MNFI Natural Community Surveys of Beaver Island Archipelago, 2017

Summary

Joshua G. Cohen’s Natural Community Surveys of Beaver Island Archipelago was prepared for the Little Traverse Bay Bands of Odawa Indians Natural Resources Department and dated December 31, 2017. The held copy is a 168-page PDF and the official MNFI report record identifies it as Report 2017-11. The survey work covered the 2015, 2016, and 2017 field seasons and was designed to give resource managers and planners standardized baseline information for biodiversity stewardship, protection, management, restoration, and monitoring.

The report is now locally held and searchable. Its methods combine field surveys, aerial-photo interpretation, historical imagery, GIS analysis, GPS ground-truthing, soil and hydrology observations, tree aging and DBH measurements, plant inventories, threat assessment, and ecological-boundary mapping. The report’s 43 site summaries are therefore a much stronger basis for this archive’s forest, wetland, dune, shoreline, species, and stewardship connections than the earlier web-indexed totals alone.

All 43 numbered site summaries now have at least one page visually checked against the held PDF and preserved as evidence images. The set covers all 16 natural-community types and every EO in the normalized occurrence index. This is a complete one-page-per-occurrence audit, not an image audit of every continuation page in multi-page summaries.

Island-by-island survey frame

The report summarizes the survey as follows:

These are element-occurrence and community-survey counts, not acres, species counts, or parcel boundaries. “New” means newly documented in this survey sequence; it does not mean newly formed habitat.

The detailed survey-results section says that 16 natural-community types were visited: bog, boreal forest, coastal fen, dry-mesic northern forest, Great Lakes barrens, Great Lakes marsh, interdunal wetland, limestone cobble shore, mesic northern forest, northern fen, open dunes, poor conifer swamp, poor fen, rich conifer swamp, sand and gravel beach, and wooded dune and swale complex. The report distinguishes the 43 occurrences visited from the larger tables of previously known occurrences, some of which were not revisited during this project.

The project now also holds a normalized occurrence index. Each of its 43 records preserves the report’s site name, community type, island, EO ID, new-or- updated status, global and state ranks, element-occurrence rank, size in acres, printed report pages, and extraction line range. It also carries the report’s site description, threat assessment, and management recommendations as source-bounded text. This makes queries such as “all Beaver Island conifer swamps with deer-browse observations” or “all outer-island shoreline EOs with invasive threats” reproducible without pretending the report supplies modern GIS geometry.

The record-level facet lists make the ecological content easier to query without turning keyword extraction into a new survey. soils_hydrology captures the report’s pH, peat, sand, marl, substrate, saturation, water-depth, seepage, and related observations; vegetation_tree_types captures canopy, understory, ground-cover, tree, and floristic descriptions; animals_and_birds captures observations such as deer, beaver, dragonflies, marsh birds, fish, crayfish, and insects; and disturbance_and_stewardship captures logging, fire, windthrow, flooding, invasive species, erosion, threats, and recommendations. Each facet is a source-worded extract from the same site summary, not an inferred presence/absence inventory.

Page map and Fox Islands scope

The downloaded PDF was verified against the held source copy: the SHA-256 is 22197bd0afaf772367eb64d02fa80ab552fbea5dc7c5f5287b72750f4d388194. A targeted page index preserves both the printed report page and the PDF page number, because the report has front matter and the two numbering systems differ.

The most useful report-level pages are:

• Printed pages 1–6 (PDF pages 7–12): introduction, Tables 1–3, and maps. These establish the survey frame as Beaver, Garden, High, and Hog Islands.
• Printed pages 7–10 (PDF pages 13–16): field methods, including plant lists, canopy structure, tree DBH and aging, soils, hydrology, disturbance, GPS, and ecological-boundary work.
• Printed pages 12–153 (PDF pages 18–159): the 43 site summaries, organized by natural-community type.
• Printed pages 154–160 (PDF pages 160–166): stewardship prioritization, Beaver-versus-outer-island comparison, and discussion.

The visual evidence subset is:

The machine-readable page index retains the corresponding PDF page numbers and SHA-256 checksums; the occurrence crosswalk links all 43 images back to their EO IDs through its visual_evidence field. There is one checked page for every numbered occurrence and coverage of all 16 community types, but the selected page does not constitute a full continuation-page audit for multi-page summaries.

North and South Fox Islands do not appear as survey islands, occurrence-table islands, or mapped occurrence records in this report. That is a documented scope boundary, not a finding that Fox Islands lack natural communities. The Fox Islands therefore need a separate source search—such as current or historical MNFI records, DNR natural-community layers, or a dedicated survey—if the archipelago analysis is to include them on equal terms.

