Wildlife and habitat evidence index for the Beaver archipelago, 2026

What this adds

The existing wildlife index records species, management, survey, and protection evidence across the seven-island frame. The MNFI soil/vegetation/ wildlife crosswalk, however, contains the strongest habitat-linked prose: it describes deer browse in forests, beaver flooding in swales, Hine’s emerald dragonfly habitat in fens, fish spawning and marsh birds, crayfish burrows, driftwood habitat, and dune or shoreline wildlife signals. This index makes those occurrence excerpts searchable beside the broader species record while preserving their different evidentiary levels.

The primary index contains 186 rows:

The 2017 rows preserve the exact occurrence excerpt, element-occurrence ID, site, community type, report pages, source lines, and adjacent soils, vegetation, and stewardship excerpts. The matcher retains source wording such as “deer,” “beaver,” “crayfish,” “shorebirds,” and “marsh birds” without silently converting every broad term into a species identification.

Outer-island habitat and study layer

The mapped-dune records add 59 primary rows for North Fox, South Fox, and Hog: 20 named animal or bird habitat examples and 39 broader habitat signals. The examples include Fowler’s toad, eastern hognose snake, Lake Huron locust, spider wasps, wolf spiders, antlions, and other source-listed taxa. The signal rows preserve functions such as shorebird feeding and roosting, waterfowl use, neotropical migrant songbird movement, cavity-nester habitat, and detritus-based food webs. These records are useful because they connect the documented dune, beach, forest, bluff, wetland, and vegetation layers to wildlife function, but the EGLE reports explicitly treat many named animals as habitat examples rather than complete occurrence lists at representative points.

Whiskey contributes a different study bundle: eight dated distribution-table taxa and two management-context rows, plus the 2010 cormorant study’s repeated study taxa and five prey taxa. The study documents Double-crested Cormorant nesting, ground/tree nest structure, modeled chick survival, and prey such as round goby, alewife, brook stickleback, white sucker, and crayfish. Those rows preserve the study’s one-season and model-based limits; they do not establish a current colony trend, island-wide occupancy, or lake-wide prey biomass.

Specialist bird studies

The index carries two specialist study bundles as a separate layer rather than folding them into checklist or occurrence counts. The Powell–Cuthbert primary study contributes nine structured records for 1987–1990. It documents systematic searches on Beaver and High, repeated nest monitoring, raised cobble and sand, foredune and inland-dune structure, beach widths, vegetation, ephemeral pools, gull colonies, storms, predators, and human disturbance. This is the project’s strongest direct bird–dune–vegetation connection.

The High Island woodcock bundle contributes five records from a secondary crosswalk. It preserves historic associations with old fields, shrub and young-tree cover, second-growth hardwood and conifer forest, soil moisture, disturbance, and the catalog-backed Whitcomb 1974 report pointer. Because the original Whitcomb scan is not held, its methods and demographic values remain secondary evidence.

Habitat-linked patterns

The evidence connects animals and birds to ecological structure in several useful ways:

• Beaver’s boreal forests, mesic forests, and rich conifer swamps repeatedly preserve deer trails, browse, herbivory, winter-deer-yard language, and concerns about cedar or hemlock regeneration.
• Beaver-influenced swales and peatlands retain beaver sign, flooding, dams, animal trails, and fine-scale soil-moisture or soil-chemistry gradients.
• Garden and Hog coastal or northern fens preserve crayfish burrows and Hine’s emerald dragonfly habitat or survey recommendations, alongside marl, peat, groundwater, and alkaline wetland context.
• Garden’s Indian Harbor marsh directly links fish spawning, shorebirds, secretive marsh birds, American bittern, Virginia rail, and sora to a named Great Lakes marsh occurrence.
• High and Beaver dune or cobble-shore records preserve Lake Huron locust, driftwood/insect/herptile habitat, and shoreline vegetation context.

These are occurrence-level relationships supported by source excerpts. They are not predictive habitat models or evidence that every similar soil or land-cover pixel supports the named animal.

Seven-island context

The index carries selected island-scale context so the research graph can be queried without implying fine-scale overlap. The 2025 NLCD frame reports 73.949 km² of forest and 55.334 km² of wetlands on Beaver, 8.477 km² of forest and 4.406 km² of wetlands on High, 10.540 km² of forest on South Fox, and smaller but distinct forest/wetland frames on Garden, Hog, Whiskey, and North Fox. The held SSURGO, NWI, circa-1800, Quaternary, aquifer, and SGMC summaries remain attached by island name.

The same context makes coverage gaps visible. The 2017 MNFI occurrence layer has no records for Whiskey, North Fox, or South Fox; the held GBIF crosswalk has zero retained observations for Whiskey and North Fox; and the public DNR forest-inventory layer is not equivalent coverage across all islands. These are source-coverage results, not evidence of ecological absence. The primary index now retains separate outer-island habitat and study rows even where the 2017 MNFI occurrence layer has no record.

The index also retains the existing 61-record archipelago wildlife index as a linked layer. That layer includes MDNR distribution-table marks, MNFI species table entries, archival survey summaries, Piping Plover protection and monitoring records, ruffed-grouse introductions, woodcock research, and other historical or management evidence. A table mark or management action remains distinct from a direct field observation.

Evidence limits

• The MNFI animals_and_birds facet is a parser-selected source field and can contain vegetation or habitat sentences alongside animal signals; the exact excerpt is retained for review.
• List entries are source-defined and list-scoped. They do not establish precise locality, abundance, current presence, or absence.
• Island-level soil, geology, NLCD, MNFI, and GBIF context is not a polygon assignment for any animal, bird, or occurrence excerpt.
• Public GBIF observation totals are coverage counts, not effort-corrected abundance or richness estimates; coordinate precision and observer effort vary.
• Zero matches are source-coverage results, not biological absence.
• Outer-island habitat examples and functions are source-preserved, not predictive habitat models; representative points are not occurrence boundaries and the mapped-dune reports are not complete island inventories.
• Whiskey distribution-table, cormorant, prey, and management rows preserve their dated source methods and must not be interpreted as current occupancy, population, or lake-wide biomass estimates.

Added 3 August 2026. Recent changes