statewide analysis and surveys to develop an approach for ...approach for evaluating the landscape...
TRANSCRIPT
Statewide Analysis and Surveys to Develop an Approachfor Identifying Priority Conservation Areas in Michigan:
2008 Progress Report
Prepared by:Yu Man Lee, Edward H. Schools, Bradford S. Slaughter, Michael R. Penskar,
Peter J. Badra, Joelle L. Gehring, Amy L. Derosier, Michael A. Kost, and Patrick W. Brown
Michigan Natural Features InventoryP.O. Box 30444
Lansing, MI 48909-7944
For:Michigan Department of Natural Resources
Wildlife DivisionP. O. Box 30444
Lansing, Michigan 48909
February 12, 2009
Report Number 2009-02
Cover photo: Dry-mesic northern forest at Pigeon River Pines in Michigan, photo by Joshua G. Cohen, MNFI
Copyright 2009 MSU Board of Trustees
Michigan State University Extension is an affirmative-action, equal-opportunity organization.
Funding for this project was provided by the Michigan Department of Natural Resources, Wildlife Division.
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008, Page - i
Table of Contents
INTRODUCTION ............................................................................................................................................ 1
METHODS ....................................................................................................................................................... 2
GIS Analysis/Component..............................................................................................................................2 Terrestrial GIS Analysis/Component ................................................................................................................. 2 Aquatic GIS Analysis/Component ..................................................................................................................... 4
Field Component ......................................................................................................................................... 4 Terrestrial Field Component .............................................................................................................................. 4 Aquatic Field Component ................................................................................................................................ 17
Aerial Photo Interpretation Component ...................................................................................................... 18
Conservation Gazetteer .............................................................................................................................. 18
RESULTS AND DISCUSSION ...................................................................................................................... 21
GIS Analysis/Component ........................................................................................................................... 21
Field Component ....................................................................................................................................... 25 Terrestrial Field Component ............................................................................................................................ 25 Aquatic Field Component ................................................................................................................................ 26
Aerial Photo Interpretation Component ...................................................................................................... 26
Conservation Gazetteer .............................................................................................................................. 27
REFERENCES ................................................................................................................................................ 27
ACKNOWLEDGEMENTS ........................................................................................................................................... 28
APPENDICES .............................................................................................................................................. A-1
List of Tables
Table 1. Thresholds for patch scores for small patch communities for area scoring. ............................................... 3
Table 2. Thresholds for patch scores for large patch communities for area scoring. ............................................... 3
Table 3. Thresholds for patch scores for core area scoring. .................................................................................. 3
Table 4. Summary of variables utilized in GIS stream/river reach disturbance model developed by Wang et al. (inpress) to predict potential quality or condition of stream/river reaches based on predicted impact rankings. ..... 5
Table 5. Crosswalk showing relationships between GIS model land cover types/classes, natural community fieldclasses, and MNFI natural community classifications. ................................................................................... 10
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008, Page - ii
Table 6. Summary of field metrics utilized to assess the ecological quality or condition of the sample cells in uplandoak and oak-pine forest and upland mesic forestnatural community field classes which are associated withupland forest cover type patches. ................................................................................................................ 11
Table 7. Summary of criteria for field metrics for evaluating the ecological quality or condition of sample cells in theupland mesic forest natural community field class. ........................................................................................ 12
Table 8. Summary of criteria for field metrics for evaluating the ecological quality or condition of sample cells in theupland oak and oak-pine forest natural community field class. ...................................................................... 14
Table 9. Summary of botanical, zoological, and aquatic community field metrics used to assess the quality orcondition of the sample cells in cover type patches across a gradient of quality based on the GIS patch analysismodel. ........................................................................................................................................................ 16
Table 10. Summary of rank factors, key ecological attributes, metrics, and ranking criteria for assessing the ecologi-cal quality or condition of photo-interpreted polygons in general. .................................................................. 19
List of Figures
Figure 1. Map showing general location of study area in Michigan.......... .............................................................. 6
Figure 2. Map showing the names and locations of the sample cells or plots that were surveyed from 2006 to 2008for the terrestrial field component of the project...............................................................................................8
Figure 3. Distribution of quality scores for all cover type patches from GIS patch analysis. .................................. 21
Figure 4. Example of GIS patch analysis results from Newaygo County. ............................................................. 22
Figure 5. Map showing general distribution of high and moderately scoring cover type patches in the Lower Penin-sula based on results from the GIS patch analysis. ........................................................................................ 23
Figure 6. Map showing general distribution of high and moderately scoring cover type patches in the Upper Penin-sula based on results from the GIS patch analysis. ........................................................................................ 24
List of Appendices
Appendix I. Summary of field metrics to assess the ecological quality or condition of the range of natural communityfield classes associated with the GIS/IFMAP land cover types or classes and MNFI natural communities. ........ A-2
Appendix II. Summary of metrics and criteria used to evaluate the quality or condition of photo-interpreted polygons bynatural community type. ......................................................................................................................................... A-7
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008 Page-1
INTRODUCTION
Land management decisions need to be basedupon a foundation of sound and comprehensiveecological information. Many types of data currentlyexist that can be synthesized into an integratedanalysis of Michigan’s natural features that iscomprehensive and scientifically sound and can informland management decisions. Increased knowledge andanalysis about the natural features on the landscape(e.g. ecological communities that perform importantecosystem services, endangered and threatenedspecies, species of concern, rare ecologicalcommunities) will improve the State’s ability to makewise decisions regarding conservation andmanagement of these resources. The MichiganDepartment of Natural Resources (MDNR) has ledseveral initiatives that have incorporated integratedand comprehensive analyses of the natural features ofMichigan. These include forest certification,development of a wildlife conservation strategy (i.e.,Wildlife Action Plan), the biodiversity/old growthinitiative, and the biodiversity conservation planningprocess.
