Ecology technical companion

Homestake: wildlife, flora and the connected wetland mosaic

Updated October 15, 2026. Local observations, ecological studies and terrain models show how water and connected habitat support Homestake Valley. The observation counts are from September 23; the terrain comparison is from the September 24 lidar analysis.

A valley held together by water

Homestake’s fens and connected wetlands form a mosaic of groundwater-supported peatlands, wet meadows, willow, beaver ponds and streams beside the forest. These places offer different kinds of water and cover through the year. A shallow breeding pool, a saturated peat soil and a cold flowing creek support functions that a deep managed reservoir changes or replaces where it inundates them.

Across 37 north-central Colorado stream reaches, Caskey and colleagues found smaller channels and a shift toward drier-site vegetation at diverted reaches. Water removal can therefore change the land along a creek as well as the water within it. The study used 16 paired and five unpaired reaches, with 100 vegetation sampling points per reach. HWF:S31, Methods, Table IV and pp.592-594

Peatlands: recovery takes more than water

Fens form where sustained groundwater keeps peat wet. CNHP describes shallow saturation into July and August as important to peat accumulation. Hillslope springs, alluvial groundwater and creek exchange contribute in different proportions across the landscape. A fen beside a creek can receive its principal supply from the hillside. CNHP fen account, Environment and Dynamic Processes

At four restored Colorado fens, Schimelpfenig and colleagues found that ditch blocking raised summer water tables. At the site rewetted 20 years earlier, disturbed surface peat still differed from reference peat. Bringing the water back and recovering the soil that stores it are different stages of recovery. The experiment concerns restoration of drained peat, rather than recovery after reservoir inundation. HWF:V3-07, Table 1 and Results

Millar and colleagues measured eight plant communities in four Colorado and Wyoming mountain fens. Growing-season net ecosystem production increased with water-table position and was greater at colder, higher-elevation sites. Sustained saturation influences both habitat and peat carbon function. HWF:V4-04, Methods and Results

Confirmed groundwater-fed fens

CNHP’s September 12, 2011 field surveys confirmed two well-established, groundwater-fed fens near Gold Park in Homestake Valley. Researchers measured 92 cm of peat at site 6982 and 88 cm at site 6969, recorded saturated soils and established wetland plant communities, and found no disturbance or noxious weeds at either site. Together, these observations provide affirmative evidence of healthy, functioning fen ecosystems at the surveyed sites in 2011. HWF:FEN-2011-S01, Table 1 and printed pp.92-99

Protecting these established ecosystems means protecting the groundwater supplies and ecological conditions that sustain them. The field accounts describe groundwater arriving from adjacent slopes and flowing through the peatland system.

Inventory classes and reservoir exposure

Homestake’s fens are established features of the valley’s ecosystem. Wilderness Workshop and Eagle River Coalition describe them directly in their accounts of field work with ecologists. Wilderness Workshop, Homestake community science · Eagle River Coalition, July 23, 2026

The inventory confidence labels below describe particular mapped areas. The 2025 White River National Forest inventory distinguishes field-confirmed fen (>40 cm peat), field-confirmed peat-accumulating wetland (<40 cm), likely fen, possible fen and low-confidence fen. Appendix A reports the following confirmed/likely totals. HWF:S03, Table 1 and Appendix A

Appendix A groups confidence classes 7, 6 and 5: Confirmed Fen, Confirmed Peat-Accumulating Wetland and Likely Fen. The table preserves that combined inventory grouping.

CNHP’s Fens2025V2 inventory identifies 15 Confirmed Fen polygons within the Homestake Creek watershed, along with 160 Potential Fen polygons. The interactive valley map shows confirmed fens as their own class and displays the broader map’s Confirmed Peat-Accumulating Wetland class separately. These are inventory polygons; the two 2011 Gold Park survey sites lie within one confirmed polygon. The complete watershed comparison uses the Homestake outlet boundary from USGS NLDI and agency geometry retrieved October 10, 2026. HWF:FEN-MAP-V2, map source and classification

