Vegetation Response to Glacier Retreat in the Trinity Alps

By Justin Garwood, Mike van Hattem, Tony LaBanca, Ken Lindke, Gordon Leppig, and Michael Kauffmann

A BIOGEOGRAPHIC CROSSROADS

High in the Klamath Mountains of northwest California—where floras from the Cascades, Sierra Nevada, Coast Ranges, and Great Basin converge, and more than 3,500 plant taxa have been documented—the landscape has long been shaped by geologic and climatic complexity (Kauffmann and Garwood 2022). Each ridgeline and cirque supports a distinctive combination of rock types, microclimates, and species. Because of the range’s coastal proximity and latitude, its highest peaks accumulate snowpacks comparable to those of the Sierra Nevada or Cascades, yet here, at the southern edge of the Pacific Northwest, deep winter snow persists several thousand feet lower in elevation. Over time, the region has absorbed major environmental shifts while sustaining an exceptional concentration of endemic plants and communities that endured through both glacial and interglacial periods. For most of the last ten thousand years, change across these elevations unfolded gradually.

That trajectory has now shifted. As across much of the mountain West, species adapted to cool summers face sustained warming, snowpacks melt earlier and return later, subalpine seeps dry more quickly, and meadows show increasing sensitivity to drought (Selkowitz et al. 2002). Transformations that once occurred over millennia are now apparent within decades.

One of the clearest local signals of this acceleration was the disappearance, in 2022, of Grizzly Glacier below Thompson Peak (2,744 m; 9,002 ft)—the last glacier in the Klamath Mountains (Garwood et al. 2020). Thompson Peak, a granitic summit within the Trinity Alps Wilderness, is the second-highest peak in the Klamath Mountains behind Mount Eddy (2,773 m; 9,037 ft). The glacier’s loss marked not only a climatic threshold but also the opening of new ecological space. For decades, Grizzly Glacier occupied a bare, ice-sculpted void amid otherwise continuous vegetation, including forests of whitebark pine (Pinus albicaulis) extending upslope above the ice. These forests reach the summit of Thompson Peak on the south slope—a rare juxtaposition in California, where glaciers elsewhere only persist well above treeline. As the ice receded, newly exposed terrain provided a unique opportunity to directly observe early plant establishment on ground previously covered by ice (Ficetola et al. 2024).

THE LAST GLACIER OF THE TRINITY ALPS—A BRIEF HISTORY

Some of the highest perennial surface waters in the Klamath Mountains originate above 2,450 m (8,000 ft), historically sustained by two small but persistent ice bodies: Salmon Glacier and Grizzly Glacier. Though separated by less than one kilometer (0.6 miles), their meltwaters followed divergent paths—one draining into the South Fork of the Salmon River, the other flowing toward the North Fork of the Trinity River—before ultimately reuniting at Weitchpec in the Klamath River.

Grizzly Glacier was always modest in size. Evidence suggests it formed around 1300 CE, as the Medieval Climate Anomaly gave way to the cooler conditions of the Little Ice Age (Sharp 1960; Graham 2013). Snow that lingered through successive summers gradually compacted into firn and then into a slow-moving mass of ice that met the technical definition of a glacier.

By the late nineteenth century, Grizzly Glacier reached its most recent maximum extent—approximately 24 hectares (60 acres), as indicated by the innermost moraine (Graham 2013)—and began to retreat soon thereafter. Throughout the twentieth century, scientists debated whether Grizzly and Salmon glaciers remained active, yet both persisted longer than many comparable ice features elsewhere in the region. As late as the 1970s, Grizzly Glacier still covered roughly six hectares (15 acres), shrinking to 4.6 hectares (11 acres) by 1994 (Garwood et al. 2020).

That persistence gradually eroded under sustained warming, shifting precipitation patterns, and repeated drought. By the end of 2015—during the most severe regional drought in millennia (Griffin and Anchukaitis 2014; Ullrich et al. 2019)—Salmon Glacier had vanished (Garwood et al. 2020). Continued drought, followed by the River Complex Fire in 2021, further accelerated Grizzly Glacier’s decline, as ash from burned whitebark pine forests darkened the ice surface and increased melt.