An important component of these initiatives isthe identification of areas of high ecological value orsignificance for conservation and management. Thesecan include high quality examples of natural features,functional representative native ecosystems, and areaswith concentrations of rare, threatened or endangeredspecies or natural communities. Ideally, identificationof these high ecological value areas as priorities forconservation should be based on information collectedfrom systematic and comprehensive field inventoriesof natural features and other areas of high ecologicalsignificance across the state. However, suchinventories have not been completed to date due tofunding and time constraints. In lieu of havingcomplete information from field surveys, GIS-basedmodels have been developed to help identify potentialareas of high ecological value or significance.However, these models need to be tested, verified, andrefined based on conditions in the field before they canbe applied with confidence to help guide conservationand management decisions.
The goal of this project is to improve theMDNR’s ability to identify terrestrial and aquatic areasof high ecological value. This would be achieved by
developing an effective approach for evaluating andidentifying these areas on the landscape remotely andsystematically surveying these areas to assess theirecological condition and associated plants and animals.Specific objectives of this project include the following:1) develop a GIS model that identifies areas of highecological value statewide using currently availabledata; 2) evaluate air photo interpretation resultscompared to GIS model results; 3) develop and applyfield metrics to assess ecological value; and 4) testGIS model and potentially other existing GIS andremote sensing models with field metrics/data.
This project is envisioned to be the first part of a10 to 20 year effort that will lead to a scientificallysound and comprehensive statewide analysis,identification, and survey of high ecological valueareas (as defined by this project, see below) aspriorities for conservation. This project will addressseveral statewide research needs identified in theMichigan’s Wildlife Action Plan. These includedeveloping and providing tools or approaches forlandscape assessment and identification of high qualityrepresentative occurrences and providing informationto address conservation needs that address statewidepriority threats and issues, and needs to addressecosystem representation and network issues. Theultimate goal of this project is to provide informationthat will inform conservation and land use decisions byfederal, state, and local government agencies and otherentities.
This report provides a summary of projectactivities and progress made on the Statewide Analysisand Survey project during 2008. This project wasinitiated in 2006 and continued in 2007 and 2008. Asubset of the data collected from 2006 through 2008will be summarized and analyzed in 2009. A finalreport summarizing project accomplishments andfindings from 2006 through 2009 will be prepared andcompleted in 2009. A project evaluation also will beconducted in 2009, and results of that evaluation willbe provided in the 2009 final report.
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008 Page-2
METHODS
This project has two major components. Thefirst component is developing an effective and efficientapproach for evaluating the landscape and identifyingor predicting areas of high ecological value orsignificance remotely. To accomplish this, a GISanalysis of the state was conducted by MNFI toidentify terrestrial areas of high ecological value, asdefined by this project. A GIS model assessing thequality or condition of stream reaches that wasdeveloped by Wang et al. (2006) at the Institute forFisheries Research was utilized to identify potentialaquatic areas of high ecological value. The purpose ofthe GIS analysis or component (“GIS component”) isto provide credible, scientifically sound guidance basedupon currently available ecological information toidentify areas with high ecological value and to providethe information in a short time frame for the entirestate.