These are watershed inventory totals. The reservoir overlay is a separate calculation using the retained September 18 inventory and our modeled full-pool A-D water surfaces for the potential dam sites. It finds about 81.4 acres of mapped wetland and Potential Fen overlap, with overlapping areas counted once, shared by all four models, excluding mapped water categories. The wider wetland system connected to Homestake Creek contains 362.1 mapped acres in 36 linked complexes across our Whitney-to-Eagle assessment, reaching the Eagle River confluence near Red Cliff. Additional water removal would ultimately impact all connected downstream wetland ecosystems. Connected wetland acreage and methods Potential Fen remains a source mapping class. HWF:MODEL-C001, overlap calculation

The fen layer differs between inventory versions. None of the current Fens2025V2 fen polygons intersects the four stored modeled full-pool boundaries. The published potential-fen overlap uses the September 18, 2026 inventory, shown as a separately labeled historical layer; mapped wetlands also overlap the modeled pools. This is a difference in mapped fen coverage, rather than a field finding that wetlands or groundwater effects are absent. Current and historical map-source comparison

In plain terms, we placed each modeled water surface over the wetland and Potential Fen maps, counted the ground covered in every alternative, and counted overlapping map features only once. The technical area rule is area[(W ∪ F) ∩ A ∩ B ∩ C ∩ D] in the projected EPSG:32613 coordinate system, converted from square metres to acres; W includes only non-water wetland inventory categories, and F is Potential Fen. For a single option, replace A ∩ B ∩ C ∩ D with that option's water surface. The plain-language map method, formulas and source records explain the calculation; the underlying data preserve input hashes, inventory dates, class counts and per-option areas.

A September 23, 2026 comparison of the same downstream query envelope returned 18 Potential Fen features from the older CNHP service and six from Fens 2025. Five new features overlap older polygons, one does not, and none is geometrically identical. CNHP’s 2025 report explains that the older mapping rated existing NWI wetland polygons, which could combine fen and other wetland habitat; the update redraws boundaries around fen features. The reservoir-overlap calculations preserve the dated inventory’s original classifications and geometry; the interactive viewer uses the newer Fens2025V2 polygons. CNHP 2025, printed pp.7-9 Geometry comparison and retained query results

Water, carbon and refuges in a changing climate

Homestake's fens, wet meadows, willow margins, ponds and streams support several functions together. The studies below explain those functions at their named locations. Their significance for Homestake is the value of retaining the water, soils and connections that make them possible.

Cool water and dry-season resilience

Groundwater exchange can supply cool patches where aquatic organisms escape summer heat. An eight-year Oregon study found that natural and artificial beaver dams moderated daily temperature extremes and created local cool-water refuges. Shading by streamside plants also limits solar heating. Protecting a varied network of flowing water, groundwater connections and cover preserves thermal choices as streams warm. Weber et al., 2017 · CNHP, wetland functions

Wetland water is stored in saturated soils, shallow groundwater, ponds and floodplain sediments. In Rocky Mountain National Park, beaver dams moderated summer water-table decline across adjoining land. A Canadian Rocky Mountain fen shifted between transmitting, storing and releasing water through a dry season. These functions can sustain wet habitat after peak runoff. Many fens are groundwater-discharge environments; the direction and timing of exchange determine how water reaches the creek. Retained ecological water and extra downstream discharge are separate water-budget outcomes. Westbrook et al., 2006 · Streich and Westbrook, 2020 · Nash et al., 2018

Floods, sediment and water quality

Vegetation, pond storage and multiple flow paths can slow runoff and delay flood peaks. Wetland roots stabilize soil and banks, while ponds retain sediment, including material arriving after fire. Research in Colorado and Wyoming documented post-fire sediment accumulation in beaver ponds. The response depends on available storage and the event. Westbrook et al., 2020 · Dunn et al., 2024

Living soils also process nutrients. At Colorado's East River, field measurements and modeling linked a beaver dam to stronger subsurface exchange and nitrate removal. Wetlands can retain particle-bound phosphorus, while oxygen and water chemistry influence whether particular metals are held or released. In North Saint Vrain Creek, connected beaver habitat retained water, carbon and nutrients during high flows and released them as flows declined. Water quality includes the timing and chemical form of what travels downstream. Dewey et al., 2022 · CNHP · Briggs et al., 2019 · Wegener et al., 2017

