Prepare for CWB-style scenarios by drilling the gap between definitions and decisions. For each concept — habitat use versus selection, nondetection versus absence, source versus sink, carrying capacity — practice identifying what evidence a scenario actually provides, naming the concept that evidence supports, and stating the management or documentation decision that follows. Worked examples with a plausible mistake, a better decision, and a written rationale build this skill faster than rereading notes. Finish with a self-check rubric: if you can justify each term choice and flag the limits of each inference, you are reasoning the way scenario questions reward.
Habitat Use, Selection, and Preference: Which Term Does the Evidence Support?
Use, selection, and preference describe different levels of evidence. Use is where an animal occurs; selection is disproportionate use relative to availability; preference requires ranked choice across varying availability. Match the term to the data described, not to intuition.
These terms are commonly collapsed in everyday speech, which is exactly why they are worth separating during review. Use simply documents where an animal was detected. Selection requires comparing use against availability: if forest is 20% of the landscape but 60% of radiolocations, forest is selected. Preference is a stronger claim still — it implies a ranked choice, which needs comparisons across areas or periods where availability differs. When a scenario describes only locations of detections, the strongest defensible term is use; when it gives both use and landscape proportions, selection becomes defensible.
Worked example: a telemetry study reports 60% of radiolocations in forest that covers 20% of the study area, and the proposed answer calls forest the species' preferred habitat. That overreaches — the data support selection, not preference, because no comparison across varying availability was made. The better decision is to state 'forest is selected at the landscape scale described' and note that preference cannot be inferred from one availability condition. This matters because a selection finding supports protecting existing forest cover, while a preference claim implies choices the study never tested — and a scenario answer that overclaims loses the inference points.
| Term | Evidence required | Scenario cue | Common overreach |
|---|---|---|---|
| Use | Locations or detections of animals | Where individuals were found | Calling used habitat 'selected' |
| Selection | Use compared with availability | Detections plus landscape proportions | Calling selection 'preference' |
| Preference | Ranked choice across varying availability | Multiple sites or periods with different conditions | Inferring preference from a single condition |
Nondetection Is Not Absence: Reasoning With Detection Probability
Occupancy analysis treats nondetection as ambiguous: a species may be present but undetected. Repeated visits, detection probability, and occupancy estimates let you state conclusions with uncertainty instead of declaring absence prematurely.
Occupancy (psi) is the probability a site is occupied; detection probability (p) is the probability of detecting the species given it is present. A naive estimate — the fraction of surveyed sites with at least one detection — is biased low whenever p is below one, which it nearly always is. Repeated visits to the same sites within a period when occupancy does not change let you separate the two: a site with no detections might be unoccupied, or occupied with missed detections. The more visits and the higher p, the more confidently a nondetection can be read as true absence.
Worked example: two visits to ten wetlands produce zero detections of a secretive marsh bird, and the draft report concludes the species is absent from the basin. The plausible mistake is treating nondetection as absence without estimating p — secretive species often have low detection even when present. The better decision is to report an occupancy estimate with its uncertainty, and if p is likely low, increase visit numbers before drawing management conclusions. This matters because an 'absent' finding can justify habitat changes or delisting actions that an honest 'possibly present, detection uncertain' finding would not, and scenario answers that ignore detection uncertainty reach the wrong decision.
Source, Sink, and Metapopulation Thinking in Case Scenarios
A source population produces more recruits than deaths; a sink does not and persists through immigration. Metapopulation thinking asks whether connected patches exchange individuals. Abundance or adult presence alone cannot distinguish these states.
These terms differ in what evidence establishes them. A source is demonstrated by local recruitment exceeding local mortality — typically requiring data on reproduction, survival, and dispersal, not just counts. A sink looks healthy if you only count animals, because immigrants keep arriving into habitat where births do not replace deaths. Metapopulation structure adds a patch-level view: local populations may blink out and be recolonized, so the persistence of the network, not any single patch, is the unit of concern. Scenario evidence like 'high counts in a meadow' or 'juveniles seen occasionally' must be mapped against this hierarchy before choosing a label.
Worked example: nest boxes added to a forest patch attract many adult birds each spring, and the recommendation is to expand the program because the patch appears productive. The plausible mistake is reading adult abundance as evidence of a source. Follow-up might show low fledging success due to nest predators, with adults recruited from surrounding habitat — a sink. The better decision is to measure nesting success and, where possible, movement before scaling up. This matters because investing in sink habitat can divert resources from genuinely productive patches and even attract individuals into conditions where their fitness is lower, a decision a scenario answer should catch.
Carrying Capacity and Density Dependence: Applying the Model to the Data
Carrying capacity (K) is the population size an environment can sustain under current conditions; density dependence means vital rates shift as density changes. Both are context-dependent, and responses to management can lag behind habitat change.