Terrestrial GIS Analysis/ComponentTerrestrial areas of high ecological value that
have been identified by the GIS analysis have beendefined as high quality cover type patches. A patchanalysis was conducted to identify high quality covertype patches. A cover type patch is a spatially definedarea of the same land cover class or type delineatedusing GIS, based on the MDNR Integrated ForestMonitoring, Assessment and Prescription (IFMAP)land cover dataset (MDNR 2001). The IFMAPdataset is a raster dataset of 30-m resolution cellsderived from remotely-sensed Landsat ThematicMapper imagery. A total of 35 different land coverclassifications are identified in IFMAP. The IFMAPland cover data were reclassified and aggregated intocover type patches of 16 general land cover typeswhich comprised of the following: Upland DeciduousForest, Upland Coniferous Forest, Upland MixedForest, Lowland Deciduous Forest, LowlandConiferous Forest, Lowland Mixed Forest, Grassland,Shrub, Non-forested Wetland, Pines, Pasture/Parks,Agriculture, Sand/Soil, Bare Soil/Rock, Urban, andWater. The agriculture, bare soil/rock, urban, andwater cover types were excluded from the GISanalysis. Because of file size limitations, the state wasdivided into 13 different regions for the analysis, 10 inthe Lower Peninsula and 3 in the Upper Peninsula.
ARCGIS, FRAGSTATS, and SPSS were usedto complete the GIS analysis or model to identify high
quality cover type patches for this component of theproject. Each cover type patch was scored from 1-4,with 4 indicating highest quality, for each of thefollowing three variables or criteria: area, core area,and proximity to similar patches. We initially hadincluded two additional variables in the GIS analysis,edge contrast and shape. However, subsequent testingshowed that core area and edge adjacency werepositively correlated with each other while area andshape were negatively correlated with each other.Core area and edge adjacency correlated becauseFRAGSTATS uses the adjacent cover types whencalculating the patch core area. The negativecorrelation between area and shape was likely anartifact of the IFMAP dataset and the methodFRAGSTATS uses to calculate shape. When testing araster dataset, FRAGSTATS assigns a perfect shapescore to a square. The greater the deviation from asquare, the lower the shape score. In the IFMAPdatasets, larger areas tend to be convoluted patchesconnected together by narrow strips. This gives largepatches a relatively high edge to area ratio and asubsequent low FRAGSTATS shape score. Anotherproblem scoring patches for shape was the inherentlinear shape of some cover type patches. For example,lowland riparian forests in the southern LowerPeninsula and beach areas tend to naturally occur aslinear patches, which resulted in lower scores for thesecover type patches in the analysis. As a result, edgecontrast and shape were removed from the analysis,and only area, core area, and proximity to similarpatches were used in the final GIS analysis.
Area ScoringSome land cover types occur over much larger
areas than do certain other types in Michigan. Thus,cover types were classified into large patch or smallpatch communities for area scoring. Large patchcommunities were Grassland, Upland DeciduousForest, Upland Coniferous Forest, Upland MixedForest, Lowland Deciduous Forest, LowlandConiferous Forest, Lowland Mixed Forest, Shrub,Pines, and Pasture/Parks. Small patch communitieswere Non-forested Wetland and Sand/Soil. Smallpatch communities were given a score from 2 to 4 forarea or size of the patch using the thresholdssummarized in Table 1. Large patch communities weregiven a score from 1 to 4 for area or size of the patchbased on the thresholds summarized in Table 2.
GIS Analysis/Component
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008 Page-3
Patch score Threshold
2 0 <= area < 2 hectares
3 2 hectares <= area < 20 hectares
4 20 hectares <= area
Table 1. Thresholds for patch scores for small patch communities for area scoring.
Patch score Threshold
1 0 < area < 20 hectares
2 20 hectares <= area < 1000 hectares
3 1000 hectares <= area < 2000 hectares
4 area => 2000 hectares
Table 2. Thresholds for patch scores for large patch communities for area scoring.
Core area was determined using FRAGSTATS.FRAGSTATS utilizes user determined depth impactsof adjoining cover types to determine the amount ofcore area. Core area scores for each patch weredetermined using the thresholds summarized in Table3.
Core Area ScoringPatches were given a score from 0 to 4 for the
percentage of the patch that is considered core area.
Proximity ScoringPatches were given a score from 0 to 4 for
proximity to similar patches. FRAGSTATS assigns anopen ended proximity value to each patch. Eachpatch was scored using SPSS to do a two-step clusteranalysis of values assigned by FRAGSTATS to thepatch. The cluster analysis was performed individuallyon each cover type. Those patches given a zeroproximity value by FRAGSTATS were given aproximity score of zero.
Patch score Threshold
0 No core area
1 0 < core area <= .1
2 0.1 < core area <= 0.25
3 0.25 < core area <= 0.5
4 0.5 < core area
Table 3. Thresholds for patch scores for core area scoring.
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008 Page-4
catchment size, catchment surficial geology, catchmentslope, catchment land use, discharge, nutrientconcentration, summer water temperature, valleycharacter, valley width, valley wetlands, channel
Area, core area, and proximity scores were thensummed for each patch, resulting in a total score foreach patch ranging from 3 to 12, with 12 indicating thehighest quality. A score of 12 would indicate a largecover type patch with a high percentage of core areaand in close proximity to other similar cover typepatches. The Pines and Pasture/Parks cover typepatches were scored for each criterion, but theirsummed criteria scores were halved, resulting in totalpatch scores ranging from 1.5 to 6. This was donebecause IFMAP does not differentiate betweennatural pine stands and pine plantations.
