Seven-island ecology connectivity matrix, 2026

Summary

This matrix is the archipelago-wide comparison layer that the individual ecology indexes were missing. It places seven island records side by side and retains the source-specific measures for tree and woody-vegetation evidence, plant evidence, MNFI natural-community occurrences, animals and birds, aquatic fauna and fisheries, DNR forest stands, NLCD canopy, NLCD land cover, SSURGO soils, and physical geology.

The matrix does not pretend that these sources measure the same thing. A tree row is an evidence match, an MNFI record is a named occurrence, a DNR stand is an inventory polygon, a canopy value is a 30 m remote-sensing summary, and a soil or geology value is mapped regional context. The output makes those different units easier to compare while retaining their labels and limits.

Seven-island comparison

These values are navigation signals, not rankings. Beaver has much larger tree and plant counts partly because the DNR public-forest comments and MNFI occurrence descriptions are concentrated there. Garden, High, and Hog have substantial MNFI occurrence coverage. Fox now also has structured mapped-dune vegetation entries, while Whiskey has only broad cover-type labels; neither should be read as equivalent to a complete plant inventory or current DNR stand coverage.

What the connectivity makes visible

Forest structure and named trees

The tree index contributes exact-island GLO, MNFI, guide, and DNR-comment evidence, with the largest source concentration on Beaver. The current canopy layer extends quantitative structure to all seven islands, including North and South Fox. The DNR inventory adds stand and covertype detail only for the returned Beaver frame. A high canopy value therefore cannot be translated into a particular cedar, pine, spruce, maple, oak, or beech community without the source-specific tree and natural-community evidence.

Plants, soils, and natural communities

The MNFI occurrence matrix connects plants and tree taxa to named bogs, fens, marshes, dunes, boreal forests, northern forests, and conifer swamps on four islands. The outer-island mapped-dune indexes add structured vegetation and tree-type entries for Hog, North Fox, and South Fox, while Whiskey contributes only broad cover types. The SSURGO and physical-geology sections extend the comparison to all seven islands, but only at island context scale. The matrix keeps the useful question—where do source records describe similar trees, plants, substrates, and water regimes?—separate from the stronger and usually unsupported claim that a particular soil polygon overlaps a particular occurrence.

Historical GLO ecology

The historical GLO layer adds 43 exact-island rows from the 1845–1849 survey sequence to the matrix. These are not a second modern habitat survey: they are page-bounded surveyor descriptions that put soil, shore material, water, tree names, vegetation, fire, windfall, timber cutting, fishery use, and occasional bird language on a dated historical frame. Beaver contributes 21 rows; Garden 8; Hog 4; North Fox 4; High 3; Whiskey 2; and South Fox 1. The 1942 Shoe and Pismire records are kept in the same domain but remain outside the exact seven-island counts.

This makes historical change queryable without pretending that a survey route is a forest stand or a page description is an ecological inventory. The source records retain the handwritten or typed survey date, page image, township and range, source wording, and explicit limits. Group-scoped Garden/Hog language, small-island observations, and the 1942 form are not split or normalized into unsupported modern island assignments.

Invasive species and stewardship

The stewardship layer adds 84 exact-island source rows while preserving five different evidence classes. Beaver’s 51 rows combine the 31 source-defined 2024 invasive-plant list entries, seven GEI survey and planning records, twelve MNFI 2017 occurrence stewardship records, and one MNFI 2019 Beaver case-study record. Garden, High, and Hog retain MNFI occurrence stewardship evidence; North and South Fox retain current mapped-dune invasive or disturbance signals. Whiskey has no exact-island row in this source bundle. Seven additional MNFI 2019 action-framework records remain regional or unassigned.

These rows connect invasive plants to forests, dunes, marshes, fens, shorelines, soil and hydrology concerns, rare-species safeguards, and colonial-waterbird planning. They must not be read as a current invasive distribution map: a source-defined list is not a point occurrence, a stewardship recommendation is not a treatment result, and the unavailable GEI appendices are not reconstructed from downstream accounts.

Wetlands and hydrology

The wetland/hydrology domain adds direct mapped context where the held sources support it. The NWI and EGLE hydric-soil summaries cover Beaver, Garden, High, and Hog in the held environmental crosswalk; Beaver has 318 NWI features over 47.824 km² and 207 EGLE hydric-soil features over 64.536 km², while the smaller islands have their own asymmetric values. The named-place crosswalk adds 18 exact-contained aquatic or wetland anchors within the seven-island frame, including 16 on Beaver and one each on Garden and High. Point features that are only nearest to an island are not assigned to it.

These are map-context joins, not wetland delineations for parcels or species. They are useful for reading fens, marshes, swamps, ponds, lakes, streams, and hydric soils beside the forest, plant, invasive, bird, and aquatic-fauna records, while preserving missing NWI/EGLE context for Whiskey and both Fox Islands as a source-coverage gap.

Animals and birds

The wildlife/bird layer preserves species, distribution, survey, management, and observation records across the archipelago. It can be read alongside forest, wetland, canopy, soil, and geology context, but the matrix does not infer habitat suitability or occupancy from those layers. An island with zero MNFI occurrence records or zero DNR stands is not an island with zero animals, birds, trees, or natural communities.

Fox Islands and source asymmetry

North and South Fox are intentionally present as full rows rather than being lost inside a Beaver-centered summary. Their rows show the current canopy and soil/geology frames, sparse tree evidence, source-declared wildlife/bird records, and the absence of returned public DNR stands and MNFI occurrence records in the held indexes. Those asymmetries identify research priorities; they do not establish ecological absence. The same rule applies to Whiskey.

Aquatic fauna and fisheries

The aquatic layer adds 23 exact-island rows to this seven-island frame. Beaver has source-defined pondsnail and lake-trout/whitefish-management records; Garden and Hog have wetland-fauna signals tied to coastal fens, marl pools, crayfish burrows, and Hine’s emerald dragonfly; High has both aquatic-fauna and GLO fish-shanty evidence; and Whiskey and North Fox retain page-bounded GLO fishery-use observations. South Fox has no exact-island row in this particular index. The remaining 14 rows concern Ojibwa, Trout, Gull, and Gull Island area scopes and are intentionally not folded into the seven named islands.

The distinction matters: a crayfish burrow or pondsnail list entry is aquatic fauna evidence; a GLO fish shanty or fishing village is historical human-use evidence; and a federal egg-planning or fry-distribution table is agency activity or planning evidence. None is by itself a current fish inventory, catch series, stocking-site polygon, or population estimate.

Method and limits

The generator joins existing records by exact island names or consumes an existing island-level summary. It does not perform a new spatial overlay or assign a soil, geology, canopy, land-cover, or DNR polygon to an MNFI occurrence, wildlife observation, tree mention, or parcel. Group labels such as “Garden and Hog Islands” remain group-scoped in the tree index instead of being split into two records.

Evidence-row counts are not counts of individuals or abundance. Wildlife and bird records have source-specific survey and distribution meanings. Aquatic- fauna rows, GLO fishery-use observations, and federal fish-management records are distinct evidence classes; fishery activity is not an aquatic occurrence, and planning or stocking tables are not proof of a local release, collection station, abundance, or current occupancy. NLCD canopy and land-cover values are remote-sensing summaries. SSURGO and physical geology categories are regional map context. The DNR inventory is asymmetric and public-inventory based. The underlying source indexes, source excerpts, geometries, rasters, and provenance files should be consulted before making a site-level claim.

Added 3 August 2026. Recent changes