Keep accumulated carbon in functioning soils

Saturation limits oxygen entering peat and helps preserve accumulated organic carbon. Drainage changes both carbon and the physical soil system: four Colorado fen restoration sites showed altered peat properties, with disturbed surface soil still different from reference peat at the site rewetted 20 years earlier. Carbon also accumulates in floodplain soils outside mapped fen boundaries. Schimelpfenig et al., 2014 · Sutfin and Wohl, 2017

Protecting Homestake’s fens preserves established carbon stores and the hydrology that sustains them. Growing-season carbon uptake varies with water tables, vegetation and temperature. Millar et al., 2017

Refuges through heat, drought and fire

Across the Cameron Peak, East Troublesome and Mullen megafires, beaver-dammed river corridors had lower burn severity. Wet patches can retain living structure within a burned landscape, complementing their sediment-retention role during recovery. Warming and earlier snowmelt increase the importance of maintaining the water that supports such habitat. A Colorado fen study found winter drought altered summer vegetation and evaporation even with near-normal summer precipitation, connecting winter water supply to summer conditions. Fairfax et al., 2024 · Blanken, 2014

A food web that crosses the water's edge

In Colorado's Kimball Creek, a greater share of beaver-pool habitat was associated with more aquatic insect biomass emerging per length of stream. Adult insects move that food resource onto land. Different pond and flowing-water communities contribute to a varied habitat mosaic. Water also shapes amphibian breeding opportunity, winter fish refuge, willow structure used by birds, and forage and cover for mammals. Dewey et al., 2025 · Hossack et al., 2015 · Lindstrom and Hubert, 2004 · Baril et al., 2011

Fen chemistry and saturation support distinct plant and moss communities. A San Juan Mountains survey recorded 20 fen communities across 182 sites. A Colorado experiment with streamside bluebells showed that water availability changed floral traits, seed set and pollinator visitation. Peat itself preserves environmental history: a Colorado fen record reconstructs hydroclimate and ecosystem change since the last glacial maximum. Intact wetlands therefore hold biodiversity, living reference systems and natural archives alongside their value for community science and learning. Chimner et al., 2010 · Gallagher and Campbell, 2017 · Todd et al., 2025

The conservation case is cumulative: changing groundwater or seasonal flow can affect peat, plants, beaver habitat, food supplies and refuges together. Flooding the valley floor also replaces a mosaic of shallow and flowing habitats with managed open water. Protecting Homestake means keeping those functioning connections intact.

Seasonal water and keystone species

Runoff connects channels and wet ground before summer water falls. Persistent pools and flowing reaches become refuges; severe drying can remove them. A keystone species has an unusually large influence on how its ecosystem functions. Beavers reinforce parts of this network by holding and spreading water. Groundwater-fed fens retain their distinct water sources and peat-forming processes.

Wet and dry years change habitat opportunities

The federal Beaver Restoration Guidebook describes off-channel habitat along larger rivers and seasonal main-channel dams at low flow. HWF:BEAVER-GUIDE, 2023 version 2.02, printed p.3 / PDF p.13

In North Saint Vrain Creek, Colorado, water-level and temperature records indicated a side-channel surface connection to the main creek in June and July 2015, while an off-channel beaver pond maintained relatively stable levels through October. On the East River, Colorado, a main-channel dam was built during the exceptionally low-water summer of 2018. These findings establish changing habitat opportunities and dam placement. As seasonal flows fall, persistent ponds and connected waterways become increasingly important habitat. HWF:BEAVER-SAINTVRAIN, Wegener et al. 2017, sections 2.1 and 3.1, Figure 5 · HWF:BEAVER-EASTRIVER, Dewey et al. 2022, seasonal hydrology and figures

Protecting Homestake’s ecosystem means sustaining its water, vegetation and connected habitat. Keystone species such as beavers help maintain the ponds, wet meadows, willow and channels that support other wildlife. Reduced water and destruction of connected wetland vegetation can constrain shelter, forage, movement and dam maintenance. Reduced beaver activity can then weaken water retention and habitat for other wildlife. This is a component-supported Homestake risk pathway. CPW beaver account · HWF:BEAVER-HABITAT, Tesky 1993, Forest Service habitat and cover