By the close of the 2022 melt season, Grizzly Glacier had disappeared entirely, exposing fresh moraine, polished granite slabs, and glacier foreland that had not received direct sunlight since long before Euro-American settlement of California. This loss was especially notable: the Klamath Mountains retained their final glacier into the twenty-first century at an elevation approximately 517 m (1,700 ft) below the lowest extant glaciers in the Sierra Nevada and on Mount Shasta (Garwood et al. 2020).

Study Goals

The objectives of this study were to:

• Document the vascular plant flora within the innermost Little Ice Age (LIA) moraine and the surrounding Grizzly Glacier foreland.
• Identify plant taxa occupying terrain newly released from ice and compare this flora with historical collections by William J. Ferlatte.
• Determine which plant taxa have persisted since Ferlatte’s work and which have established following glacier retreat.
  Assess how species richness and diversity change with increasing distance from former glacier margins.

By integrating Ferlatte’s historical floristic records with our decade-long effort to develop a comprehensive contemporary flora and apply plot-based ecological methods, we created a temporal benchmark—a detailed snapshot of vegetation development spanning the glacier’s final ~130 years of retreat.

Methods

Since 2012, our team conducted thirteen backpacking expeditions into the remote Grizzly Glacier cirque. Documentation of the glacier’s retreat combined on-the-ground GPS surveys with airplane overflights, aerial and satellite imagery, and repeat photography (Garwood et al. 2020). Botanical fieldwork formed the core of the project and included revisiting Ferlatte’s general collection localities, establishing quantitative vegetation plots, collecting voucher specimens, and assembling an up-to-date vascular plant flora for the study area (Figure 1).

THE NEWLY REVEALED FLORA—PATTERNS AND ECOLOGICAL MEANING

Where ice once pressed against stone, the landscape now consists of raw till and skeletal soils—fluvial silts washed from beneath the glacier, cobbles pushed forward by past ice movement, and polished granite slabs smoothed by centuries of abrasion. Meltwater channels braid across this newly exposed terrain, creating a fine-scaled mosaic of wet and dry microhabitats within the 17-hectare (42-acre) study area (Figure 1).

A Floristic Snapshot of a Newly Revealed Landscape

Across the inner moraine and glacier foreland, we documented 82 plant taxa representing 63 genera and 29 families, encompassing five life forms—trees, shrubs, herbs, graminoids, and ferns. All recorded taxa are native to California (Table 1, Table 2).

Asteraceae was the most species-rich family, accounting for 13% of the flora, followed by Poaceae at 10%. Species were unevenly distributed among families: nearly half of all taxa belonged to just six families, while nine families were represented by a single species. This pattern reflects both the ecological heterogeneity of the site and the early-seral nature of habitats exposed by glacier retreat.

Ecologically, the flora is characteristic of subalpine communities, including meadow specialists, talus-dwelling herbs, and species adapted to bedrock outcrops. The lower boundary of the study area (2,347 m; 7,700 ft) lies well within the subalpine zone of the Klamath Mountains, which Ferlatte (1974) and Sawyer (2006) define as elevations above approximately 2,103 m (6,900 ft).
Nearly half of the documented taxa also occur broadly across western North American mountain ranges. These include whitebark pine, mountain hemlock (Tsuga mertensiana), bog laurel (Kalmia polifolia), pink mountain-heather (Phyllodoce empetriformis), hairy arnica (Arnica mollis), wandering daisy (Erigeron glacialis var. glacialis), Jeffrey’s shooting star (Primula jeffreyi), western pasqueflower (Anemone occidentalis), showy sedge (Carex spectabilis), Mertens’ rush (Juncus mertensianus), and pullup muhly (Muhlenbergia filiformis).

An additional 25% of taxa are restricted to the California Floristic Province, particularly the Sierra Nevada, southern Cascades, and Klamath Mountains. These include Shasta red fir (Abies magnifica var. shastensis), Parish’s yampah (Perideridia parishii), broad-seeded rockcress (Boechera howellii), mountain jewel flower (Streptanthus tortuosus), Sierra laurel (Leucothoe davisii), quill-leaved lewisia (Lewisia leeana), Sierra primrose (Primula suffrutescens), Brewer’s reedgrass (Calamagrostis breweri), and California squirreltail (Elymus elymoides var. californicus).

