Plant and vegetation evidence index for the Beaver archipelago, 2026

What this adds

This index makes plant evidence searchable across several different source scales while keeping their meanings separate. The 2024 MNFI guide and appendices provide source-defined plant lists; the 2017 MNFI report provides plant names inside named natural-community occurrence descriptions; the EGLE/ MNFI mapped-dune indexes provide structured vegetation and tree-type entries for named outer-island occurrences; the MDNR MIFI layer provides plant names in modern public-forest stand comments; and the Whiskey management baseline provides three broad cover-type labels.

The generated output contains 2,196 evidence rows across 240 normalized plant or vegetation labels. A row preserves the matched common or scientific-name alias, source excerpt or structured entry, source class, island label, year when supplied, and source locator. Repeated mentions are retained because they point back to different evidence contexts; they are not counts of plants, sites, acres, abundance, or current presence.

Island scope and source asymmetry

The synchronized archipelago matrix exposes how unevenly these plant sources cover the named-island frame:

These are exact-island assignments within this index. Beaver’s rows combine MNFI occurrence descriptions, MIFI public-stand comments, and the three source-defined 2024 guide lists; Garden and High are represented by MNFI occurrence-description matches; Hog combines those matches with mapped-dune vegetation entries; North Fox and South Fox are represented by structured mapped-dune occurrence entries; and Whiskey has only three broad management cover-type labels. The outer-island rows are therefore not equivalent to the Beaver occurrence, list, and public-stand coverage. The guide’s Beaver-wide lists are not copied onto outer islands merely because the list is regional or archipelago-related.

Connecting plants to habitat

The MNFI occurrence matches are especially useful because the source excerpts retain their ecological facets. A plant name can therefore be read alongside the same occurrence’s soils and hydrology, vegetation structure, animal and bird observations, and disturbance or stewardship notes. The crosswalk covers bogs, fens, marshes, dunes, beaches, boreal forest, northern forest, and conifer swamps across four islands. The mapped-dune entries extend that structured vegetation frame to North Fox and South Fox, including foredune grasses, beach and dune herbs, shrubs, tree layers, and an interdunal wetland; Hog adds forest and cobble-shore layers. This supports questions such as how tamarack, cedar, sedges, rushes, pitcher plants, or dune plants are described within particular substrate and water regimes without asserting that the crosswalk is a complete flora for any site.

The MIFI subset adds a different kind of connectivity. Its exact matches are sentences from Beaver Island public-forest stand comments, linked to stand and cover-type attributes. They can be compared with the separate tree/woody vegetation index, but the comments remain narrative observations and the layer does not cover private land or provide equivalent coverage for the other islands.

Whiskey remains a deliberate low-resolution case. Its three new rows are the 2017 plan’s “mixed conifer,” “balsam fir,” and “aspen” cover-type labels, not species-level survey records. The same source explicitly reports a natural- heritage survey gap, so the plant index does not manufacture a species list from the cormorant study’s nesting vegetation context.

Community-level vegetation and habitat frame

The companion MNFI natural-community connectivity matrix provides a more structured way to read the plant matches. It joins the 43 named 2017 occurrences to source-preserved plant and tree taxa, soil/hydrology terms, and wildlife signals. The following selected community types show the range of connections available in the held report:

The counts are occurrence-level summaries, not species richness or habitat quality rankings. “Plant taxa” and “tree taxa” count normalized labels present in the occurrence records; wildlife rows can include broad source wording such as “deer” or “shorebirds.” The substrate and water columns describe terms in the same occurrence evidence, not a claim that every listed taxon occupies every mapped soil or geology unit. The matrix joins island-scale context by island name and does not create an occurrence-specific polygon overlay.

Native, rare, and invasive status

The 2024 list records preserve the source’s categories and reported status text. “Native,” “rare,” and “invasive” are not silently reconciled into a current legal or ecological status, and a list-scoped Beaver Island entry is not assigned a precise locality. The index retains this list layer beside the 2017 occurrence matches so that a researcher can distinguish a guide appendix entry from an occurrence-level description.

The regional 2019 invasive-action framework supplies planning context for the broader archipelago and Great Lakes islands: its priority scope is 93 invasive species across 13 taxa. That figure describes the framework’s regional scope, not current Beaver Archipelago presence. It should be used with the dated Beaver list, the 2017 MNFI threats and stewardship text, and later field or management records.

Evidence limits

• Exact-name matching is a retrieval aid. It can miss synonyms, spelling variants, or names absent from the 2024 lexicon, and it does not prove that a matched name refers to a separate individual.
• Guide and appendix rows are source-defined lists. Their island scope does not assign every plant to a parcel, stand, community polygon, or current occurrence.
• MNFI matches remain bounded by named occurrence records and preserved source excerpts. They do not establish complete site flora, abundance, or absence.
• MIFI matches are limited to Beaver Island public-inventory comments and do not replace formal plant surveys or a complete stand-composition inventory.
• Mapped-dune entries are structured examples tied to named source records; representative points are not natural-community boundaries, and their vegetation lists are not complete island floras.
• Whiskey’s cover-type entries are broad management-plan labels and preserve a known survey gap rather than filling it with inference.
• Zero matches are source-coverage results, not evidence of absence.

Added 3 August 2026. Recent changes