The researchers compared the hermaphroditic Matías Rodríguez population with the San Nicolás Tecoaco population, which had previously been classified as trioecious. The sites were about 6 km apart in Hidalgo, Mexico: Matías Rodríguez was located in pine–oak forest, while San Nicolás Tecoaco occupied xerophytic scrub. The observational and experimental study combined floral measurements, pollen and ovule counts, five controlled pollination treatments, fruit and seed measurements, germination assays and opportunistic observations of floral visitors.

All flowering plants observed in 2015 at both sites were hermaphroditic, including plants at the trioecious site. Male and female floral morphs had been documented there in other years, but none flowered during the study period. Consequently, the reproductive analyses for both populations were restricted to hermaphroditic flowers, and sex could not be included as a factor in the statistical models. Researchers sectioned 68 floral buds from Matías Rodríguez and 20 from San Nicolás Tecoaco for reproductive measurements, with each bud collected from a different plant.

Flowers at Matías Rodríguez had a modeled mean of 670 ovules per flower, with a 95% confidence interval of 611–736, compared with 448 ovules at San Nicolás Tecoaco, with a 95% confidence interval of 383–514. A negative-binomial model estimated an approximately 33% reduction in ovule production at the trioecious site. Pollen production did not differ detectably between the populations. The mean pollen–ovule ratio was 349 at Matías Rodríguez, with a 95% confidence interval of 296–412, and 479 at San Nicolás Tecoaco, with a wider interval of 360–645. The population effect on this ratio was marginal rather than conclusive.

The study tested automatic self-pollination, manual self-pollination, cross-pollination, supplemental pollination and open pollination. Fruit set was highest under open pollination in both populations, accounting for approximately 62–65% of flowers at Matías Rodríguez and 57–80% at San Nicolás Tecoaco. The trioecious population generally produced more seeds per fruit, with treatment means ranging from 128 ± 157 to 286 ± 188, compared with 83 ± 130 to 178 ± 148 in the hermaphroditic population. Estimated fruit volume also tended to be greater at San Nicolás Tecoaco, although treatment-specific comparisons varied between sites.

A beta-binomial mixed model detected higher germination among seeds from the trioecious population than from the hermaphroditic population. Estimated marginal germination proportions were 0.302 at San Nicolás Tecoaco, with a 95% confidence interval of 0.245–0.366, and 0.213 at Matías Rodríguez, with a 95% confidence interval of 0.169–0.265. The model estimated a 59% higher germination probability at the trioecious site, with an odds ratio of 1.59, and indicated substantial variation among plants. In contrast, a non-parametric comparison of medians, 0.29 versus 0.20, did not reach statistical significance.

Pollination treatment had no detectable effect on germination in the hermaphroditic population, with estimated probabilities ranging from 0.19 to 0.26. In the trioecious population, autonomous self-pollination had an estimated germination proportion of 0.44, with a 95% confidence interval of 0.29–0.60, compared with 0.24 for open pollination, with a 95% confidence interval of 0.15–0.35. The mixed model estimated that autonomous self-pollination increased germination relative to the control. Manual self-pollination had an estimated proportion of 0.37, while cross-pollination had a proportion of 0.22; neither was reported as significantly different from the control in the mixed-model comparisons. Overall, the treatments showed that autonomous selfing was functionally viable in both populations and that no inbreeding depression was detected in the evaluated early fitness traits.

Observed visitors included solitary bees in the genera Anthophora, Macrotera and Tetraloniella, as well as nitidulid beetles. Within Matías Rodríguez, the number of Macrotera visits was positively associated with germination, while Tetraloniella and Anthophora visits were negatively associated with it. Nitidulidae showed a marginal positive association. In San Nicolás Tecoaco, Macrotera showed only a marginal positive association, and the other visitor groups had no detectable associations with germination. These are statistical associations, not measurements of pollination effectiveness, and they do not establish that any visitor caused the differences in germination.

Visitor observations were opportunistic, sampling effort differed slightly between sites, and researchers did not quantify pollen deposition or standardized visitation rates. The study was also restricted to a single flowering season, with only hermaphroditic flowers analyzed in 2015 despite the trioecious population’s known male and female morphs. The sites differed in environment, sampling duration and number of observed flowers. The results therefore document population-level differences in reproductive allocation, early seed performance and visitor assemblages, but do not establish pollinator effectiveness, long-term evolutionary change or causes of the site differences.