Beaver dams can hold water beyond the pond edge. Monitoring two dams along 1.5 kilometres of stream in Rocky Mountain National Park found slower seasonal groundwater decline across areas of 9 and 12 hectares within a 58-hectare study area. These measured areas describe those two dams during 2002-2005. HWF:S12, Westbrook et al. 2006

Fens are groundwater-fed peat wetlands; wet meadows and beaver ponds have distinct water regimes and histories. In one Rocky Mountain fen, four years with intact dams and two after flood breaches showed that dams connected the peatland to the stream, raised and stabilized water tables and increased storage. The reported 150 m radius of influence belongs to that study. For Homestake, the relevant connection is the actual creek, groundwater, pond and vegetation network sustaining each wetland. HWF:BEAVER-FEN, Karran et al. 2018, abstract

Reinforcing loss and the wider habitat network

Where less water or cleared vegetation limits beaver habitat, reduced maintenance and dam deterioration can further reduce lateral inundation and wetland water levels. Stressed willow then supplies less forage and construction material. Groundwater, tributary inputs, floods, browsing and climate shape the response. A 2025 Rocky Mountain National Park study examined interacting willow decline, beaver loss, groundwater, flow and climate. The study shows how willow decline, beaver loss and changing water supplies can reinforce one another. HWF:V3-06, Cooper et al. 2025

The local record and the area comparisons

Six beaver observations in the September 23, 2026 BioBlitz snapshot span June 2024 to July 2026, with four Research Grade and two Needs ID. The records may include the same animals. CSU’s 2005 Homestake field account describes intricate stream-wetland connections and Geyer’s, mountain and diamond-leaf willows. HWF:BEAVER-OBSERVATIONS, dated records · HWF:BEAVER-LOCAL-FIELD, Appendix A.1, printed A-39 / PDF p.43

Historical inventory records from 1983 imagery identify beaver-modified wetlands along lower Homestake Creek. The whole-complex inventory follows mapped wetland and pond connections: 36 linked complexes contain 362.1 unique acres of mapped wetland and potential-fen habitat across the Whitney-to-Eagle assessment. This includes one 49.2-acre complex crossing the upstream assessment endpoint. The lower reach from Blodgett to the Eagle River confluence contains 137.0 of those acres. These areas describe mapped habitat; drying varies with flow, pond levels and groundwater. Creek-fed side channels, ponding and shallow groundwater connect habitat across the valley floor; lowering the creek can increase drainage even where other sources also contribute water. Whole-complex inventory and methods

Inundation or clearing of a pond-channel-willow complex removes established habitat where it occurs. Beyond the flood line, altered creek and groundwater supply can change pond persistence, forage access and wetland function. Those connected effects extend the ecological concern beyond the inundation boundary.

Amphibians need time in the water

CNHP's 2000 Eagle County survey documents boreal-toad breeding in upper Homestake tributaries. CPW's 2006-2007 status report adds historical pond drying and uncertain recruitment at an upper-drainage site. The record establishes breeding habitat in the valley's history, with present occupancy unresolved at those sites. HWF:V3-17, printed p.159 / PDF p.162 · HWF:V3-24, printed p.36

Eggs and larvae need water long enough to reach metamorphosis. An experiment with boreal chorus frog tadpoles from Colorado populations found limited ability to accelerate development under simulated early pond drying. A pond can appear suitable in spring and still dry before young frogs finish that transition. HWF:V4-03, laboratory experiment, abstract

CPW identifies beaver pond complexes, including shallow sun-warmed abandoned or blown-out ponds, as boreal-toad breeding habitat. The different depths, temperatures and connections within a pond network matter to its use. CPW boreal-toad account

Cold-water habitat and food for fish

The 2000 CNHP survey records a Colorado River cutthroat population in the upper drainage. CPW's 2023 and 2024 distribution schedules each list 1,000 approved CR1 fish for Homestake Creek water code 20622. CR1 denotes a Colorado River cutthroat hatchery strain; the schedules record approval, with completed stocking, survival and current fish lineage unresolved. 2023 schedule, PDF p.34 and legend · 2024 schedule, PDF p.35 and legend

