects of
catastrophes
- Overestimating variation
- Using the wrong standard deviation
- Not correcting truncations
- Phantom individuals
- Duration (simulation time horizon) too
long or too short
- Confusing uncertainty and variability
Modeling Spatial
Structure
- Ignoring spatial structure
- Too many (or too few) populations
- Ignoring spatial correlation
- Wrong dispersal rates
- Symmetric dispersal rates
- Too high (or too low) map resolution
Presenting Results
- Implicit assumptions
- Ignoring uncertainty
- Emphasizing deterministic (not probabilistic)
results
- Using absolute (not relative) predictions
- Estimating risk of extinction rather than
decline
Impact Assessment
- Failing to incorporate cumulative
impacts
- Selecting the wrong spatial scale (or a
single scale)
- Reference population includes impact
effects
- Uncertainty masking impact
- Underestimating impacts of toxicity,
habitat loss, habitat fragmentation, and harvest
- Difference between experimental and
model time steps
- Wrong time horizon or threshold
References
Last modified: 7 December 2009