Two taxa exhibited notable distributional disjunctions. Lyall’s goldenweed (Tonestus lyallii), widespread in the northern Rocky Mountains, occurs at only a handful of sites in California, including the study area. Bud saxifrage (Micranthes bryophora), common at high elevations in the Sierra Nevada, is rare in both the Klamath Mountains and the central Rocky Mountains.

SPECIES PERSISTING FROM FERLATTE’S FLORA

A core group of long-lived subalpine species persisted in the same locations where Ferlatte recorded them in 1974. These included widespread circumboreal taxa—holly fern (Polystichum lonchitis), alpine goldenrod (Solidago multiradiata), mountain sorrel (Oxyria digyna), creeping sibbaldia (Sibbaldea procumbens), purple reedgrass (Calamagrostis purpurascens), and alpine timothy (Phleum alpinum)—which together constituted about 11% of the flora. Regionally endemic species, including Whitney’s tarweed (Hazardia whitneyi var. discoidea), Gray’s catchfly (Silene grayi), Mount Eddy stonecrop (Sedum kiersteadiae), cobwebby paintbrush (Castilleja arachnoidea), Merriam’s alumroot (Heuchera merriamii), strawberry saxifrage (Saxifragopsis fragarioides), and Piper’s woodrush (Luzula piperi), were also consistently present.

Four taxa that reach their southernmost range limits within or near the Klamath Mountains—Lemmon’s holly fern (Polystichum lemmonii), Cascade aster (Doellingeria ledophylla var. covillei), yellowish lousewort (Pedicularis bracteosa var. flavida), and pincushion beardtongue (Penstemon procerus var. brachyanthus)—persisted in habitats similar to those described by Ferlatte (1974). This underscores the resilience of these cold-adapted species under shifting climate conditions. Importantly, every taxon Ferlatte recorded within the upper cirque was rediscovered in our surveys.

Newly Observed or Early Establishing Species

In 2014, we documented the uppermost distribution of vascular plants below Grizzly Glacier by mapping 183 plant occurrences along the entire lower glacier margin (Figure 1). Glacier retreat over the preceding decades was documented by Garwood et al. (2020) using aerial imagery spanning 59 years (1955–2014). Of the mapped plant locations, 56% extended beyond the 1955 glacier margin, and 38% extended beyond the 1972 margin. No plants occurred within the 1994 glacier boundary; however, the average distance from these early-establishing plants to the 2014 ice edge was just 54 meters (range = 18–144 m) (Figure 1).

Species occurring closest to the 2014 glacier margin included club-fruited willowherb (Epilobium clavatum), pullup muhly, Tolmie’s saxifrage (Micranthes tolmiei), showy sedge, green sedge (Carex vernacula), Merten’s rush, and mountain sorrel (see Figure 2).

Across the study area, we documented ten plant taxa not recorded by Ferlatte (Table 2)—a modest but meaningful addition given the relatively small spatial extent of our surveys compared to Ferlatte’s broader collections within the Grizzly Lake cirque. These taxa reflect two general ecological pathways: early-establishing species associated with recently exposed substrates and generalist species expanding into newly available habitats.

Table 2: Download Entire Flora (PDF)

Beyond the earliest species occurring along the lower glacier margin, additional early-establishing taxa occupied the upper foreland as ice retreated. These included rock sword fern (Polystichum imbricans ssp. imbricans), holly fern, Sierra willow (Salix eastwoodiae), Scouler’s willow (Salix scouleriana), dunhead alpine sedge (Carex phaeocephala), green sedge, and Piper’s woodrush. These species were most frequently associated with disturbed ground, early-seral substrates, or moist microsites adjacent to meltwater channels.

Generalist species expanding into newly exposed moist, gravelly areas included Lee’s lewisia, bog laurel, club-fruited willowherb, Lewis’ monkeyflower (Erythranthe lewisii), mountain sorrel, Jeffrey’s shooting star, showy sedge, Drummond’s rush (Juncus drummondii), and Merten’s rush.

Together, these patterns of plant establishment and expansion closely parallel observations from other recently deglaciated landscapes, where species track shifting moisture gradients and take advantage of newly exposed substrates (Zumsteg et al. 2012; Mainetti et al. 2022).

MORAINE AND FORELAND PLOT ANALYSIS

In addition to compiling a comprehensive flora, we quantified patterns of plant distribution and relative species occurrence by sampling 125 spatially balanced, randomized plots in 2014. Plots were divided between the inner moraine and glacier foreland, reflecting the expectation that plant assemblages would differ across these contrasting geomorphic surfaces (Figure 1).