Flow maintains connected pools and carries food. In the upper Shasta River, California, a wet-2019 study found late-season drifting invertebrate prey approaching zero below a diversion even where streambed biomass could be higher. Food on the bed and food reaching a fish can respond differently to water removal. The study included 13 sampling events and limited fish-pool replication. HWF:V3-ST11, Methods, Results and Discussion

Moose, forest cover and passage

A Colorado study used 1,672 summer GPS locations from 21 adult moose in and near Rocky Mountain National Park during June-August 2017 and 2018. Mean distance to mapped flowing or standing water was about 110 metres; 90% of locations were within about 285 metres. Water, willow forage and adjacent forest belong together in the summer habitat picture. HWF:V3-ST01, Methods and Table 1

Colorado lynx research connects snowshoe hare availability and horizontal forest cover with habitat use. A western Colorado highway-crossing study of 13 tracked lynx found fewer crossings than a random movement expectation for five animals, with fine-scale crossings associated with nearby vegetative cover. These studies concern cover and movement in their respective study landscapes. HWF:V4-07 · HWF:V4-06

The Save Homestake archive separately identifies a Canada lynx photographed near the Blodgett reservoir option on April 2, 2017. The campaign supplies the date and locality; the original photographer is not named in the supplied archive record. This archival photograph is distinct from present camera coverage and from the BioBlitz snapshot. Archive image and attribution

Lidar: the modeled route climbs onto the hillsides

We compared the easiest terrain route between the same two valley gateways in the present landscape and each modeled reservoir. We then counted how much nearby ground still offers a route within a modest extra-cost allowance. This asks whether there is room for alternative paths as well as whether one path remains open. The exact routing grid, cost equation, gate locations and source terrain are in the linked methods below.

The September 24 lidar addendum uses 2016 bare-earth terrain on a native 2 m routing grid, fixed gates and the existing modeled full-pool reservoir outlines. For the matched central down-valley route, cumulative ascent rises from 73.2 metres under current conditions to 191.6 metres in A-C and 191.3 metres in D. The reservoir scenarios replace easier valley-floor passage with routes that climb the sides.

Cumulative ascent sums uphill portions of a route. Overall terrain cost combines 3D distance and signed slope through a relative cost function. They describe different aspects of the same passage. All 200 tested paths remained connected. Costs are relative terrain units; species-specific energy use and passability were not measured.

The opportunity analysis counts between-gate cells that can participate in a route within a fixed cost budget above the current-condition optimum. The percentage loss depends on how much extra terrain cost is allowed.

These are reductions in computational route-opportunity area. They show fewer inexpensive alternatives, with wider cost budgets admitting more routes. The existing 10 m reservoir visualizer supplies the modeled full-pool boundaries; the lidar analysis supplies a separate, finer terrain-routing comparison. Maps, methods, 200 route records and 15 envelope summaries

Local occurrence and plant methods

We used three distinct records: the dated community BioBlitz, a screened search of historical and submitted GBIF occurrences, and a separate comparison of plant names with the Corps' regional wetland-plant list. An observation is one submitted record, a taxon concept is an identification category, and an exact plant-name match is a lookup against a fixed list. The selection rules and source records follow.

The live BioBlitz query was refreshed at 2026-09-23 14:40:29 UTC. The snapshot retains 618 Research Grade, 778 Needs ID and four Casual observations. The 1,400-record table includes observation IDs and dates, identification grades, taxon assignments and positional accuracy. The 229 reported terminal concepts include higher-rank identifications; 190 are species rank. 491 records lack a terminal report-taxon assignment and remain unassigned in this snapshot. S01: live project

The GBIF search used “Homestake,” United States and Colorado, then screened county/locality fields. Of 379 discoveries, 242 matched Eagle County plus the locality name; nine “Homestake Peak School” records were excluded as place-name false positives. The resulting 233 records are a discovery set, not a complete drainage inventory. The 129 Cornell records contain 53 provider taxon-name concepts, including a genus-level identification; one additional historical bird specimen occurs outside that Cornell subset. Dataset lineage identifies submitted eBird-derived occurrence records; original checklist effort is unavailable in the provider fields used here. Merlin detections are not counted as eBird submissions. V3-35: query · V3-36: Cornell EOD dataset · S15: Cornell Sound ID guidance

The screening arithmetic is explicit: 242 locality matches − 9 school-name false positives = 233 retained discoveries. The source table keeps the provider ID and locality fields for each screened occurrence discovery.