Aquatic GIS Analysis/ComponentTo develop an approach for identifying potential
aquatic areas of high ecological value, a GIS modelthat predicts stream/river reach quality or conditionthat was developed by Wang et al. ( in press) at theMDNR’s Institute of Fisheries Research was used forthis component of the study. A reach is generally astretch of stream from confluence to confluence. Thisanalysis or model assessed the quality or condition ofstream/river reaches based on a number of variablesincluding number of active mines, percentage ofnetwork watershed in agricultural and urban land use,number of point source facilities and toxic releasesites, number of road crossings, road density, damdensity, nitrogen loading, and watershed area treatedwith manure (Table 4).
Each stream/river reach in the model also isassociated with a valley segment (i.e., VSEC) and ariver type. Valley segments are aggregates of reachesbased on a number of variables. Variables used in theclassification of VSECs include the following:
character, channel sinuosity, and key fish species.Valley segment boundaries were determined byapplying the following priority criteria: 1) junctions ofsimilar order tributaries, 2) corresponding breaks inland surface form, 3) changes in local groundwatersource, 4) abrupt changes in major land uses, and 5)observed or expected changes in key fish species.River types are MNFI’s draft river naturalcommunities which are based on catchment size,water temperature, and gradient. Four size classes aredefined using drainage areas of VSECs, followingdefinitions outlined in Michigan’s Wildlife Action Plan:1) headwaters and small tributaries are less than 40mi2; 2) medium rivers are between 40-179 mi2; 3)large rivers are between 180-620 mi2; and 4) verylarge rivers are greater than 620 mi2 (Eagle et al.2005). Three classes of water temperatures aredefined as follows: cold (<19�C), cool (19-22�C),and warm (>22�C). Three classes of gradient aredefined as follows: low (<0.001), moderate (0.001-0.006), and high (>0.006). Gradient classes weredefined using the 25th and 75th percentiles of allstream reach gradients across Michigan. All of thesevalues were pulled from the latest VSEC data(VSECxEDUxHUC8_noshoreline.shp).
Based on the variables assessed in the GISmodel, each stream/river reach in the model has beenassigned one of the following six rankings of expectedimpact: reference, no impact, detectable, moderate,heavy, and severe. These rankings can be comparedand combined with field metrics to evaluate anddetermine if this GIS model can be used to accuratelyassess the quality or condition of stream/river reachesin Michigan. For example, a no impact ranking mightindicate higher site quality or higher ecological value asdefined by this project whereas a moderate, heavy orsevere impact ranking might indicate lower site qualityor lower ecological value as defined by this project.
Field Component
The second major component of this projectinvolved developing and applying field metrics toassess the ecological value or condition of the covertype patches and stream/river reaches identified by theGIS models to test and verify the models. The fieldmetrics or data will be used to test the GIS modeldeveloped by MNFI as part of this project. Thesemetrics also could potentially be used to test otherexisting GIS and remote sensing models that identifyareas of high ecological value or conservation priority.
The purpose of the field component (“fieldcomponent”) of this project is to verify and improvethe scientific foundation of the GIS models and provideinformation to enhance future analyses or models thatidentify areas of high ecological value andconservation priority.
Terrestrial Field ComponentTerrestrial areas of high ecological value as
identified by the field metrics for this project are
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008 Page-5
Variables for all streams
Active mining (#/10000 km2)
Network watershed agricultural land use (%)
Network watershed urban land use (%)
MDEQ’s permitted point source facilities (#/100 km2)
MDEQ’s permitted point source facilities having direct connection with stream (#/100 km2)
USEPA’s toxic release inventory sites (#/10000 km2)
Population density (#/km2)
Road crossing (#/km2)
Road density (km/km2)
Total nitrogen plus (phosphorus*10) loading (kg/l/yr)
Watershed area treated with manure from barn yards (m/km)
Additional variables for coldwater streams
Total nitrogen plus (phosphorus*10) yield (kg/l/year)
Additional variables for warmwater streams
Dam density (#/100 km2)
USEPA’s toxic release inventory sites discharging into
surface water (#/10000 km2)
Table 4. Summary of variables utilized in GIS stream/river reach disturbance model developed by Wanget al. (in press) to predict potential quality or condition of stream/river reaches based on predicted impactrankings.
referred to as high quality natural community fieldclasses. Natural community field classes are groups ofnatural communities derived from MNFI naturalcommunity types that were evaluated for quality in thefield based on various field metrics (Table 5). Fieldmetrics were developed and applied to assess theecological condition or quality of the natural communityas well as the quality or condition of the plant andanimal communities. Assessment of the animalcommunity focused on breeding bird communities forthis first phase of the project.
