Contemporary Evolution of Plant Growth Rate Following Experimental Removal of Herbivores
Evolutionary Ecology of Plant Defenses and Herbivore Interactions
Macroevolutionary and Microevolutionary Timescales:
Herbivorous interactions with plants extend across more than of evolutionary history, driving the diversification of an array of plant defense strategies.
For herbivory to guide the diversification of plant defense strategies across macroevolutionary epochs, natural selection exerted by herbivores must simultaneously direct microevolutionary adaptations within localized plant populations over ecological timescales ().
The foundational prerequisites for short-term adaptive microevolution are broadly present across plant taxa:
Ubiquitous heritable genetic variation for both resistance and tolerance traits.
Measurable differential selection pressures imposed directly on these traits by herbivores.
Dichotomy of Plant Defensive Strategies:
Resistance Traits: Chemical, physical, or morphological attributes that directly deter herbivores or diminish the total quantity of damage sustained.
Tolerance Traits: Physiological mechanisms that mitigate, compensate for, or neutralize the negative fitness consequences incurred following herbivore damage.
Empirical Gaps in Direct Experimental Evidence:
Prior insights into plant microevolution driven by herbivores have relied disproportionately on observational studies.
Observational approaches have established positive correlations between herbivory pressure and the expression of defensive secondary metabolites or tolerance.
Observational correlations are insufficient to prove causality; theoretical models demonstrate that spatial correlations can arise from alternative evolutionary processes independent of biotic natural selection.
Manipulative field experiments that explicitly exclude herbivores over extended durations are essential to isolate direct and indirect evolutionary outcomes.
Direct Versus Indirect Selection Pressures by Vertebrate Herbivores:
Vertebrate grazers impose evolutionary pressures fundamentally distinct from those imposed by phytophagous arthropods.
Direct Selection: Direct tissue consumption, defoliation frequency, and close cropping alter morphological architecture, palatability, and relative tissue loss.
Indirect Selection: Chronic grazing cascades through the ecosystem by altering abiotic and biotic parameters, including:
Soil nutrient dynamics and cycling.
Total net primary productivity ().
Canopy light penetration.
Interspecific competitive hierarchies and community dominance.
Eco-Evolutionary Feedbacks:
Contemporary evolutionary trait shifts within a single plant population have the potential to feed back into the community, restructuring interspecific interactions, competitive exclusion, and species coexistence.
Ecological Context of the Silwood Park Acid Grassland System
Geographic and Environmental Setting:
Located at Silwood Park, Imperial College, Ascot, Berkshire, United Kingdom ().
Habitat consists of species-poor lowland acidic grasslands situated on coarse, sandy, nutrient-depleted soils with co-occurring monocots and eudicots.
Keystone Role of the European Rabbit (Oryctolagus cuniculus):
Introduced to England thousands of years ago, with populations sustained at high densities at Silwood Park since the 1950s.
Functions as an ecological keystone species by arresting ecological succession, preventing open grasslands from transitioning into closed-canopy woodland.
Rabbits graze plants down to ground level, suppressing individual plant size and preventing reproductive flowering.
Quantitative Impacts of Rabbit Grazing on Sward Structure and Ecosystem Properties:
Vegetation Height: Rabbit grazing reduces mean community sward height by ( in ungrazed swards down to in grazed swards).
Standing Biomass: Grazing reduces mean standing aboveground biomass by ( dry weight per in ungrazed swards down to dry weight per in grazed swards).
Primary Productivity: Long-term exclusion of rabbits for results in an approximate increase in net primary productivity (annual aboveground biomass production).
Soil Fertility: Rabbit exclusion over drives an approximate increase in soil nutrient availability.
Community Shifts and Population Trajectories Under Grazing:
Grazing Increasers: Species that increase in relative abundance in the presence of continuous grazing (Holcus lanatus, Senecio jacobaea, and Trifolium repens). These species typically exhibit fast growth rates and high physiological tolerance to repetitive defoliation.
Grazing Decreasers: Species that decline significantly under rabbit grazing pressure (Achillea millefolium, Vicia sativa, and Festuca rubra).