Selected dated records:

The occurrence tables retain provider IDs, record dates, taxonomic strings, basis, licenses, uncertainty and GBIF quality flags. Multiple repositories or dates do not guarantee independent animals or populations. Specimens were not reexamined. Taxonomic conflicts, including a goshawk accepted-name mismatch and genus-level hawk record, remain visible. The plant records include introduced species and species of unresolved local origin.

Flora and plant-community interpretation

We matched the plant names in the local observation and historical records against one fixed regional wetland-plant reference. A match says how often that named plant tends to occur in wetlands across the reference region. The list version, counting rules and exact-name decisions follow.

The plant-name comparison contains 197 taxon/concept rows, including all 163 BioBlitz plant concepts, additional historical specimens and the historical cottongrass report. Across the combined table, 125 exact names match the Corps regional list; 72 remain unmatched. The 163 BioBlitz concepts alone yield 106 exact matches, including 44 OBL/FACW concepts.

For the full crosswalk, the count rule is 197 rows − 125 exact matches = 72 unmatched rows. The BioBlitz subset is reported separately; its 106 matches are part of the 125, not an additional set to add. The source table retains each original name, reference-list result and page locator so the classifications can be checked row by row.

The fixed reference is final 2022 NWPL version 3.6, Western Mountains, Valleys and Coast, with 3,407 parsed entries and per-name PDF page locators. The ratings describe where each plant tends to grow: OBL means obligate wetland, found almost exclusively in wetlands; FACW means facultative wetland, usually in wetlands; FAC means facultative, found in both wetland and non-wetland habitats; FACU means facultative upland, usually outside wetlands; and UPL means upland, found almost entirely outside wetlands. They do not delineate a wetland, establish peat depth or distinguish a fen from a marsh. USACE, 2022 WMVC list, cover/background and individual rows

No rating was assigned merely because a similar name exists. Caltha chionophila differs from the list’s C. leptosepala, Trollius albiflorus from T. laxus, and Vaccinium cespitosum from V. caespitosum; those pairs are not treated as exact matches. Bryophytes and genus/higher-rank records remain unrated when no exact entry exists. Unmatched is not automatically UPL.

Current CNHP exact matches add useful priorities: Aulacomnium palustre is G5/S1S3, fully tracked; Sphagnum angustifolium is G5/S3S4, watchlisted; historical GBIF Eriophorum chamissonis is G5/S3, watchlisted. The 33-row plant-community comparison table includes Betula glandulosa / Sphagnum Shrub Fen (G2/S2) and willow/sedge associations with differing ranks. The table lists potential community associations for comparison with the Homestake plant record. CNHP current tracking list, exact named rows

The existing water setting

Recorded Homestake Tunnel exports averaged about 23,400 AF/year in WY2007-2024, while about 21,500 AF/year passed Gold Park near the campground. Export represents 52.09% of the combined recorded export and gauge flow. The five-collector record uses a different interval, WY2014-2024, and places 88.26% of capture in May-June. Its water enters later reservoir and tunnel accounts, so it overlaps with export.

Annual totals establish scale. Seasonal water availability shapes pond persistence, peat saturation, cold-water refuge and streamside vegetation. Homestake’s habitats depend on different water sources, seasonal flows and biological conditions. Water methods and matched records

Evidence scope

BioBlitz collection boundaries, historical locality records, HUC12 inventory watersheds, modeled pools and gauge catchments describe different areas. Individual record dates, identification grades, positional uncertainty and licenses remain in the data tables. Precise breeding, rare-species and camera positions are excluded from public outputs.

Regional studies explain mechanisms. The reservoir overlays estimate exposure to our modeled full-pool boundaries. The corridor analysis evaluates terrain routes. Together they connect a documented resource with a plausible effect while preserving the scale at which each finding was established.