Subject of Research: Comparative reproductive ecology of two Mexican Opuntia robusta populations

Article Title: Contrasting reproductive strategies and their functional consequences in two populations of Opuntia robusta (Cactaceae)

Article References: Mendoza-Reyes, C. H. S., Miranda-Pérez, A., Núñez-Farfán, J., & Janczur, M. K. (2026). Contrasting reproductive strategies and their functional consequences in two populations of Opuntia robusta (Cactaceae). BMC Plant Biology, 26(1), Article 1629. https://doi.org/10.1186/s12870-026-09625-0

Image Credits: AI Generated

DOI: 10.1186/s12870-026-09625-0

Keywords: Opuntia robusta, mating systems, autonomous self-pollination, pollen–ovule ratio, seed germination, floral visitors, reproductive allocation, trioecious population, hermaphroditic population, Hidalgo Mexico, Contrasting, reproductive

An international pattern long recognized by ecologists holds that pollen–ovule ratios, a conservative indicator of a plant’s breeding system, tend to rise with elevation across diverse floras. The conventional interpretation is that plants at altitude, facing sparse and unreliable pollinators, invest more heavily in pollen to increase the odds of outcrossing. But quantity is only half the story. Reproductive success depends not only on how much pollen is available but on how well that pollen performs, a distinction central to the pollen quantity–quality framework proposed by Aizen and Harder in 2007. Until now, simultaneous measurements of pollen quantity and quality within a single species along a natural elevational gradient have remained scarce.

Researchers from the Department of Botany at the University of Kashmir, led by Latif Ahmad Peer with Sadaf Wani and Taliyah Manzoor as equal first authors, set out to close that gap. Between July and August 2024, at the peak flowering season, the team sampled five populations of Thymus linearis across the Kashmir Valley, from a low-elevation site at 1,800 meters to the alpine heights of Pir ki Gali at 3,499 meters. At each site they measured sixteen vegetative and reproductive traits in fifteen randomly selected mature individuals, and took finer-grained pollen measurements from three individuals per site, a sample size comparable to other alpine reproductive studies. In total, the team counted and stained thousands of individual pollen grains under the microscope, using acetocarmine to assess stainability, a widely used comparative proxy for pollen functionality.

The results were unambiguous in their direction. The pollen–ovule ratio climbed from roughly 3,269 pollen grains per ovule at the lowest site to 4,740 at Pir ki Gali, an increase of about 45 percent that rose significantly with elevation. Over the same gradient, pollen stainability fell from 91.3 percent to 70.8 percent, a 22 percent reduction in functional quality. Plotted against each other, the two variables traced a strong negative relationship across populations, with the pollen–ovule ratio and stainability explaining more than 70 percent of each other’s variation. That direct relationship narrowly missed conventional statistical significance, a nuance the authors acknowledge candidly, but the opposing elevational trends were each individually significant and together are consistent with a genuine trade-off in which high-altitude plants compensate for deteriorating pollen quality by manufacturing more of it.

Every population examined fell above a pollen–ovule ratio of 2,900, placing all of them squarely in the xenogamous, or outcrossing, category under Cruden’s classic 1977 classification. What changed with elevation was the intensity of that investment. The authors suggest several non-exclusive mechanisms behind the quality decline. Intensifying ultraviolet-B radiation at altitude can damage DNA during the delicate stages of pollen development known as microsporogenesis, while the cold that characterizes high mountains is well known to disrupt normal male gamete formation. Nutrient-poor alpine soils may also force plants to choose between producing abundant pollen and producing functional pollen. Distinguishing among these mechanisms, the authors stress, will require experimental manipulation that goes beyond their correlational field design.

The second major surprise of the study concerns the leaves. A substantial body of literature, from the worldwide leaf economics spectrum onward, predicts that leaf size should shrink with elevation as plants adopt more resource-conservative phenotypes in cold, windswept environments. Thymus linearis defied that expectation. Three of the four measured leaf dimensions increased significantly toward the top of the gradient, and lower leaf width showed the most dramatic differentiation of any trait in the entire dataset, expanding from 0.18 centimeters at 1,800 meters to 0.46 centimeters at 3,499 meters, a 156 percent difference between the lowest and highest populations. The team interprets this through a photosynthetic-compensation hypothesis: at Pir ki Gali the growing season lasts only three to four months, compared with five to six months at lower elevations, and larger leaves may allow plants to maximize carbon gain during that compressed window, outweighing the costs of increased water loss and heat load.