To adequately evaluate the efficacy of the GISmodel, the goal of the terrestrial field component wasto sample and collect field metrics from at least 30cover type patches in each land cover type or classacross a gradient of quality, as indicated by the patchscore. Of the 30 total, the goal was to collect fieldmetrics from approximately 10 cover type patchesfrom the low end of the quality spectrum, 10 from thehigh end, and 10 in the middle (exact definitions of low,medium, and high are not important at this time) tocapture a gradient of quality. Sampling in 2006 and2007 was conducted in upland forest, lowland forest,grassland/shrubland, and non-forested wetland landcover type patches. The number of samples in eachland cover type varied. To ensure sufficient sample
size from at least one land cover type to draw somemeaningful conclusions at the end of this year,sampling in 2008 focused only on one land cover typeor class.
To identify sites for field sampling in 2008, thecover type patches that were surveyed in 2006 and2007 were analyzed to address changes in thesampling scheme. Based on this analysis, it wasdetermined that most of the cover type patches thatwere surveyed in 2006 and 2007 consisted of uplandforest patches, primarily upland deciduous forestpatches, that were ranked as high to moderate qualitybased on the GIS patch analysis scores. In 2006 and2007, a total of 25 sample cells or plots were located inupland deciduous forest cover type patches that wereranked as high to moderate quality. Thus, surveyefforts in 2008 focused on surveying upland deciduousforest cover type patches ranked as low to moderatequality in order to collect field metrics from a sufficientnumber of patches in this cover type across a gradientof quality.
As in previous years of the project, theterrestrial field component in 2008 was conductedprimarily in the Newaygo County area (Figure 1).Some sampling occurred outside of Newaygo County
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008 Page-6
Figure 1. Map showing general location of study area in Michigan (counties highlighted in red).
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008 Page-7
in adjacent counties (i.e., Oceana, Muskegon, Kent,and Mecosta) and multiple ecological subsections in2008 to obtain sufficient numbers of samples of uplanddeciduous forest cover type patches ranked as low tomoderate quality (Figures 1 and 2). As in 2006 and2007, field sampling in 2008 was conducted on stateand federal public lands managed by the MDNR andthe U. S. Forest Service.
Based on project meetings and discussions in2008, several changes in the sampling scheme for theterrestrial field component of the project wereidentified and implemented in 2008. In 2006, fieldmetrics were collected in randomly selected 10-ha(24.7 ac) sample cells or plots within upland forestcover type patches. A grid of 10-ha sample cells wasintersected with the IFMAP land cover dataset. Thesample cells were further stratified in that only samplecells that contained at least 85% of the targeted covertype (i.e., upland forest, lowland forest, grassland/shrub, non-forested wetland) and were on public land(state or federal) were included in the pool of potentialsample cells for field surveys. Sample cells were thenrandomly selected from this pool for field surveys. Forthe forested cover type patches, a random set of morethan 30 sample cells was chosen from the set stratifiedby public ownership. This random set was thenrandomly ordered. If field visits determined that a cellwas not usable for some reason (e.g., incorrect landcover classification, recent management action), thatcell was removed from sampling and the next cell onthe list was sampled. In 2007, a similar method forselecting sample cells was used except that the 10-hasample cells were first located within photo-interpretedpolygons (see section on Aerial Photo InterpretationComponent) in specific cover type patches. Thesampling scheme applied in 2006 and 2007 resulted insome sample cells containing portions of multiple covertype patches, with different patch analysis scores insome cases. The sampling scheme was changed in2008 so that cover type patches were randomlyselected for field sampling instead of sample cells, andsample cells were then randomly placed within covertype patches that were selected for sampling. Thisensured that each sample cell was located within andconsisted of only one cover type patch and not multiplepatches.
Prior to field surveys, upland deciduous forestcover type patches that were ranked as low tomoderate quality, based on the GIS patch analysisscores, were identified and randomly selected for fieldsampling. Sample cells were delineated in the covertype patches selected for field sampling. The sampling
scheme in 2008 also was modified in that if a covertype patch was <10 ha in size, we conducted fieldmeasurements or metrics on the entire patch. If acover type patch was >10 ha in size, we randomlyselected or placed a 10-ha sample cell within thatpatch and conducted field metrics within the samplecell. If the patch was >10 ha but the shape of thepatch did not provide sufficient area for a square orrectangular 10-ha sample cell, we randomly selectedan irregularly shaped 10-ha sample cell for fieldmeasurements.