The current extraction finds soil/hydrology, vegetation/tree, and disturbance/stewardship material in all 43 records, and animal/bird material in 39. The four records without an animal/bird facet are not evidence of an empty fauna; they simply contain no matching observation in the extracted site summary text.

The index also preserves a report-internal discrepancy: the introduction says that seven new Beaver Island occurrences were documented, while the 17 numbered Beaver Island site summaries contain six explicit “New EO” labels. The Garden/High and Hog counts align with their explicit labels. Both the narrative totals and the site-level labels remain available; the project does not silently choose one as the corrected value.

Island-level environmental crosswalk

The project crosswalk joins the occurrence index to the held environmental layers at the named-island level. It is a useful comparative frame, but it is not a substitute for MNFI occurrence polygons:

These values are clipped to current OSM island polygons in equal-area EPSG:6933. They help compare the ecological setting around the four surveyed islands—for example, the larger hydric-soil footprint on Beaver Island versus Garden Island’s very high NWI-mapped wetland/deepwater footprint—but they do not assign any of those areas to a particular EO, landowner, or legal wetland boundary.

The Beaver Island table connects the ecological record directly to places already documented here: Fox Lake Bog, Greene’s Lake, French Bay, Little Sand Bay, Pointe La Par, Font Lake Old Growth, Lake Genesereth Old Growth, Martin’s Bluff, Southwest Old Growth, Cable Bay, Iron Ore Bay, Lookout Point, McCauley Point, McFadden Point, Sand Bay, Egg Lake, and Doty’s and Greene’s Lake swamps. The outer-island tables add Garden Island, High Island, and Hog Island sites including Indian Harbor, Jensen Harbor, Monatou Bay, Taganing Marsh and Shore, Nezewabegon Barrens and Forest, Garden Island Harbor, and Hog Island’s coastal, marsh, fen, forest, and swamp systems.

The report also records named rare or status-sensitive organisms and habitats, including dwarf lake iris, Pitcher’s thistle, Lake Huron tansy, Lake Huron locust, Hine’s emerald dragonfly habitat, and American bittern, Virginia rail, and sora observations at Indian Harbor. These observations are site- and survey-bounded; they are not a current island-wide species inventory.

The 2024 Peaine recreation plan summarizes an earlier 2016 survey slice as ten previously known Beaver Island occurrences plus five new occurrences, fifteen occurrences total, and nine community types. That difference is retained rather than treated as an error: the 2017 report includes a later field season and reports a different time slice.

Research significance

This report is the missing bridge between the project’s modern forest and wetland themes and the later planning language about natural-community protection. It provides site-level element-occurrence rankings, classifications, delineations, ecological boundaries, vegetation structure and composition, landscape and abiotic context, threats, management needs, and restoration opportunities referenced by the 2024 recreation plan.

Several cross-theme findings are especially useful:

• Beaver Island’s interior forest systems are repeatedly described as affected by deer browse, with effects on understory composition, conifer regeneration, and in some cases the apparent absence of yew. The report identifies deer density and browse pressure as a leading stewardship issue on Beaver Island, unlike the outer islands where deer were absent or nearly absent.
• Shoreline and dune systems carry a different threat profile. Invasive plants were most common in Beaver Island open dunes, while foot traffic and off-road vehicles also damaged dune systems. On Garden, High, and Hog Islands, invasive species were concentrated in shoreline ecosystems.
• Hydrology is not background detail: Great Lakes water-level fluctuation, storm disturbance, beaver flooding, peat depth, marl, alkaline sands, and groundwater seepage organize the transitions among fens, marshes, swamps, dunes, and forests. At Hog Island, high-water conditions were already shifting portions of a coastal fen toward Great Lakes marsh.
• The stewardship matrix is explicit rather than impressionistic. It combines ecological integrity, rarity, and threat severity. On Beaver Island the top scores are open-dune occurrences and two mesic northern forest occurrences; on the outer islands, Indian Harbor and Hog Island Great Lakes marsh score highest, followed by globally rare coastal fens. The report’s final synthesis identifies invasive control in open dunes, deer-browse control in forests, and invasive monitoring/control in outer-island coastal ecosystems as the leading priorities.

It also gives the archipelago a comparable survey frame. Beaver, Garden, High, and Hog Islands should not be collapsed into one ecological unit: the report surveys different numbers of occurrences and community types on each island or island group. The same caution applies to North and South Fox Islands, which are part of the broader archipelago conservation context but are not represented by these reported occurrence totals.

Evidence limits

The report is locally held, but it remains a 2015–2017 baseline rather than a current condition assessment. It should not be used to infer ownership, public access, parcel boundaries, completed restoration, or the current presence or absence of a rare species. The PDF’s maps and tables are authoritative for the report’s occurrence framework; they have not yet been converted into project GIS layers.

Added 2 August 2026 · revised 3 August 2026. Recent changes