Flipping through field notebooks…
Tracks long-term shifts in montane and alpine plant communities using historical herbarium records, GIS sampling, and nearest neighbor analysis to understand how vegetation composition and pollinator-relevant floral traits have changed across elevation gradients in western Colorado.
Alpine and montane plant communities in the Gunnison Basin sit at the leading edge of climate change in the Rocky Mountains. Short growing seasons, deep snowpacks, steep elevation gradients, and sharp topographic boundaries make these ecosystems unusually sensitive to even small shifts in temperature and precipitation. When conditions change, plants can respond by shifting their ranges upslope, by changing the timing of flowering, or by reshuffling which species grow together. Documenting these changes matters not only for understanding how mountain ecosystems work, but also for land managers in the Gunnison and Uncompahgre National Forests who must plan for shifting habitat types, altered forage for wildlife, and changing fire and water regimes.
A central challenge in this research is that detecting long-term change requires long-term data, and few mountain landscapes have been continuously monitored for many decades. Researchers therefore rely heavily on herbarium collections — preserved plant specimens that were picked in the field, pressed flat, dried, mounted on archival paper with notes on location, elevation, and date, and then stored in climate-controlled cabinets. A specimen collected in 1950 with a precise location and elevation provides a snapshot of where that species grew at that time, allowing modern botanists to ask whether the same species still occurs there today, or whether it has moved upslope, downslope, or disappeared entirely. Combined with old vegetation transects and forest classifications, herbaria turn dusty archives into powerful tools for measuring ecological change.
To interpret these changes, ecologists also study how species interact within communities. Concepts such as temporal overlap (whether species flower or grow active at the same time of year), and methods such as nearest neighbor analysis (which compares the species growing close to a target plant against what would be expected by chance) help reveal whether plants compete, facilitate one another, or simply track shared environmental conditions. Together, herbarium data, resurveys of historical plots, and analyses of community structure form the toolkit for asking how Gunnison Basin vegetation is changing — and why.
Much of the early scientific foundation for this work came from descriptive vegetation classifications in the 1980s, when researchers systematically mapped the forest and meadow communities of central and western Colorado. Alexander (Alexander, 1987) consolidated Colorado forest vegetation into 15 series and 161 habitat and community types, providing a standardized vocabulary for later ecologists. Companion classifications for the Arapaho and Roosevelt National Forests and for the Gunnison and Uncompahgre National Forests described 37 forest habitat types directly relevant to the Gunnison Basin, organized to guide both research and forest management. Even earlier work by Alexander and colleagues on partial cutting in old-growth spruce-fir forests helped establish a management context in which long-term vegetation monitoring could be interpreted.
Preserved plant specimens that are picked in the field, pressed flat, dried and attached to archival-grade paper with relevant information recorded, t...
Statistical method comparing species composition near target individuals to background composition to identify spatial associations
Method for extracting and analyzing historical herbarium specimen data to reconstruct the life and career patterns of past collectors, including tempo...
Using ArcGIS to digitize habitat polygons and generate random sampling points within appropriate community types.
Quantitative plant abundance data were collected from the same 121 sites at two time periods separated by 65 years (1948-1952 and 2012-2014) in the Co...
Quantitative plant abundance data were collected from the same 121 sites at two time periods separated by 65 years (1948-1952 and 2012-2014) in the Co...
Vascular Plants curated within the Rocky Mountain Biological Laboratory Herbarium. Specimens date to 1929 and are primarily from the East River and up...
Alongside these landscape-scale classifications, individual researchers established baseline observations of meadow plants and their pollinators near Gothic. Early reports characterized spruce forest, aspen woods, sagebrush, and grassy slope communities around RMBL (Olney, 1958), and detailed studies of individual species — from the growth substances of Veratrum tenuipetalum (Olney, 1968) to the activity of the least chipmunk (Williams, 1977) — built the natural-history foundation on which later, more quantitative comparisons would rest.
The most striking result to emerge from this body of work is that Gunnison Basin plant communities have changed substantially over the past several decades. By resurveying 121 transects originally established by Jean Langenheim between 1948 and 1952, Zorio et al. (Zorio et al., 2016) found that all four major community types — sagebrush, spruce-fir, upland-herbaceous, and alpine — had experienced significant changes in species composition and dominance after 65 years. Bare ground, grasses, and shrubs had increased, forbs had declined at higher elevations, and species' mean elevations had shifted upward by 41 meters on average. Alpine species in particular had begun to expand downward into subalpine meadows, while lower-elevation species pressed upslope. Independent work using herbarium records reached compatible conclusions: Kasunic (Kasunic, 2014) documented that four of six surveyed species, including Lupinus argenteus and Vaccinium scoparium, occurred at higher mean elevations in 2014 than in historical records, and that the maximum recorded elevation of Castilleja sulphurea had risen from 11,607 to 11,691 feet.