An aerial photo review of the randomly selectedcover type patches and sample cells was conducted toidentify and prioritize a minimum of 20 sample cells inupland deciduous forest patches with moderate to highquality scores (UDM) and low quality scores (UDL).All UDM sample cells consisted of square, 10-hasample cells, whereas all UDL sample cells consistedof irregular polygons. A minimum of 10 qualifyingUDM sample cells and 10 UDL sample cells wereidentified following aerial photo review and rejection ofapparent non-qualifying plots (e.g. plots with no visibleor reasonable access, or those lacking a sufficientextent of contiguous forest vegetation, etc.). It wasnoted that subsequent rejections could occur duringinitial field visits, and thus a list of at least 15 sites foreach category were reviewed for initial approval andfield reconnaissance.
Following the selection of sites, a folder or sitepackage was prepared and printed for Botany,Ecology, and Zoology project staff. Each site folder orpackage contained two sets of labeled aerial photos foreach of the plots (i.e., 1998 CIR and the 2005 colorimagery) in addition to a representative topographicmap layer. No plat maps were printed, as all plots fellwithin public land boundaries.
Ecological field surveys/samplingEcological field sampling was conducted from 16
June to 8 August 2008. Sample cells or plots in 10UDM and 10 UDL cover type patches were surveyedin 2008. The ecological quality or condition of thenatural community within the sample cell was assessedbased on up to 11 field metrics/criteria, with eachmetric having a score from 1-4 (4 indicating excellentcondition and 1 indicating poor condition). Field metricswere developed in 2006 were for each naturalcommunity field class (Appendix 1). These metricswere developed to score indicators of structural andfunctional biodiversity within a variety of landscapes(Noss 1990, 1999, McElhinny et al. 2005, Lindenmayeret al. 2000, 2006). These metrics can be divided into
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008 Page-8
Figure 2. Map showing the names and locations of the sample cells or plots that were surveyed from2006 to 2008 for the terrestrial field component of the project. Ecological, botanical, and zoological(breeding bird) surveys were conducted in these sample cells. Sample cells were primarily located inNewaygo County and subsection VII.3, but some sample cells also were located in surrounding countiesand subsections.
VII.4
VII.3
VII.2.1
VI.3.2VI.4.2
MECOSTA
NEWAYGO
MUSKEGON
OCEANA
KENT
#
18476
#
16408
#
16303#
16179#
14651#
14527
#
9333
#
9464
#
17393#
17646#
18783#
18388
#
23215
#
21399#
21627
#
20360
#21753#
21246#
20231# 20490
#
19413
#
17644
#
9461
#
8705
#
17851
#UDL12
#
UDL11
#
UDL5
#
UDL8
#
UDL4
#
UDL14
#
UDL7
#
UDL3
#
UDL6#
UDL9
#
UDM5
#
UDM4#
UDM3
#
UDM7
#
UDM11
#
UDM2
#
UDM10
#
UDM9#
UDM8
#UDM1
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008 Page-9
three categories evaluating the landscape context,present condition, and threats facing of the naturalcommunity.
The upland forest land cover type/class in theGIS patch analysis model includes two naturalcommunity field classes, upland oak and oak-pineforest and upland mesic forest, which, in turn, relate toseveral MNFI natural community types (Table 5).Thus, the field metrics that were utilized for evaluatingthe condition of the natural community in the UDMand UDL sample cells consisted of the field metricsthat have been developed for the upland mesic forestand upland oak and oak-pine natural community fieldclasses (Table 6). The specific set of metrics thatwere used depended upon the natural community classthat occurred in the field in the sample cell. Thecriteria that were used to evaluate and score the fieldmetrics for these natural community field classes aresummarized in Tables 7 and 8. The same ecologicalmetrics were utilized in 2006 and 2007. Scores forthese field metrics were derived by walking throughand assessing the field metrics for the entire samplecell in the targeted cover type patch.
Botanical field surveys/floristic samplingEarly season floristic sampling ensued on June
16, and early surveys were conducted on 20 samplecells or plots (10 UDLs and 10 UDMs) through thefirst week of July. A handheld Ipaq computer wasused for all field surveys to ensure thorough coverageof the sample cell and in many cases simply to locateboundaries. A comprehensive flora list was compiledduring meander-searches of all sample cells.Specimens were collected and pressed for all speciesthat could not be identified in the field and which weresufficiently fertile or in a condition suitable for keyingand subsequent identification. The time upon enteringand leaving each sample cell was recorded, and inaddition, five representative photos were recorded foreach sample cell. One sample cell was subsequentlyrejected as a sample site when examined in the field,and avoidance of this sample cell as a suitable samplesite was immediately communicated to the Zoologyand Ecology Section staff. When this sample cell wasrejected, the next available sample cell on the list wasqueued up for sampling.