The third pattern emerged in the plant’s overall architecture. Populations at mid-elevation sites such as Drung and Dara were taller, produced more secondary branches, and carried more inflorescences. At the highest site, the plants were stockier and more heavily invested in support structures: stem circumference increased by 71 percent and root circumference by 76 percent relative to the lowest site, while inflorescence numbers dropped by 64 percent. Regression analysis revealed that stem girth thickened by roughly 0.024 centimeters for every 100 meters of elevation gain, and root girth by 0.029 centimeters over the same distance. These structural traits were tightly coupled with one another and with leaf dimensions, correlation coefficients ranging from 0.97 to 0.99, painting a picture of a coordinated stress-tolerant syndrome rather than isolated trait shifts. The pattern is broadly consistent with the stress-tolerant expectations of Grime’s CSR framework, although the authors are careful to note that their measured traits do not permit a formal classification of ecological strategies.

To capture how these traits move together, the researchers performed a principal component analysis on fourteen morphological traits. The first two components accounted for a remarkable 88.17 percent of total variation, with the first axis alone explaining 65.62 percent and primarily representing a gradient of structural investment. The highest-elevation population occupied the extreme positive end of that axis, associated with the thickest stems, largest root systems, and broadest leaves, effectively the strongest stress-tolerant phenotype in the dataset. A second analysis that included the two pollen variables produced an essentially identical population structure and, tellingly, loaded the pollen–ovule ratio and pollen stainability in opposite directions on the first axis, providing independent multivariate corroboration of the quantity–quality trade-off.

The study’s most consequential contribution may be what it implies for interpreting community-level data. Previous work in southwestern China documented rising pollen–ovule ratios across 84 species along an elevation gradient and interpreted the trend as a shift toward outcrossing. The Kashmir results suggest such interpretations may be incomplete: if pollen quality declines as quantity rises, then elevated pollen–ovule ratios could partly reflect a compensatory response to stress rather than a straightforward increase in outcrossing. A recent global meta-analysis finding that reproductive responses to elevation are largely species-specific reinforces the value of single-species studies like this one, which can reveal mechanistic trade-offs invisible to broader surveys.

The authors are equally clear about the limitations. Pollen traits were measured in three individuals per population rather than fifteen, reflecting the labor-intensive nature of pollen counting, and acetocarmine staining measures stainability rather than true germination capacity. The observational design cannot disentangle which of the many covarying environmental factors along the gradient drives the observed patterns, nor can it separate phenotypic plasticity from genetically based local adaptation. Common-garden and reciprocal-transplant experiments, direct pollinator observations, and pollen germination assays are the logical next steps. Yet the climate implications are difficult to ignore. As temperatures rise, mountain species are expected to shift upward to track suitable conditions, and Thymus linearis may already be pressing against a ceiling: its highest-elevation population has the lowest pollen stainability of any site studied. If warming pushes populations further upslope or disrupts growing seasons, pollen quality could deteriorate further, creating a reproductive bottleneck precisely where species are forced to retreat. The pronounced differentiation among populations along the gradient, whether plastic or heritable, argues for conserving multiple populations across the full elevational range rather than one or two flagship sites, so that the functional diversity this thyme has evolved to cope with its mountains is not lost before it is fully understood.

Subject of Research: Elevational variation in pollen quantity, pollen quality, and stress-tolerant morphology in the Himalayan plant Thymus linearis

Article Title: Elevational shift in pollen quantity–quality trade-off and stress-tolerant morphology in Himalayan Thymus linearis

Article References: Wani, S., Manzoor, T., & Peer, L. A. (2026). Elevational shift in pollen quantity–quality trade-off and stress-tolerant morphology in Himalayan Thymus linearis. Discover Plants, 3(1), Article 389. https://doi.org/10.1007/s44372-026-00871-y

Image Credits: AI Generated

DOI: 10.1007/s44372-026-00871-y

Keywords: Thymus linearis, Kashmir Himalaya, elevational gradient, pollen-ovule ratio, pollen stainability, plant reproduction, leaf traits, stress tolerance, phenotypic variation, climate change, outcrossing, plant morphology