Grazing Neutrals (Stable Species): Species whose relative abundance remains invariant regardless of grazing pressure (Rumex acetosa, Stellaria graminea, and Veronica chamaedrys).
Biology and Demography of Rumex acetosa L. (Polygonaceae):
Perennial dioecious herb (common sorrel) maintaining stable community cover under both grazed ( cover) and rabbit-excluded ( cover) conditions (, ).
Reproduction within the grassland is overwhelmingly dominated by clonal vegetative growth; grazing completely suppresses flowering and seed production.
Inside exclosures, flowering and seed set occur freely, yet dense perennial root networks and light competition heavily restrict seedling recruitment.
Population differentiation across the area operates under clonal lineage selection, with local adaptation occurring across spatial scales of just a few meters.
Long-Term Manipulative Field Experiment and Greenhouse Testing
Exclosure Chronosequence Setup:
Twenty-two distinct rabbit exclosures were established across the Silwood Park grassland landscape between 1984 and 2009.
Maintained continuously without rabbit grazing for durations ranging from ().
Fencing manipulations selectively excluded European rabbits while leaving plant populations unmanipulated.
Seed Provenance and Sampling Protocol:
Seeds of Rumex acetosa were harvested in 2009 from haphazardly sampled fruiting stems across all 22 populations.
To obtain seeds representing the continuous-grazing baseline (where grazing normally prevents flowering), new temporary exclosures were erected in 2009 within the grazed matrix to permit flowering and seed development.
Each exclosure was designated an independent evolutionary population.
Common Greenhouse Environment:
Conducted in fall 2009 at North Carolina State University to test for genetic differentiation in a common environment.
Seeds were germinated simultaneously on moist filter paper to standardize germination timing and remove maternal effects related to emergence speed.
Seedlings were transplanted individually into , diameter pots containing a sand:potting soil mixture (Fafard no. 2).
Soil was supplemented with approximately of slow-release fertilizer (Osmocote N:P:K).
Populations were represented by an average of (ranging from depending on seed availability).
Factorial Experimental Treatments:
Implemented a fully factorial design per population with 6 replicates:
Simulated Herbivory: All aboveground biomass was clipped with a razor blade to above the soil surface at day and day post-transplanting.
Interspecific Competition: A single individual of Oenothera biennis (common evening primrose, Onagraceae) of a standardized North American genotype was planted as a phytometer at a fixed distance of from the R. acetosa seedling.
Harvest: At day , entire root and shoot systems were destructively harvested, rinsed thoroughly of soil, dried, and weighed.
Mathematical Formulations of Measured Traits:
Intrinsic Growth Rate: Calculated as total dry biomass accumulation divided by total growth duration:
Relative Tolerance to Herbivory: Quantified on a relative biomass basis:
Relative Competitive Ability: Quantified on a relative biomass reduction basis:
Phytometer Response: Evaluated by contrasting the final taproot biomass of O. biennis grown in monoculture versus in competition with R. acetosa.
Biochemical Assays and Maternal Investment Analysis
Tissue Preparation:
Plants grown from seed in the standardized greenhouse environment were harvested at the stage of .
Leaf tissues were freeze-dried using a VirTis Genesis freeze dryer and homogenized into a fine powder.
Protein-Binding Soluble Tannin Quantification:
Measured via a radial diffusion assay based on protein precipitation activity.
of homogenized leaf tissue was extracted three times in aqueous acetone at room temperature.
Acetone was evacuated using a vacuum centrifuge; the residual extract was redissolved in deionized water.
Gels were prepared with agarose and bovine serum albumin () and cast in petri dishes at per dish.
Center wells ( diameter) were punched into the solidified gel, and of extract was loaded.
Plates were incubated at for , and precipitation ring diameters were measured with digital calipers to determine the percentage of protein-binding phenolics.
Oxalate Quantification:
Quantified using the kinetic-catalytic spectrophotometric reaction method.
of homogenized tissue was suspended in of deionized water and heated in an water bath for .
Samples were centrifuged, and the supernatant was diluted to a concentration with deionized water.