Late season floristic sampling took place fromearly to mid-September. Prior to sampling, all samplecell or plot data, with the exception of the collectedspecimens, were entered into the Michigan FQAprogram to prepare printed plant lists for each samplecell, such that an alphabetized list could be carried in
the field for each sample cell. When encountered,new species were appended to the list, and anyadditional unknowns were collected for subsequentkeying and verification. All 20 sample cells wererevisited for the late summer inventory, with eachsample cell producing a significant number ofadditional species.
All pressed specimens were dried and placed incollection folders. Once identification of thesespecimens is confirmed after consultation withspecialists at The University of Michigan, thesespecies will be added to the sample cell or plot data asappropriate, and all related FQA-measures (e.g. FQI)will be recalculated and summarized. Approximately50-100 individual species collections were made fordetermination.
Plant information collected during the early andlate season floristic sampling was used to assess thequality or condition of the plant community in thesample cells. The quality or condition of the plantcommunity was determined primarily using the FloristicQuality Assessment (FQA) system or program(Herman et al. 2001). This program or approachassesses quality based on the following metrics: 1)plant species presence/composition, 2) speciesrichness (i.e., total number of species present), 3) theFloristic Quality Index (FQI) (i.e., for each speciesand sample cell), 4) mean coefficient of conservatism,5) mean coefficient of wetness, and 6) summary ofphysiognomic classes for both native and non-nativeplant species (Table 9).
Zoological field samplingBirds have been considered good indicators of
ecosystem integrity due to their relatively low birthrates and long life spans which make them sensitive tochanges in the environment (Maurer 1993). Inaddition, the diversity of avifauna is strongly correlatedwith habitat type and quality making habitatspecialization common (MacArthur and MacArthur1961). This strong association with habitat and theprevalence of habitat loss and alteration in recenthistory has been frequently linked together to explainthe population declines of many Nearctic migrant birdsthat winter in the tropics (Rappole and McDonald1994). These characteristics make avifauna a usefulfocal group to examine the relationships among habitatcharacteristics, habitat quality, conservation, andspecies presence or absence.
Breeding bird surveys were conducted in Mayand June of 2008. These surveys were conducted in
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008 Page-10
GIS Model Land Cover Type/Class
Natural Community Field Class MNFI Natural Community M
NF
I Map
pin
g C
od
e(s)
Up
lan
d F
ore
st
Lo
wla
nd
Fo
rest
No
n-f
ore
sted
Wet
lan
d
Gra
ssla
nd
/Sh
rub
Acidic peatlands
Bog 6121 xPoor fen 6223 xPoor conifer swamp 4234 xMuskeg 6125 x
Intermittent wetlandsCoastal plain marsh 62253 xIntermittent wetland 62251 x
Upland prairies and savanna
Dry sand prairie 31111 xMesic sand prairie 31112 xOak barrens 332 xOak openings 336 xOak-pine barrens 334 xPine barrens 333 x
Upland oak and oak-pine forests
Dry southern forest 4123 xDry-mesic northern forest 4311 xDry-mesic southern forest 4122 x
Upland mesic forestsMesic northern forest 4112 xMesic southern forest 4111 x
Lowland mixed and coniferous forests
Hardwood-conifer swamp 432 xRich tamarack swamp 4232, 4321 xRich conifer swamp 4231, 4242 x
Lowland deciduous forests
Southern hardwood swamp 414 xNorthern hardwood swamp 414 x
Floodplain forest Floodplain forest 415 x
Shrub-dominated wetlands
Inundated shrub swamp 6123 xNorthern shrub thicket 6122 xSouthern shrub-carr 6126 x
Emergent and submergent marshes
Emergent marsh 6221 xInland salt marsh 6224 xSubmergent marsh 621 x
Wet meadows, wet and wet-mesic prairies, and
calcareous fen
Wet-mesic sand prairie 625 xNorthern fen 6262 xNorthern wet meadow 6223 xSouthern wet meadow 6223 xWet prairie 625 xWet-mesic prairie 625 x
* Aspen was typed as 413, then assigned to upland mesic or upland oak and oak-pine forest class.
Table 5. Crosswalk showing relationships between GIS model land cover types/classes, natural commu-nity field classes, and MNFI natural community classifications.
