Kestrel Ridge: deer, wolves and a missing trophic level
Kestrel Ridge Reserve covers 1,200 ha of temperate oak forest. The last wolf was shot in 1962. Deer density is now estimated at 38 adults per km²; rangers record that 71% of oak seedlings are browsed, and only 40 saplings per hectare survive in the shrub layer, against 900 per hectare in a 1968 survey. Bramble and tall ground vegetation have thinned sharply. The reserve's black grouse, which nest in tall ground cover, have fallen from 55 displaying males in 2000 to 12 in 2024. The warden proposes a 35% annual cull, arguing it will push deer below the level at which oak regenerates. A local hunting group opposes any cull and wants a predator-proof fence around 200 ha of core forest instead. Fencing costs €340,000 to install and €28,000 a year to maintain. The reserve's total annual budget is €210,000, of which €60,000 is already committed to fixed costs.
1. Explain how the removal of wolves could reduce black grouse numbers, even though wolves never eat grouse. Identify the direction of each link and state what kind of feedback loop has been lost.
Read the worked answer
The chain is a trophic cascade, an indirect density-mediated effect:
- Wolves extirpated (1962) → predation pressure on deer removed (negative link deleted).
- Deer density rises to 38/km² (negative link on deer removed, so deer increase).
- Browsing rate reaches 71% of oak seedlings → sapling density collapses from 900/ha to 40/ha (deer exert a strong negative link on vegetation).
- Shrub layer and bramble simplify → tall nesting cover disappears.
- Black grouse lose nesting habitat → displaying males fall 55 → 12 (negative link on grouse).
Wolves affect grouse only through the deer–vegetation pathway, so the effect is indirect and depends on density, not on direct predation.
Feedback: the wolf–deer link was a negative feedback that held deer near a lower equilibrium. Without it, deer are now limited mainly by food and disease (bottom-up), so they can overshoot, degrade the vegetation, then crash. The system has shifted to a new state with simplified understorey.
Caution: these are correlated measurements, not an experiment. Before accepting the cascade, rule out conifer encroachment, climate, and disturbance as alternative drivers of the grouse decline.
2. Evaluate the warden's 35% cull against the fenced enclosure. Use the figures given and at least two ecological concepts in your judgement.
Read the worked answer
The cull as stated is too small. 38 × 0.65 = 24.7 deer/km² remaining. Browse is already 71% at 38/km² and regeneration fails at 40 saplings/ha; a target of roughly 8–10 deer/km² for oak regeneration would require a 74–79% reduction, not 35%. A 35% cull is cheap to repeat but leaves browsing above the regeneration threshold, so the indicator will not move.
The fence is partly unaffordable. 200 ha is 17% of the reserve. Discretionary budget is €210,000 − €60,000 = €150,000 a year. Capital cost (€340,000) is 2.3 times that; maintenance (€28,000) is 19% of it every year, permanently. Ecologically, excluding deer from 200 ha concentrates 456 animals on the remaining 1,000 ha, raising density there to roughly 46/km² and worsening browse outside the fence. It also isolates the enclosed population and creates edge effects.
Judgement: do the cull, but at a size that actually crosses the regeneration threshold, and use temporary exclosures only on the most sensitive 20–50 ha rather than a perimeter fence. Concepts: carrying capacity, threshold, trophic cascade, resilience. The cull is reversible and cheap; the fence is expensive, permanent and displaces the problem.
3. The manager wants a single indicator to judge within three years whether the decision is working. Recommend one, reject one alternative, and justify both choices.
Read the worked answer
Recommend: oak sapling density in the 0.5–2 m height class, measured in fixed plots, together with browse rate.
Why: it is the state variable that the management action directly controls, it responds within two to three seasons, it can be measured objectively with repeatable plots, and it has a defensible target (returning towards the 1968 benchmark of 900/ha, with an interim target of ~250/ha). Browse rate separates 'few saplings' from 'saplings eaten', which distinguishes a deer problem from a light or soil problem.
Reject: black grouse counts. Grouse are a lagging, low-signal indicator. At 12 displaying males, year-to-year variation from weather and disturbance swamps any real trend, and detection is noisy. Grouse also respond to factors the reserve does not control.
Design: 40 permanent 10 × 10 m plots stratified by habitat, sampled each August, plus a control exclosure in each stratum to separate browsing from climate. Review after year 3; if sapling density is flat while browse rate stays above 50%, the cull is too small and must be increased or the exclosure strategy scaled up. The indicator must be tied to a pre-agreed trigger for changing course, otherwise it is just monitoring.