Exploitation Ecosystems Hypothesis

The Exploitation Ecosystems Hypothesis (EEH) (also known as the Oksanen–Fretwell model) is a quantitative theory of food-web structure and trophic regulation formulated by Finnish ecologist Lauri Oksanen, Stephen D. Fretwell, Joseph Arruda, and Pekka Niemelä in 1981 (building on Fretwell's 1977 food-chain length synthesis). Generalizing the classic 1960 Green World Hypothesis of Hairston, Smith, and Slobodkin (HSS) across global gradients of abiotic primary productivity, the Exploitation Ecosystems Hypothesis proposes that the number of dynamically functional trophic levels in an ecosystem is governed by potential Net Primary Productivity (NPP), and that as trophic levels are added one by one along a productivity gradient, the regulation of plant and herbivore biomass alternates systematically between bottom-up resource limitation and top-down consumer control.

Because endothermic vertebrate herbivores and carnivores require minimum threshold densities of prey biomass to offset their basal metabolic maintenance costs, harsh, low-productivity environments support short food chains (one or two trophic links), whereas warm, moist, high-productivity environments support three- or four-link food chains. Consequently, the HSS "green world" represents one specific regime—the three-trophic-level forest zone—along a broader continuum of biospheric productivity regimes.

Consumer–Resource Differential Equations Across Productivity (G)

Oksanen and colleagues (1981) formalized the Exploitation Ecosystems Hypothesis using a vertical chain of Rosenzweig–MacArthur differential equations describing the standing biomass of plants (P), herbivores (H), primary carnivores (C), and secondary/apex carnivores (T) as a function of potential primary productivity G (proportional to maximum plant carrying capacity K):

dP / dt = r P (1 − P / K) − f_H(P) · H

dH / dt = [ e_H f_H(P) − m_H ] · H − f_C(H) · C

dC / dt = [ e_C f_C(H) − m_C ] · C − f_T(C) · T

Here, f_i(X) = a_i X / (1 + a_i h_i X) is the saturating Holling Type II functional response, e_i is the trophic assimilation-production efficiency, and m_i is the per-capita metabolic maintenance and mortality rate of consumer guild i.

For a consumer guild i to persist (dX_i / dt ≥ 0 at low consumer density), its underlying resource biomass must exceed its zero-net-growth isocline (ZNGI) threshold X* = m_i / (a_i [e_i − m_i h_i]). As potential primary productivity K increases from zero across a latitudinal, elevational, or precipitation gradient, the ecosystem steps through four discrete dynamic regimes separated by sharp threshold transitions.

The Four Discrete Productivity Regimes of the EEH

1. Regime I — One-Link Barren / Polar Desert (K < P*_H; H* = 0, C* = 0):

In hyper-arid deserts, high-alpine talus fields, and high-arctic polar deserts where annual above-ground net primary production falls below ~20 to 50 g dry matter m⁻² yr⁻¹, sparse plant cover K lies below the minimum foraging threshold P*_H required to sustain resident vertebrate herbivores. Standing plant biomass P* = K increases linearly with abiotic amelioration and is resource-limited.

1. Regime II — Two-Link Herbivore-Grazed Tundra / Arid Steppe (P*_H < K < K_C; H* > 0, C* = 0):

Once primary productivity surpasses P*_H (roughly 50 to 700 g m⁻² yr⁻¹, characteristic of Arctic tundra, alpine snowbeds, and arid shortgrass steppes), resident herbivores (such as lemmings, voles, reindeer, and saiga antelope) establish permanent populations, but total herbivore productivity remains too low to sustain resident endothermic carnivores (H* < H*_C). In this two-link world, herbivores are food-limited and graze plant biomass down to a flat, constant ceiling P* = P*_H. Any further increase in primary productivity K is channeled into higher herbivore biomass H*, while standing plant biomass remains low and heavily cropped (a "brown/grazed world").

1. Regime III — Three-Link Predator-Protected Green Forest (K_C < K < K_T; H* > 0, C* > 0, T* = 0):

When primary productivity exceeds the threshold K_C (>700 g m⁻² yr⁻¹, characteristic of boreal, temperate, and tropical forests), herbivore production becomes sufficient to support resident specialist and generalist carnivores (weasels, foxes, lynx, raptors, and insectivorous birds/parasitoids). Carnivores pin herbivore biomass at the predator threshold H* = H*_C, releasing vegetation from grazing control. Standing plant biomass P* resumes climbing with productivity K, reproducing the classic HSS Green World Hypothesis.

1. Regime IV — Four-Link Apex-Predator System (K > K_T; T* > 0):

In productive pelagic aquatic ecosystems (and certain productive subtrophic guilds), a fourth trophic level of secondary carnivores or piscivores (T) establishes, suppressing primary carnivores (C*) and releasing herbivores (H*) to once again exert strong top-down grazing pressure on primary producers (P*).

1. Fennoscandian Alpine Altitudinal Gradients (Oksanen 1981–2000)

Lauri Oksanen and colleagues tested the EEH along steep altitudinal productivity gradients on Mount Nuolja and across northern Fennoscandian fells (Finnmarksvidda and Joatka), where summer temperature and snow-free growing season length control primary productivity across a few hundred meters of elevation:

• High-Alpine Snowbeds (Regime I): Near the upper limit of vascular plant growth (~1,100–1,400 m elevation), dwarf willow (Salix herbacea) and moss mats experience minimal vertebrate grazing because forage density is too sparse for Norwegian lemmings (Lemmus lemmus) to maintain positive winter energy balance.
• Middle-Alpine Tundra Heaths (Regime II): At intermediate elevations with moderate productivity, lemmings and gray-sided voles (Myodes rufocanus) reach high peak densities in the absence of resident least weasels (Mustela nivalis) and stoats (Mustela erminea). Long-term wire exclosure experiments demonstrated that excluding rodent herbivores caused dramatic increases in standing bryophyte and dwarf-shrub biomass, confirming intense top-down herbivore control in the two-link zone.
• Low-Alpine Betula Scrub and Subalpine Birch Forest (Regime III): In productive lower-slope birch forests where mustelid predators, red foxes (Vulpes vulpes), and avian raptors maintain resident populations, rodent densities are held down by predation, herbivore exclosures show weak effects on standing plant biomass, and plant competition for light dominates community structure.