Microplate setup: of crystal violet solution was mixed with of diluted extract.
Reactions were initiated by dispensing of a solution containing and .
Absorbance was read at precisely post-initiation (BioTek PowerWave X model UPRWIE).
Maternal Provisioning Controls:
To evaluate potential non-genetic maternal seed provisioning differences among exclosure ages, 50 seeds per exclosure were weighed.
Exclosure age exhibited no significant relationship with seed mass (, , ).
Synchronization of germination dates and the absence of seed mass differences eliminate maternal effects as drivers of subsequent trait variation.
Evolutionary Trajectories of Plant Growth Rate

Contemporary Evolution of Intrinsic Growth Rate:
Intrinsic growth rate decreased linearly as a function of the number of years populations were protected from rabbit grazing.
This evolutionary reduction in growth rate was consistent across both defoliation and competition treatments:
Undamaged Plants without Competition: Growth rate declined by over 26 years (, , ).
Damaged Plants without Competition: Growth rate declined by over 26 years (, , ).
Damaged Plants with Competition: Growth rate declined by over 26 years (, , ).
Undamaged Plants with Competition: Growth rate declined by over 26 years, demonstrating a consistent negative trajectory (, , ).
Rates of Microevolutionary Change:
Evolutionary rates of phenotypic change reached a maximum of (standard deviations of change per year, assuming a generation time of ).
Calculated across the 26-year chronosequence, growth rate decreased by at least per year.
These rates align with typical rates of contemporary microevolution reported for wild plant populations responding to strong selective agents.
Evolutionary Stasis in Secondary Metabolites, Tolerance, and Competition

Stasis of Constitutive Chemical Defenses:
Protein-binding soluble tannin concentrations in leaf tissue exhibited no relationship with exclosure age (, , ).
Oxalate concentrations in leaf tissue showed no directional evolutionary change over time (, , ).
Evolutionary rates for both defensive chemical metabolites were near zero.
Stasis of Tolerance to Defoliation:
Simulated herbivory severely reduced overall plant biomass by (, , ).
Despite this strong treatment effect, tolerance to defoliation did not change across populations as a function of exclosure age (, , ).
Mixed-effects modeling showed no significant interaction between exclosure age and damage treatment (, , ).
Interspecific Competitive Dynamics:
The presence of the phytometer Oenothera biennis reduced Rumex acetosa biomass by (, , ).
Competitive ability against O. biennis did not evolve following the exclusion of rabbits (, , ).
Mixed-effects modeling revealed no interaction between exclosure age and competition treatment (, , ).
Competition was asymmetric: Rumex acetosa presence had no detectable competitive effect on the biomass of O. biennis taproots (, , ).
Correlations and Trait Trade-Offs:
A significant negative genetic correlation was detected between intrinsic growth rate and tolerance to herbivory (, , ).
No significant correlations were observed between intrinsic growth rate and secondary metabolite concentrations (tannins or oxalates, all ).
No significant correlation emerged between competitive ability and tolerance to herbivory ().
Theoretical Mechanisms of Growth Rate Defense and the Resource Availability Hypothesis
Intrinsic Growth Rate as an Evolutionary Defense:
A defensive trait is defined as any phenotypic characteristic that enhances plant fitness in environments with active herbivore pressure.
Intrinsic growth rate functions directly as a defense against severe vertebrate grazing, independent of induced compensatory regrowth.
Under continuous rabbit cropping, rapid tissue generation enables genotypes to continuously replace lost photosynthetic tissue, preventing exhaustion of root reserves and avoiding local mortality.
Chronic defoliation actively selects for fast-growing genotypes while purging slow-growing genotypes.
In the absence of herbivores, rapid growth incurs physiological or metabolic costs, selecting for slower-growing genotypes over time.
Challenges to the Resource Availability Hypothesis (RAH):
The Resource Availability Hypothesis assumes that plant growth rate evolves primarily as an adaptation to abiotic resource availability (e.g., nutrients, light, water), and that this growth rate then sets limits on defense allocation.
RAH predicts that high resource availability favors fast growth and reduces allocation to constitutive defenses, generating an inverse correlation between growth rate and chemical defenses.
