The use of plant-derived volatile organic compounds (VOCs) to deter herbivorous animals from crops, gardens, and natural habitats represents a long-standing area of inquiry in both applied pest management and plant chemical ecology. The theoretical appeal is well-founded: aromatic plants have co-evolved with herbivore pressure over millions of years and have developed complex, constitutive, and inducible VOC profiles that function—in part—as chemical defences against herbivory. Many aromatic species are empirically observed to be avoided by large mammalian herbivores under naturalistic field conditions, a pattern reported consistently across independent horticultural and ecological literature.

Despite this theoretical promise, the published literature evaluating botanical VOC preparations as deterrents against large herbivorous mammals has returned largely discouraging results. Studies assessing commercial and homemade odour repellents against wild boar (Sus scrofa), deer (Cervidae), and other species have consistently found preparations to be either ineffective or effective only transiently, with rapid habituation documented across multiple independent studies [1]-[3]. These findings have shaped a prevailing consensus that scent-based botanical deterrents are not a viable management tool.

Here, we identify a critical and, to our knowledge, previously unremarked structural flaw in the methodological foundation of this literature: every published study in this field has used a processed, extracted, harvested, or synthesised plant material as its VOC source. Not one has tested the deterrent effect of a living, metabolically active, growing aromatic plant in a controlled experimental setting. We argue that this is not a minor methodological detail but a fundamental category error—one that renders the existing consensus on botanical deterrent inefficacy premature and potentially invalid as a general conclusion about aromatic plants.

This manuscript presents: 1) a systematic characterisation of the six principal dimensions by which a living plant’s VOC output differs from its extracted equivalent; 2) a formal critique of the conclusions that extraction-based studies can and cannot support; 3) a review of existing evidence that indirectly supports the living-plant deterrence concept; and 4) a structured experimental agenda to test the hypothesis directly.

A review of published studies on botanical VOC deterrence against large herbivorous mammals reveals a consistent and unremarked pattern across all study types: the VOC source is invariably a preparation, not a plant.

The VOC emissions of a living plant are not a static chemical mixture but a dynamic, regulated, and environmentally responsive output of an active metabolic system. The following six dimensions are each independently documented in the plant physiology and chemical ecology literature, and each represents a way in which an extraction fundamentally fails to represent the living organism as a deterrent source.

If the six dimensions described in Section 3 produce a VOC output that is qualitatively and quantitatively distinct from that of an extracted equivalent—and the plant physiology literature provides strong a priori reason to expect this—then the following conclusions of the existing botanical deterrence literature must be understood as methodologically constrained:

“Botanical scent repellents are ineffective against wild boar” [1]-[3]: Established using applied preparations only. Valid as a conclusion about those specific preparations. It cannot be validly extended to living aromatic plants, which have never been tested.“Animals rapidly habituate to scent-based deterrents”: Documented for static, degrading preparations. Kimball and Taylor (2010) [12] identify four mechanisms of olfactory repellence—neophobia, irritation, conditioned aversion, and flavour modification—of which only conditioned aversion and irritation are likely to remain effective under sustained herbivore pressure. A living plant presenting a continuously replenished, HIPV-augmented VOC signal activates multiple mechanisms simultaneously.Single-compound identification studies: Identifying the primary active compound from an extract may isolate a relevant molecule while missing the full synergistic VOC blend of the living organism, which may differ substantially in both composition and concentration.Cross-species deterrence rankings: Comparative rankings of plant deterrent potential based on extraction assays may not reflect the relative deterrent efficacy of the corresponding living organisms, particularly if species differ in HIPV upregulation capacity or compound lability profiles.

The most significant implication is structural: if living plants prove materially more effective deterrents than their extracted equivalents, then the foundational experimental unit of botanical repellent research requires revision. Controlled studies must be designed around the living organism, not its preparation.

No study has directly compared a living plant with its extracted equivalent as an herbivore deterrent. However, multiple independent bodies of evidence provide indirect but substantive support for the proposition that they would differ significantly in deterrent efficacy.

Testing the living-plant deterrence hypothesis requires experimental designs not previously employed in this field. The following programme is proposed:

Study 1—Direct three-way comparison: Present target herbivores (e.g., Susscrofa or Cervidae) with three simultaneous conditions: 1) a living aromatic plant—with Ribes nigrum L. as the recommended initial candidate given its well-documented constitutive VOC profile and field-observed mammalian avoidance, 2) an extraction of the same plant applied to a scent-neutral inert substrate of equivalent size, and 3) an untreated control. Behavioural data will be analysed using Cox proportional-hazards models to compare time-to-first-approach and time-to-first-contact across conditions. This is the critical experiment that directly tests whether the living plant and its extract are equivalent deterrents.