In the logistic model, growth slows as a population approaches K because resources or space become limiting; in real populations, density dependence can appear in fecundity, survival, dispersal, or body condition, and it can lag. K is not a fixed species trait — it moves with forage quality, winter severity, water availability, and disturbance. A scenario that gives you a density figure and habitat trend is asking you to reason about which vital rate is likely affected and over what timescale, not to recite the equation. Watch for whether the scenario supports an equilibrium claim at all.
Worked example: shrub restoration on a deer winter range is completed, and a plan projects the herd will reach its new higher K within one season. The plausible mistake is assuming carrying capacity responds instantly to habitat improvement — shrubs need years to reach browse value, and density-dependent feedback in recruitment appears over multiple cohorts. The better decision is to set interim expectations, monitor body condition and fawn-to-doe ratios across several years, and treat K as an estimate to be revised. This matters because overestimating near-term capacity can drive harvest or stocking decisions that outpace what the restored habitat actually supports.
Applied Safety and Ethics in Field Scenarios: Choosing the Conservative Option
Applied scenario items ask you to match a described field situation to standard precautions and legal constraints. The reliable pattern is to identify the hazard or protected status, choose the option that avoids unnecessary handling and preserves evidence, and document.
Field scenarios reward a decision sequence rather than improvisation: first name the hazard or the legal status of what you found; second, choose the action that avoids unlicensed handling, unnecessary proximity, or disturbance; third, record what you observed and report it through the appropriate channel. Wildlife fieldwork commonly involves disease precautions around carcasses, hazard awareness around large or defensive animals, and permit conditions governing handling of protected species. In a paper scenario you cannot inspect or intervene — you reason from the facts given, and the conservative, documented option is generally the defensible one.
Worked example: during an avian survey you find a dead raptor with no visible cause of death, and you consider collecting it to give a rehabilitation or research contact. The plausible mistake is assuming good intentions permit handling — many protected species are covered by permit regimes that restrict who may possess carcasses or parts. The better decision is to document location, time, condition, and photographs without moving the bird, then report per your project's permit conditions or to the relevant wildlife agency. This matters because the ethical and legal weight falls on proper reporting and chain of custody, and the scenario is testing whether you know that possession itself can be the violation.
Methods and Documentation: What Makes a Survey Write-Up Defensible
A defensible write-up states the objective, study design, sampling frame, effort, detection limitations, and the boundary between observation and inference. Scenario questions reward recognizing which of these elements a described study is missing.
Learn to audit a described study against a fixed checklist: What was the objective? What population or area does the sampling frame cover? How were sites or transects chosen — randomly, systematically, opportunistically? What effort was applied, and over what period? What detection limitations are acknowledged? Finally, does each conclusion trace to a specific observation? Opportunistic sampling, for example, can support a statement that a species occurs in an area but not an estimate of how commonly, because the sampling was not representative by design.
Worked example: a report states 'the survey confirms the species is widespread across the county' based on sightings collected along public roads during one month. The plausible mistake is accepting the confirmation claim at face value. The better decision is to identify two limits — roadside sampling skews toward certain habitat edges, and a single month cannot support 'widespread' across seasons — and restate what the data support: documented occurrence at roadside points during that period. This matters because interpretation questions hinge on matching claim strength to design, and being able to name the missing design element is the skill being tested.
A Preparation Sequence and Self-Check Rubric
Sequence: build the term distinctions, drill scenario translation with written rationales, run a small structured observation exercise, then mixed practice sets. Check readiness against a rubric of justified inferences, not a single score.
A workable sequence: first, spend early sessions writing one-sentence definitions of each key pair — use versus selection, occupancy versus abundance, source versus sink, K versus realized density — plus the evidence each requires. Second, drill translation: for each practice scenario, write the concept the evidence supports, the decision that follows, and one limitation, before checking the answer. Third, run the observation exercise below. Fourth, mix topics in later practice so you must first diagnose which concept a scenario is about — the step raw definition review never trains.
Exercise: choose a visible local species (birds work well) and one site type. Visit the site three times over a week; each visit, record detections, time, weather, and one habitat covariate. Expected observations: detection varies between visits even if the species stays present, and covariates explain part of that variation — a direct felt demonstration of detection probability. Self-check rubric: (1) Can you state what each visit's data do and do not support? (2) Can you name a conclusion that three visits cannot support? (3) Would you report a nondetection as absence — if so, revisit the occupancy material. A score of 3 of 3, with written rationales, is a learning milestone indicating the translation skill is forming, not a prediction of any exam outcome.
- Readiness check: given any scenario, you can name the concept, the required evidence, and one inference limit — in writing, within a few minutes.
- Readiness check: you can explain why two figures differ, such as a naive occupancy rate versus a modeled estimate, without looking at notes.
- Readiness check: you can take a weak management recommendation and identify which design element (frame, effort, detection, or inference) it overreached on.
References and further reading
Use these references to explore the concepts and check the latest information from the relevant organizations.