The Silwood Park system violates this fundamental premise: growth rate itself evolved as an anti-herbivore defense rather than an abiotic adaptation.
Relaxing herbivory led to an evolutionary decrease in growth rate despite increases in soil nutrients () and primary productivity () inside exclosures.
When herbivores directly select for accelerated growth, RAH paradoxically predicts that herbivores would select for reduced defenses.
Constraints on the Evolution of Tolerance:
The trade-off between growth rate and tolerance () indicates that fast-growing plants may have limited capacity for further induced compensation.
Selection for high growth rate under intense rabbit grazing can directly constrain the evolution of tolerance mechanisms.
Synthesis of Herbivore Removal Studies Across Plant Taxa
Empirical Literature Comparisons:
Prior common-garden studies tracking phenotypic evolution following the exclusion of vertebrate herbivores were conducted exclusively on grasses (Poaceae):
Carman and Briske (1985): Cattle removal; Growth rate: ; Competitive ability: ; Tolerance: .
Polley and Detling (1990): Prairie dog removal; Growth rate: ; Competitive ability: ; Tolerance: .
Tomas et al. (2000): Cattle removal; Growth rate: ; Competitive ability: ; Tolerance: .
Briske and Anderson (1992): Cattle removal; Growth rate: ; Competitive ability: ; Tolerance: .
Jefferies (1988): Lesser snow goose removal; Growth rate: ; Competitive ability: ; Tolerance: .
Loreti et al. (2001): Cattle removal; Growth rate: ; Competitive ability: ; Tolerance: .
Polley and Detling (1988): Prairie dog removal; Growth rate: ; Competitive ability: ; Tolerance: .
Detling and Painter (1983): Prairie dog removal; Growth rate: ; Competitive ability: ; Tolerance: .
Jaindl et al. (1994): Cattle removal; Growth rate: ; Competitive ability: ; Tolerance: .
Jaramillo and Detling (1988): Prairie dog removal; Growth rate: ; Competitive ability: ; Tolerance: .
Painter et al. (1989): Prairie dog removal; Growth rate: ; Competitive ability: ; Tolerance: .
Turley et al. (This System): Rabbit removal on Rumex acetosa (Polygonaceae); Growth rate: ; Competitive ability: ; Tolerance: .
Distribution of Growth Rate Evolutionary Trajectories:
Across the 12 vertebrate herbivore removal studies evaluated:
documented an evolutionary increase in plant growth rate following herbivore removal.
documented an evolutionary decrease in plant growth rate following herbivore removal.
showed no significant evolutionary change in plant growth rate.
Mechanistic Divergence in Growth Rate Evolution:
Decreased Growth Rate After Herbivore Removal: Occurs when continuous, severe grazing directly selects for high intrinsic growth to sustain rapid leaf replacement, while rapid growth becomes maladaptive or costly once grazing ceases.
Increased Growth Rate After Herbivore Removal: Occurs when herbivory selects for high investment in physical or chemical defenses that trade off with growth, such that excluding herbivores releases plants from defense allocation costs, enabling selection for rapid vegetative growth.
Eco-Evolutionary Dynamics and Community-Level Consequences
Absence of Detectable Interspecific Feedbacks:
Despite contemporary evolution reducing Rumex acetosa growth rate by to , this evolutionary shift did not feed back to alter competitive suppression of the phytometer Oenothera biennis or modify R. acetosa's relative competitive performance.
Contemporary microevolution does not inevitably generate cascading eco-evolutionary feedbacks within surrounding multi-species plant communities.
Community Context and Ecological Redundancy:
Pronounced eco-evolutionary dynamics are most frequently documented in simple ecosystems, chemostats, or microcosm communities with specialized trophic interactions.
In species-rich grassland assemblages characterized by high niche overlap and diffuse competitive interactions, shifting environmental conditions tend to produce rapid community compositional turnover rather than observable feedbacks driven by intra-specific evolution.
Intra-specific microevolution within individual species can broaden ecological niche breadth, stabilizing population persistence and dampening community shifts under variable herbivore pressure.