Study 2—HIPV induction trial: Expose living candidate plants—beginning with Ribes nigrum—to simulated herbivory (mechanical damage plus exogenous jasmonic acid application) and measure changes in headspace VOC profile before and after induction using Gas Chromatography-Mass Spectrometry (GC-MS). Present induced versus non-induced living plants to target herbivores in paired trials. VOC composition data will be analysed using multivariate ordination permutational multivariate analysis of variance (PERMANOVA) to quantify profile shifts; herbivore behavioural responses will be compared using paired Wilcoxon signed-rank tests.

Study 3—In situ headspace versus extraction VOC profiling: For candidate deterrent species, conduct parallel GC-MS analysis of: a) dynamic headspace from intact living plants (Solid-Phase Microextraction in situ), b) SPME surface sampling of living plant emissions, and c) solvent extraction of equivalent plant mass. Ribes nigrum is recommended as the priority candidate given the published characterisation of its bud and leaf VOC profile. Compound-by-compound differences across conditions will be assessed using principal component analysis (PCA) and non-parametric permutation tests. Submit TRPA1-activating candidate compound classes for in vitro mammalian TRPA1 activity screening.

Study 4—Longitudinal biomass and VOC output tracking: Monitor total headspace VOC emission from individual perennial aromatic plant specimens—with Ribes nigrum as the priority species—across three or more growing seasons using annual SPME headspace sampling. Leaf area index and above-ground dry mass will be recorded at each sampling point. Linear mixed-effects models will be used to test whether total VOC emission scales with biomass across seasons, with individual plant identity as a random effect.

Study 5—Field border planting trial: Establish living aromatic border plantings—using Ribes nigrum as the initial candidate, given its native range overlap with Sus scrofa in Europe and its documented deer-resistance—around agricultural or garden plots in areas with confirmed boar activity. Monitor herbivore damage via camera traps, track surveys, and crop-loss assessment over a minimum of two full growing seasons. Compare with extract-treated and untreated control plots. Deterrence efficacy will be assessed using generalised linear mixed models (GLMMs) with plot as the random effect and treatment, season, and plant age class as fixed factors. We also note that there could be unintended consequences from overstimulation. We argue that because VOC production escalates and varies when the plant is disturbed, due to the unique blend of HIPVs, the overstimulation will not occur. This is an aspect of the hypothesised increased deterrent effect of the upregulation. However, this will require additional consideration and monitoring if mammals are documented contacting the proposed barrier shrub on multiple occasions.

The botanical VOC deterrence literature has studied the plant but not the organism. In every published study, living chemistry has been extracted, killed, and asked whether its residue repels animals. That the answer has so often been negative or inconsistent may reflect the inadequacy of the experimental approach more than it reflects the deterrent potential of aromatic plants. A living perennial shrub—continuously biosynthesising its VOC complement, dynamically responding to herbivore contact with upregulated HIPV emission, projecting a concentration gradient into the surrounding air, and growing more chemically productive every growing season—is a categorically different deterrent source from a spray bottle of extract. The two have never been compared.

The six dimensions identified here—continuous biosynthesis, HIPV upregulation, environmental modulation, biomass scaling, compound lability, and plume architecture—are each individually documented in the plant physiology and chemical ecology literature. Together they constitute a strong a priori case that extraction-based deterrence studies are not valid proxies for living-plant deterrence experiments.

The right experiment—presenting a living aromatic plant to a large herbivore under controlled conditions—has not yet been done. This manuscript proposes that it should be, and that it may substantially revise current understanding of the deterrent potential of the plant kingdom.

This work received no external funding.

C.D.: Conceptualisation, hypothesis formulation, methodological critique, literature review, manuscript preparation.

No experimental data were generated or analysed in this study. This manuscript constitutes a methodological critique and research proposal.

No animal or human subjects research was conducted in preparation of this manuscript. Proposed Studies 1, 2, and 3 would require institutional animal ethics approval prior to commencement.

During the preparation of this work, the author used Claude/Sonnet 4.6 in order to format the research into an acceptable version for publication/peer-review. The author also used Google Scholar in order to locate existing literature relevant to the manuscript. After using these tools/services, the author reviewed and edited the content as needed and took full responsibility for the content of the publication.