Vascular plants were recorded in 87 of the 125 plots. Fifty of the plots contained non-vascular mosses, an early-seral group commonly associated with recently deglaciated terrain (Jones and Henry 2003). Crustose lichens were detected in only two foreland plots, consistent with studies showing that lichen abundance increases with surface age and substrate stability (Haugland and Beatty 2005). In total, 42 plant species—approximately 51% of the 82 taxa documented in the full flora—were detected in plots, highlighting the rarity and patchy distribution of many species across the study area.

Across all plots, the most frequently encountered species were club-fruited willowherb, showy sedge, and pullup muhly (Table 3, Figure 2). Species richness increased with distance from the former glacier margin. When moraine and foreland plots were combined, the upper half of the study area (0–154 m from the ice) contained 22 species, while the lower half (154–310 m from the ice) contained 37 species. Overall, the moraine supported 41% greater species richness (39 species) than the foreland (23 species) (Table 3).

Foreland Patterns

The foreland contained no trees and was strongly structured by meltwater availability. As ice retreated over the past century, newly established streams created narrow but persistent wet habitats that supported early-establishing species. In the upper foreland (0–136 m from the ice), 11 species were recorded; three were closely tied to streams, while the remainder consisted of habitat generalists capable of occupying both wet and dry microsites.

Many of these species were widely distributed across the foreland, reflecting broad ecological tolerance (Figure 3).
In the lower foreland (136–272 m from the ice), 22 species were recorded, including 12 species not present in the upper foreland plots. None of these were obligate wetland species; most were generalists or upland taxa, suggesting increasing habitat differentiation and soil development with distance from the ice.

Moraine Patterns

Vegetation on the moraine differed markedly from the foreland (Figure 3). Most moraine plots were dominated by upland species adapted to well-drained slopes, although facultative wetland species occurred where fine sediments, shade, or seepage created localized moisture. Mesic species were confined to plots near the moraine base, particularly in large talus with deep, shaded cavities and small seeps.

The upper half of the moraine (0–155 m from the ice) contained 14 species, six of which were found only in this zone. The lower half (155–310 m) supported 33 species, including 25 species unique to that zone. A small portion of the terminal moraine supported a distinct dwarf shrub community with well-developed soils, characterized by ericaceous shrubs and shrubby conifers—features consistent with later-successional conditions.

Ecological Implications

Together, these results reveal multiple distinct plant communities within a small area, shaped by distance from ice, substrate stability, and moisture availability. Many species exhibited broad niche breadth, occupying both wetland and upland settings—likely facilitated by the cool subalpine climate and extended runoff season. Mosses played an important facilitative role in wetter microsites, particularly for species dependent on fine, moist substrates. As snow and ice continue to decline, some wet habitats may become more ephemeral, with implications for species reliant on sustained moisture.

REVISITING FERLATTE—LINKING PAST AND PRESENT

William J. Ferlatte’s Flora of the Trinity Alps (Ferlatte 1974) remains one of the foundational botanical works for the region. His species lists and descriptive notes shaped generations of field botanists and provided one of the earliest comprehensive floristic treatments of the Trinity Alps high country. That work, however, was conducted in a landscape still influenced by a larger glacier and by perennial snowfields that persisted longer into the summer season.

Comparing past and present required careful interpretation. Ferlatte (1974) often recorded broad localities rather than precise coordinates, yet the continuity between his flora and ours was striking. Every species he documented within the upper cirque was also recorded during our modern surveys, indicating that the core subalpine flora has persisted through more than five decades of climatic warming. This stability mirrors patterns observed in other high-elevation floras, where long-lived, stress-tolerant species endure despite substantial environmental change (Franzén et al. 2019; Mainetti et al. 2022).

More revealing than persistence, however, were the new arrivals—species that Ferlatte (1974) did not record and could not reasonably have overlooked. Notably, the 15 willow individuals documented in our surveys, some approaching the size of a small vehicle, were absent from his flora. These willows now occupy drainage lines in the modern foreland, growing as close as 150 m from the 2014 glacier margin, providing strong evidence of establishment following late twentieth- and early twenty-first-century glacier retreat. Their expansion is consistent with early-seral shrub establishment documented in glacier forelands worldwide (Fickert 2020; Bayle et al. 2023).

