Beaver Island ArchiveA primary-source archive & encyclopedia45°40′N 85°33′W · Lake Michigan
Colour scheme
Search

Dataset

GBIF wildlife observations and island ecology crosswalk, 2026

Also GBIF habitat context crosswalk · public wildlife and island ecology index

Draft — sources verified, prose evolving

TypeDataset
Dates1904 – 2026
WithinBeaver Island archipelago
ErasModern era
On this page
Dataset
Dates
1904 – 2026
Eras
Modern era
Concepts
wildlifebirdsanimalsforestsvegetationwetlandssoilsgeologyconservation
Within
Beaver Island archipelago

What this adds

The existing GBIF snapshot contained 5,097 public observations, but its interpretation stopped at raw island totals. This crosswalk turns it into an aggregate comparison layer: 36 selected taxa × 7 islands = 252 cells. Each cell records public-observation counts, years, coordinate-precision classes, historical wildlife-evidence counts, and a conservative island-level ecology context.

Island GBIF observations Selected taxa observed 2025 NLCD forest km² 2025 NLCD wetlands km² MNFI 2017 occurrences Public DNR stands
Beaver Island 4,908 25 73.9494 55.3338 17 424
Garden Island 65 8 0.0909 16.9335 12 0
High Island 112 13 8.4771 4.4064 8 0
Hog Island 6 5 5.4198 2.6847 6 0
South Fox Island 6 4 10.5399 0.0450 0 0
Whiskey Island 0 0 0.2520 0.0405 0 0
North Fox Island 0 0 2.8593 0.0810 0 0

The table shows observation and inventory coverage together without equating them. Beaver’s 4,908 observations reflect public attention and accessibility; they do not make it biologically richer by that ratio. The zeroes for Whiskey and North Fox are retained GBIF results for the selected taxa, not evidence of species absence. The zero DNR counts on six islands are likewise a public- inventory coverage result documented in the forest layer.

Wildlife and bird connections

The largest public-observation totals are Bald eagle (967), Common loon (872), Common tern (606), Caspian tern (572), Wild turkey (484), Merlin (319), and Ruffed grouse (304). These are modern public-observation counts, not population estimates.

Piping plover is a particularly useful connection. The GBIF snapshot retains 18 observations—11 on Beaver and 7 on High—while the historical evidence layer contains records or survey context for Beaver, Garden, High, Hog, South Fox, and Whiskey. The mismatch is informative: a modern public-observation layer and a survey/management archive answer different questions and should not be collapsed into one distribution map.

American woodcock has 106 retained observations across Beaver, Garden, High, Hog, and South Fox. That creates a useful comparison with the High Island woodcock research trail, but the crosswalk does not turn those points into woodcock habitat or occupancy claims. Common loon, terns, grouse, sparrows, meadowlarks, bluebirds, and raptors similarly gain a searchable modern layer alongside the MNFI, MDNR, and newspaper records.

Several historically documented taxa have no retained GBIF observation in the selected snapshot, including Northern goshawk, Lake Huron locust, Hine’s emerald dragonfly, Little brown bat, Coldwater pondsnail, Campeloma spire snail, Great Lakes physa, Giant northern pea clam, and Lake floater. These are priority gaps for targeted survey or source retrieval, not negative findings.

Connecting animals to soils, forests, and vegetation

The ecological context is deliberately island-level. For each island the crosswalk preserves:

  • 2021 and 2025 NLCD forest, wetland, developed, barren, and open-water areas, plus the number of changed pixels between the two years;
  • MNFI occurrence counts and natural-community types;
  • leading SSURGO map units and their intersected areas;
  • NWI, hydric-soil, Quaternary-geology, and circa-1800 cover categories; and
  • public MDNR stand counts, areas, and cover groups where the held service returned stands.

This allows questions such as whether the islands with public observations also have documented forest, wetland, dune, or historic-cover context, while preserving the correct direction of inference: an island’s forest or soil context does not prove that a particular observation occurred in that class.

Method and limits

The build groups the retained GBIF records by common name and island, retains date ranges, unique years, source datasets, institutions, and coordinate- precision classes, then compares each cell with matching records from the curated 2017–2024 wildlife evidence index. It attaches the existing island context summaries by island name. Raw coordinates are not copied into this aggregate output.

GBIF is a public-observation source, not a standardized survey. Many records have unknown coordinate uncertainty, and observer effort is uneven. The selected taxa are inherited from prior archive work rather than being a full flora or fauna inventory. Environmental layers are broad or survey-bounded, and no point-level habitat, soil, vegetation, ownership, or legal-wetland assignment is made.

See also

Sources

  1. Project-generated GBIF wildlife observations and island ecology crosswalk (2026).
    Contains 252 island-by-taxon cells for 36 selected taxa, aggregate public-observation dates and precision classes, matching historical evidence counts, and island-level NLCD, MNFI, SSURGO, NWI/geology/circa-1800, and public forest-inventory context. Raw observation coordinates are intentionally omitted. No public copy of this file is known; the archive holds one and can provide it on request.
    research · wildlife · indices · gbif_archipelago_ecology_crosswalk
  2. Project build script for the GBIF ecology crosswalk.
    Reproduces the aggregate grouping and context joins without assigning an observation to a habitat polygon. No public copy of this file is known; the archive holds one and can provide it on request.
    scripts · build_gbif_archipelago_ecology_crosswalk
  3. Supplies 5,097 retained public observations within the seven held island polygons, with original dates, taxonomic fields, publishers, and coordinate-precision metadata preserved in the source index.
    research · wildlife · indices · gbif_archipelago_observations
  4. Project-generated Annual NLCD island ecology crosswalk (2021 and 2025).
    Supplies island-level forest, wetland, developed, barren, open-water, and transition context; NLCD classes are broad land cover, not species habitat or tree-species maps. No public copy of this file is known; the archive holds one and can provide it on request.
    research · land_cover · crosswalks · annual_nlcd_island_ecology_crosswalk_2021_2025_2026-08-02
  5. Project-generated MNFI occurrence and soil/vegetation/wildlife crosswalks.
    Supplies island-level natural-community occurrence and environmental-layer context while preserving the occurrence-level and island-level distinction. No public copy of this file is known; the archive holds one and can provide it on request.
    research · island_farming · crosswalks · mnfi_occurrence_environmental_crosswalk
  6. Held SSURGO, NWI, Quaternary geology, circa-1800 cover, and MDNR public forest-inventory summaries.
    Provides broad physical and forest context for comparison; these layers are not joined to individual public observations. No public copy of this file is known; the archive holds one and can provide it on request.
    soils · ssurgo_island_soil_ecology_crosswalk

Added 3 August 2026. Recent changes

Cite this article

Beaver Island Archive, “GBIF wildlife observations and island ecology crosswalk, 2026”, https://beaverislandarchive.org/documents/gbif-archipelago-ecology-crosswalk-2026/, accessed [date], citing Project-generated GBIF wildlife observations and island ecology crosswalk (2026). (and 5 further sources listed on the page).

Last revised 2026-08-03. Fill in [date] with the date you read it — this page is static and cannot know that.

BibTeX for “GBIF wildlife observations and island ecology crosswalk, 2026”RIS for “GBIF wildlife observations and island ecology crosswalk, 2026”