The research was carried out by Ayşe Özlem Tursun of Malatya Turgut Özal University, who combined experimental plant physiology with species distribution modeling, a pairing that remains surprisingly uncommon in climate-impact studies. Conventional distribution models treat a species as a passive follower of climate envelopes, predicting where conditions might be tolerable but saying little about whether the organism will perform better or worse once it arrives. By integrating experimentally derived physiological responses, Tursun’s framework attempts to capture how photosynthesis, growth and secondary metabolite production in lemon balm shift under the novel thermal and atmospheric regimes of the coming decades, and then translates those responses into a scenario-based sensitivity analysis of range change.

The modeling backbone of the study is the Maximum Entropy algorithm, or MaxEnt, one of the most widely used tools for mapping potential species distributions from presence-only records. Tursun assembled 117 spatially independent occurrence records for lemon balm, drawing on a formally documented download from the Global Biodiversity Information Facility, and paired them with bioclimatic variables from the WorldClim version 2.1 database. To keep the projections honest, model complexity was tuned using spatial block cross-validation, a technique that separates training and testing data geographically so the model cannot simply memorize local conditions. The resulting statistics were strong: a training area under the curve of 0.825, a test AUC of 0.878, and a True Skill Statistic of 0.620, values that indicate the model discriminates suitable from unsuitable habitat considerably better than chance.

Two climatic variables emerged as the decisive levers shaping where lemon balm can live. The minimum temperature of the coldest month, known among modelers as Bio6, contributed most by variable importance, while the annual mean temperature, Bio1, ranked highest by permutation importance. In plain terms, lemon balm’s Turkish distribution is governed above all by how cold winters get and how warm the year runs on average. That makes intuitive sense for a herb of Mediterranean and Irano-Turanian affinity, and it also explains the geography of the projected expansion: as winters soften and annual temperatures climb, the cold walls that currently fence the species into parts of western and southern Türkiye begin to dissolve.

The spatial pattern of the forecast is striking. Under all scenarios and both future time horizons, 2050 and 2070, the model projects progressive habitat expansion rather than contraction, with newly suitable territory emerging along the Black Sea coastline and across the northern Central Anatolian plateau. The expansion is pronouncedly poleward and upward in elevation, a signature consistent with the general expectation that species will track their thermal niches toward higher latitudes and altitudes. For a country whose medicinal and aromatic plant sector is economically and culturally significant, the map carries practical weight: it identifies where lemon balm cultivation might plausibly be established in the coming decades and where current strongholds could face shifting competitive and climatic conditions.

What lifts the study beyond a standard MaxEnt exercise is its treatment of carbon dioxide. Rising atmospheric CO2 does not merely warm the planet; it also fertilizes photosynthesis in many plant species, a physiological effect that pure correlative models routinely ignore. Drawing on earlier experimental work in which lemon balm was grown under different temperature and CO2 concentrations, Tursun ran a scenario-based sensitivity analysis to estimate how much of the projected expansion could be attributed to this fertilization effect. The answer was roughly 25.1 percentage points of the total expansion, a substantial share, though the analysis concluded that rising temperatures remain the dominant driver of the range shift. The result is a rare quantitative partitioning of the climatic and physiological components of a forecast range change.

Robustness was addressed through an ensemble assessment across three CMIP6 general circulation models. The coefficient of variation across the ensemble came out at just 5.6 percent, a low figure indicating that the expansion signal is not an artifact of one particular climate model’s quirks. The projections were run under two Shared Socioeconomic Pathways, SSP2-4.5 representing a moderate emissions trajectory and SSP5-8.5 representing a fossil-fuel-intensive future, and the direction of change held under both, with magnitude scaling with emissions. In a field where projections often swing wildly between climate models, that degree of agreement is notable and lends credibility to the headline number of a doubling of suitable area by mid-to-late century.

Yet the study’s own framing makes clear that a larger map is not an unalloyed good. Climate change poses what the author calls dual challenges to medicinal and aromatic plants: the contraction of suitable habitats on one hand, and the alteration of secondary metabolite biosynthesis under new thermal and atmospheric conditions on the other. For lemon balm, the commercial and medicinal value lies in its essential oils and phenolic compounds, and the scientific literature has long documented that environmental stressors, temperature and CO2 among them, can reshape the chemistry of such plants. A habitat that is climatically suitable may therefore produce raw material of different quality and potency than today’s harvests, a dimension that distribution maps alone cannot capture.

The methodological choices also deserve attention from practitioners. The occurrence data were spatially thinned to reduce sampling bias, the model was evaluated with metrics appropriate to presence-only modeling, and the threshold selection for converting continuous suitability scores into binary habitat maps followed established procedures for such data. The R script underlying the analysis was released as supplementary material, and the occurrence dataset carries a persistent citable DOI, making the workflow reproducible end to end. In an era when species distribution models increasingly inform conservation planning and agricultural policy, that transparency matters as much as the headline projection.

For Türkiye, the practical message is one of climate-smart agricultural planning. The identification of the Black Sea coast and northern Central Anatolia as zones of emerging suitability offers a forward-looking guide for growers, cooperatives and policymakers in the medicinal plant sector, suggesting where cultivation trials, land-use investments and conservation measures might be directed before the climate arrives. At the same time, the study is a reminder that even the winners of climate change live in a transformed world: a lemon balm field in 2070 will grow under different skies, breathe different air, and likely synthesize a different bouquet of compounds than its predecessors. The expansion of its habitat, dramatic as it is, is only the first chapter of the story.

Subject of Research: Projected habitat expansion of lemon balm in Türkiye under CMIP6 climate scenarios using integrated physiological and distribution modeling

Article Title: Integrating experimental physiological responses with species distribution modeling to forecast the future habitat expansion of Melissa officinalis L. in Türkiye under CMIP6 scenarios

Article References: Integrating experimental physiological responses with species distribution modeling to forecast the future habitat expansion of Melissa officinalis L. in Türkiye under CMIP6 scenarios. (n.d.). https://doi.org/10.1007/s00704-026-06613-0

Image Credits: AI Generated

DOI: 10.1007/s00704-026-06613-0

Keywords: lemon balm, Melissa officinalis, species distribution modeling, MaxEnt, CMIP6, climate change, Türkiye, CO2 fertilization, medicinal plants, habitat expansion, bioclimatic variables, SSP scenarios