These shifts in where plants grow are accompanied by patterns in how plant communities are assembled. Across five Rocky Mountain meadow communities, floral traits showed significant phenotypic overdispersion compared with randomly assembled communities (Stansberry, 2009), suggesting that competition for pollinators helps structure which species coexist. Consistent with this, Gray et al. (Gray et al., 2018) found systematic shifts in pollinator-relevant floral coloration along an elevation gradient near RMBL: reflectance of shorter wavelengths (ultraviolet through blue) declined linearly with elevation, color saturation in both bee and fly visual spaces increased with elevation, and the proportion of species perceived as bee-blue decreased upslope while bee-UV-green species peaked at intermediate elevations of 3,400–3,600 m.
Reproductive biology adds a further layer to community-level change. In gynodioecious populations of Geranium richardsonii — a species with both female and hermaphroditic plants — Williams et al. (Williams et al., 2000) showed that female flowers are smaller, receive fewer pollinator visits, and capture less pollen than hermaphrodites, and that self-pollination within hermaphrodites accounts for much of the difference. Building on earlier observations of pollinator choice in this species (Drew, 1997), herbarium-based work confirmed that on average only about 7% of G. richardsonii flowers in a population are female (Kasunic, 2014). Such fine-scale reproductive patterns determine which species can persist and migrate as climate shifts community composition.
Early work in the late 20th century focused on classifying communities and describing individual species. Studies from the 2000s and 2010s, including Williams et al. (Williams et al., 2000), Kasunic (Kasunic, 2014), Zorio et al. (Zorio et al., 2016), and Gray et al. (Gray et al., 2018), marked a clear pivot: researchers began using historical baselines — herbarium specimens, mid-century transects, and pollinator records — to quantify change rather than simply describe the present. Recent work since 2020 has broadened the lens further, examining the deeper landscape context in which plant communities sit. Clausen (Clausen, 2020) mapped 37 mountain passes crossing the continental divide around the Colorado River headwaters, with floor elevations from 2,655 to 3,825 meters, and Clausen (Clausen, 2021), (Clausen, 2021) extended this work to the Yampa-Colorado and North Platte-South Platte drainage divides, arguing that flow reversals during regional uplift shaped the topography that today controls where alpine habitat exists.
The emerging frontier combines these threads: pairing high-resolution historical data from herbaria and old transects with GIS-based digitization of habitat polygons and statistical methods such as nearest neighbor analysis, researchers are increasingly able to ask not just whether species have moved, but which species are moving together, which are being left behind, and how the underlying topography constrains future migration.
Many important questions remain. We do not yet know how much of the observed upslope shift in Gunnison Basin plants is driven by warming temperatures versus changes in snowpack, grazing, or competitive interactions among species. It is unclear whether the loss of forbs at high elevations represents a transient reshuffling or a longer-term decline that will alter pollinator networks, and whether species with specialized reproductive systems such as Geranium richardsonii can keep pace with climate as their populations move. The relationship between flower color shifts along elevation and the actual pollinator communities encountered also remains to be tested directly. Over the next decade, expanding herbarium digitization, repeating mid-century transects on a regular cycle, and linking these biological records to the topographic and hydrologic framework of the Gunnison Basin offer the most promising path toward predicting which alpine communities will persist, which will transform, and which may disappear.
Alexander, R. R. (1987). Classification of the forest vegetation of Colorado by habitat type and community type. →
Alexander, R. R., Colo. (1972). Initial partial cutting in old-growth spruce-fir. →
Alexander, R. R., et al. (1988). Forest vegetation of the Gunnison and parts of the Uncompahgre National Forests: a preliminary habitat type classification. →
Clausen, E. (2020). Topographic Map Analysis of Mountain Passes Crossing the Continental Divide Between Colorado River Headwaters and North and South Platte River Headwaters. Journal of Geography and Geology. →
Clausen, E. (2021). Topographic Map Analysis of the North Platte River-South Platte River Drainage Divide Area, Western Larimer County, Colorado. Earth Science Research. →
Clausen, E. (2021). Yampa River-Colorado River Drainage Divide Origin Determined from Topographic Map Evidence, Southern Routt County, Colorado. Open Journal of Geology. →
Drew, J. (1997). Pollinator choice of flower size and sexual morph in Geranium richardsonii. →
Stansberry (2009). Factors influencing floral traits in Rocky Mountain meadows: competition, environmental filtering, and phylogeny. →
Gray, M., et al. (2018). Consistent shifts in pollinator-relevant floral coloration along a Rocky Mountain elevation gradient. Journal of Ecology. →
Hess, K., et al. (1986). Forest vegetation of the Arapaho and Roosevelt National Forests in central Colorado: a habitat type classification. →
Kasunic, J. (2014). Expansion of herbaria data based on historically surveyed herbaceous plants in the Crested Butte area, Colorado. →
Olney, H. (1958). Combined reports to cover Englemann spruce forest, immature grassy slope, aspen woods, and sage community. →
Olney, H. (1968). Growth substances from Veratrum tenuipetalum. Plant Physiology. →
Williams, C. F., et al. (2000). Floral dimorphism, pollination, and self-fertilization in gynodioecious Geranium richardsonii (Geraniaceae). American Journal of Botany. →
Williams, D. (1977). Activity and feeding behavior of the least chipmunk, Eutamias minimus, near Gothic, Colorado. →
Zorio, S., et al. (2016). Sixty-five years of change in montane plant communities in Western Colorado, USA. Arctic, Antarctic, and Alpine Research. →