Together, Ferlatte’s historical records and our contemporary, plot-based surveys reveal a plant community anchored by long-term residents but increasingly reshaped by new arrivals exploiting newly exposed substrates. This integrated perspective provides a critical baseline for tracking ecological responses to continued glacier loss and climate warming in the Trinity Alps.

CLIMATE CHANGE AT THE CROWN OF THE KLAMATH RANGE

The disappearance of Grizzly Glacier is both a symbol and a data point—a local expression of the global acceleration of ice loss (Zemp et al. 2019). Subalpine environments are often described as “islands in the sky,” and concepts from island biogeography apply well to these systems: limited area, strong environmental boundaries, and heightened sensitivity to isolation and climate change (Helm et al. 2020). As conditions warm, the effective “island size” of high-elevation habitats contracts, dispersal pathways narrow, and species already confined to small patches face increasing pressure.

Like true islands, subalpine systems respond rapidly to shifts in temperature, moisture, and disturbance regimes. Studies of glacier forelands across Europe and North America consistently document rapid ecological change, including high rates of species turnover and upslope migration as warming intensifies (Franzén et al. 2019; Stibal et al. 2020).

In the Klamath Mountains, the stakes are particularly high. The region supports one of the richest conifer assemblages anywhere (Sawyer 2006, Kauffmann and Garwood 2022) and its meadows, serpentine savannas, and formerly glaciated lake basins harbor relict species and narrow endemics shaped by millennia of climatic stability. Accelerating warming compresses ecological niches, forces snowpack-dependent species upslope, and reshapes the fine-scale mosaic of communities that defines the high country (Selkowitz et al. 2002).

For conservation practitioners, the newly exposed glacier foreland represents both warning and opportunity. Understanding how plant communities assemble on freshly deglaciated terrain can help anticipate broader ecological shifts across the subalpine zone (Robbins and Matthews 2009). These insights can inform monitoring priorities and refine predictions about which species are likely to expand, contract, or disappear from the highest elevations as climate change continues.

CONCLUSION—WHAT REMAINS AFTER THE ICE

Where the glacier once crept, summer now reveals a mosaic of rock, water, and young plants. The loss of Grizzly Glacier marks an ecological threshold—the disappearance of the last glacier in the Klamath Mountains and the opening of a new chapter in mountain ecology.

This study stands at that threshold. By integrating Ferlatte’s historical perspective with contemporary, plot-based floristics, we present a portrait of a landscape in transition. The glacier’s absence is unmistakable, yet what follows is abundant life—organisms improvising, exploring, and establishing in terrain once shaped by ice. These patterns mirror those documented in deglaciated landscapes worldwide, where newly exposed ground becomes a testing ground for ecological assembly (Ficetola et al. 2024).

The glacier foreland will continue to evolve under warming climates and shifting hydrology, with species abundance and diversity changing across the gradient from moraine to bedrock. Because of this 14-year effort, however, the plants that appeared as the glacier declined and ultimately disappeared have been carefully documented, vouchered, and archived at the Cal Poly Humboldt Herbarium, providing a durable record for future comparison.

In the high, reflective light of the Trinity Alps, both absence and arrival become teachers. What persists—and what takes hold after the ice—invites us to keep watching closely. 

Download Supplemental Data (Plot Floristics, Metadata, and Procedures [zipfile])

• Justin Garwood, California Department of Fish and Wildlife. Justin.Garwood@wildlife.ca.gov
• Mike van Hattem, California Department of Fish and Wildlife. Michael.vanHattem@wildlife.ca.gov
• Tony LaBanca, California Department of Fish and Wildlife (retired). tlabanca@sbcglobal.net
• Ken Lindke, California Department of Fish and Wildlife. Kenneth.Lindke@wildlife.ca.gov
• Gordon Leppig, California Department of Fish and Wildlife (retired). gtl1@humboldt.edu
• Michael Kauffmann. Michaelkauffmann.net