2. Freshwater Phytoplankton–Zooplankton–Fish Gradients

In freshwater limnology, Lennart Persson and colleagues (1992) and Orlando Sarnelle (1992) tested the staircase predictions of the EEH across lakes of varying total phosphorus concentration (spanning ultra-oligotrophic to hyper-eutrophic productivity). In oligotrophic two-level lakes lacking planktivorous fish, phosphorus enrichment increases Daphnia zooplankton biomass while chlorophyll-a (phytoplankton biomass) remains flat; in mesotrophic three-level lakes containing planktivorous cyprinids or coregonids, Daphnia are suppressed and chlorophyll-a rises steeply with phosphorus; and in four-level lakes stocked with piscivorous pike (Esox lucius) or perch (Perca fluviatilis), planktivores are suppressed, restoring high Daphnia grazing and low phytoplankton standing crops.

Unpalatable Plant Refuges and Endotherm vs. Ectotherm Energetics

Ecological critiques by Gary A. Polis (1999), Mathew A. Leibold (1989, 1996), and Oswald J. Schmitz refined the strict step-function predictions of the original 1981 EEH model:

1. Plant Chemical Defense Heterogeneity (Leibold's Partial Refuge Model): As primary productivity K increases across a landscape, plant communities shift from palatable, fast-growing herbaceous graminoids toward slow-growing, heavily lignified and chemically defended woody shrubs and trees. When edible and defended plant guilds coexist (Leibold, 1989), nutrient enrichment increases both herbivore biomass and unpalatable plant biomass simultaneously, smoothing the sharp step-function plateaus of the homogeneous EEH into positive slopes across all trophic levels.
2. Ectotherm vs. Endotherm Metabolic Scaling: Oksanen (1988) and Yodzis and Innes (1992) noted that the classical EEH productivity thresholds apply most sharply to endothermic vertebrate food chains, where high basal metabolic rates (m_H, m_C) prevent predators from persisting in low-productivity tundra. In contrast, ectothermic invertebrate food chains (spiders, carabid beetles, parasitoid wasps, and desert scorpions) can reduce their metabolic maintenance rates by an order of magnitude during cold or dry seasons and sustain three- or four-link arthropod food webs even in arid deserts.

How does the Exploitation Ecosystems Hypothesis extend the Green World Hypothesis?

Whereas the 1960 Green World Hypothesis (HSS) assumed that all terrestrial ecosystems possess three functional trophic levels and therefore experience top-down predator protection of plants, the Exploitation Ecosystems Hypothesis shows that the HSS state requires primary productivity above ~700 g dry matter m⁻² yr⁻¹. Below that productivity threshold (as in arctic tundra or arid steppes), resident predators cannot persist, collapsing the food web to two trophic levels where food-limited herbivores heavily graze the vegetation.

What is the Paradox of Enrichment in the Exploitation Ecosystems Hypothesis?

As originally discovered by Michael Rosenzweig (1971), when primary productivity K increases inside a saturating Holling Type II consumer–resource chain, the vertical predator isocline shifts from the rising side to the falling (destabilizing) side of the prey hump-shaped isocline. Within the EEH, moving from low to moderate productivity within Regime II (or from moderate to high productivity within Regime III) destabilizes the point equilibrium into large-amplitude limit cycles—explaining why multi-year lemming and vole population cycles are most pronounced in moderately productive tundra and boreal zones.

Why do Arctic islands without wolves or foxes validate Regime II of the EEH?

High-Arctic islands such as Svalbard (Norway) and Bylot Island (Canada) exemplify two-link or simplified tundra systems where mammalian predators of ungulates (wolves) are absent or rare. On Svalbard, endemic Svalbard reindeer (Rangifer tarandus platyrhynchus) are regulated from below by winter forage availability and rain-on-snow ice crusting rather than predation, heavily grazing lichen and graminoid mats down to the two-link EEH equilibrium ceiling P*_H.

1. Fretwell, S. D. (1977). The regulation of plant communities by the food chains exploiting them. Perspectives in Biology and Medicine, 20(2), 169–185. https://doi.org/10.1353/pbm.1977.0087
2. Oksanen, L., Fretwell, S. D., Arruda, J., & Niemelä, P. (1981). Exploitation ecosystems in gradients of primary productivity. The American Naturalist, 118(2), 240–261. https://doi.org/10.1086/283817
3. Leibold, M. A. (1989). Resource edibility and the effects of predators and productivity on the outcome of trophic interactions. The American Naturalist, 134(6), 922–949. https://doi.org/10.1086/285019
4. Persson, L., Diehl, S., Johansson, L., Andersson, G., & Hamrin, S. F. (1992). Trophic interactions in temperate lake ecosystems: a test of food chain theory. The American Naturalist, 140(1), 59–84. https://doi.org/10.1086/285403
5. Oksanen, L., & Oksanen, T. (2000). The logic and realism of the hypothesis of exploitation ecosystems. The American Naturalist, 155(6), 703–723. https://doi.org/10.1086/303354