REFERENCES

• Bayle, A., et al. 2023. “Local Environmental Context Drives Heterogeneity of Plant Early Succession on Alpine Glacier Forelands.” Biogeosciences 20: 1649–1669.
• Ferlatte, William J. 1974. A Flora of the Trinity Alps, Northern California. Master’s thesis, University of California, Berkeley.
• Fickert, Thomas. 2020. “Common Patterns and Diverging Trajectories in Primary Succession of Plants in Eastern Alpine Glacier Forelands.” Diversity 12 (5): 188.
• Ficetola, Gentile Francesco, et al. 2024. “The Development of Terrestrial Ecosystems Emerging After Glacier Retreat.” Nature 332: 336–342.
• Franzén, Markus, et al. 2019. “Rapid Plant Colonization of the Forelands of a Vanishing Glacier Is Strongly Associated with Species Traits.” Arctic, Antarctic, and Alpine Research 51 (1): 366–378.
• Garwood, Justin, et al. 2020. “20th-Century Retreat and Recent Drought-Accelerated Loss of Mountain Glaciers and Perennial Snowfields in the Trinity Alps, California, USA.” Northwest Science 94: 42–61.
• Graham, J. T. 2013. Late Holocene Glacial Advances in the Klamath Mountains, Northern California, Determined from ¹⁰Be Cosmogenic Exposure Dating and
• Dendrochronology. Master’s thesis, California State University, Northridge.
• Griffin, Daniel, and Kevin J. Anchukaitis. 2014. “How Unusual Is the 2012–2014 California Drought?” Geophysical Research Letters 41 (24): 9017–9023.
• Haugland, J. E., and Stephen W. Beatty. 2005. “Vegetation Establishment, Succession, and Microsite Frost Disturbance on Glacier Forelands within Patterned Ground
• Chronosequences.” Journal of Biogeography 32 (1): 145–153.
• Helm, Aveliina, Markus Franzén, Petra Dieker, and Julian Schrader. 2020. “Vegetation Development on Recently Emerged Glacier Forelands of the Ålmajallojekna Glacier.” Arctic, Antarctic, and Alpine Research 52 (3): 381–395.
• Jones, G., and Gregory H. R. Henry. 2003. “Primary Plant Succession on Recently Deglaciated Terrain in the Canadian High Arctic.” Journal of Biogeography 30: 277–296.
• Kauffmann, Michael, and Justin Garwood, eds. 2022. The Klamath Mountains: A Natural History. Kneeland, CA: Backcountry Press.
• Kincaid, T. et al. 2008. SPSURVEY: spatial survey design and analysis. R package version 2.0.
• Mainetti, Andrea, Silvia Ravetto Enri, and Michele Lonati. 2022. “Vegetation Trajectories in Proglacial Primary Successions within Gran Paradiso National Park.” IBEX – Journal of Mountain Ecology 2022: 1–18.
• O’Kane, Katriina, and Gregory HR Henry. 2024. “Directional succession and species-specific patterns observed in repeat study of vascular plants at three glacier foreland chronosequences in the Canadian High Arctic.” Arctic Science 10 (4): 764–777.
• R Development Core Team. 2014. R: A language and environment for statistical computing. R Foundation for Statistical Computing.
• Robbins, J. A., and John A. Matthews. 2009. “Pioneer Vegetation on Glacier Forelands in Southern Norway.” Journal of Vegetation Science 20 (5): 889–902.
• Sawyer, John O. 2006. Northwest California: A Natural History. Berkeley: University of California Press.
• Selkowitz, David J., Donald B. Fagre, and Brian A. Reardon. 2002. “Interannual Variations in Snowpack in the Crown of the Continent Ecosystem.” Hydrological Processes 16: 3651–3666.
• Sharp, Robert P. 1960. “Pleistocene Glaciation in the Trinity Alps of Northern California.” American Journal of Science 258 (5): 305–340.
• Stibal, Marek, et al. 2020. “Glacial Ecosystems Are Essential to Understanding Biodiversity Responses to Glacier Retreat.” Nature Ecology & Evolution 4: 686–687.
• Ullrich, Paul A., et al. 2018. “California’s Drought of the Future: A Midcentury Recreation of the Exceptional Conditions of 2012–2017.” Earth’s Future 6: 1568–1587.
• Zemp, Michael, et al. 2019. “Global Glacier Mass Changes and Their Contributions to Sea-Level Rise from 1961 to 2016.” Nature 568: 382–386.
• Zumsteg, A., et al. 2012. “Bacterial, Archaeal, and Fungal Succession in the Forefield of a Receding Glacier.” Microbial Ecology 63: 552–564.