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008 Page-11
Upland forest
Dry
so
uth
ern
fo
rest
Dry
-mes
ic n
ort
her
n f
ore
st
Dry
-mes
ic s
ou
ther
n f
ore
st
Mes
ic n
ort
her
n f
ore
st
Mes
ic s
ou
ther
n f
ore
st
Metrics
Upland oak and oak-pine
forests
Upland mesic
forests
Landscape context**
Buffer width X X X X X
Broader land use X X X X X
Site intactness
Presence of roads, RR tracks, or other
(residential, industrial, commercial)
development X X X X X
Hydrology
Presence of drains, ditches, channelization X X X X X
Soil erosion X X X X X
ORV damage
Stream morphology
Community structure and function
Vegetative structure and composition X X X X X
Evidence of logging X X X X X
Evidence of plowing
Evidence of fire X X X
Presence of invasive non-native vascular
plant spp. X X X X X
Presence of rotting logs and snags X X X X X
Tree regeneration (oak, conifer, etc. as
appropriate) X X X
Evidence of deer browse X X X X X
Species-specific die-offs X X X X X
Condition of sphagnum peat
Cat-tail invasion (indicator of nutrient
enrichment)
Presence of red maple (Acer rubrum ) X X X
Pit-and-mound topography X X
Dominance by native grasses and/or sedges
Table 6. Summary of field metrics utilized to assess the ecological quality or condition of the sample cellsin upland oak and oak-pine forest and upland mesic forestnatural community field classes which areassociated with upland forest cover type patches.
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008 Page-12
4 =
Ex
cell
ent
3 =
Go
od
2 =
Fa
ir1
= P
oo
r
Bu
ffer
wid
thO
ccu
rren
ce i
s b
uff
ered
by
+/-
nat
ura
l
lan
d >
10
0m
on
en
tire
per
iph
ery
.
Occ
urr
ence
is
wel
l-b
uff
ered
(5
0-1
00
m)
on
>7
5%
of
per
iph
ery
.
Occ
urr
ence
is
wel
l-b
uff
ered
on
>5
0%
of
per
iph
ery
or
nar
row
ly b
uff
ered
(2
5 t
o
50
m)
on
>7
5%
of
per
iph
ery
.
Occ
urr
ence
po
orl
y b
uff
ered
(<
25
m o
n
enti
re p
erip
her
y),
or
no
t b
uff
ered
.
Bro
ad
er l
an
d u
seN
atu
ral
(mat
rix
) la
nd
cov
er d
om
inat
es
(>7
5%
of
lan
dsc
ape)
.
Lan
dsc
ape
is p
arti
ally
nat
ura
l (5
0-7
5%
),
rem
ain
der
ag
ricu
ltu
ral.
A
gri
cult
ura
l
area
s ar
e is
ola
ted
by
nat
ura
l co
ver
.
Lan
dsc
ape
pre
do
min
antl
y a
gri
cult
ura
l,
wit
h s
om
e n
atu
ral
cov
er (
33
-50
%).
Nat
ura
l la
nd
s ar
e is
ola
ted
by
agri
cult
ura
l fi
eld
s.
Lan
dsc
ape
urb
an o
r su
bu
rban
, o
r al
mo
st
com
ple
tely
ag
ricu
ltu
ral
(<3
3%
nat
ura
l,
nat
ura
l la
nd
s b
ein
g i
sola
ted
, sm
all
po
cket
s).
Pre
sen
ce o
f ro
ad
s, R
R t
rack
s, o
r
oth
er (
resi
den
tia
l, i
nd
ust
ria
l,
com
mer
cia
l) d
evel
op
men
t
Occ
urr
ence
is
un
frag
men
ted
by
ro
ads
or
oth
er d
evel
op
men
t.
Occ
urr
ence
is
ver
y l
oca
lly
im
pac
ted
by
dev
elo
pm
ent
(est
. <
10
% a
rea
imp
acte
d).
Occ
urr
ence
mo
der
atel
y i
mp
acte
d b
y
dev
elo
pm
ent
(est
. <
25
% o
f ar
ea
imp
acte
d).
R
oad
(s)
bu
ilt
thro
ug
h
wet
lan
d.
Occ
urr
ence
is
sig
nif
ican
tly
im
pac
ted
by
road
s an
d/o
r o
ther
dev
elo
pm
ent
(est
.
>2
5%
of
area
im
pac
ted
).
Ro
ads,
tra
ils,
pip
elin
es,
etc.
sig
nif
ican
tly
alt
er
wet
lan
d h
yd
rolo
gy
.
So
il c
om
pa
ctio
n/
ero
sio
n
So
il e
ssen
tial
ly u
nd
istu
rbed
, ex
cep
t
alo
ng
tra
ils
or
nea
rby
ro
ads.
So
il l
oca
lly
dis
turb
ed,
bu
t th
at p
ort
ion
occ
up
yin
g <
10
% o
f ar
ea.
So
il m
od
erat
ely
dis
turb
ed,
bu
t
un
plo
wed
.
So
il p
low
ed,
or
sig
nif
ican
tly
co
mp
acte
d
by
lo
gg
ing
eq
uip
men
t, o
r le
af l
itte
r
rem
ov
ed.
Veg
eta
tiv
e st
ruct
ure
V
eget
ativ
e st
ruct
ure
wit
hin
ex
pec
ted
ran
ge
of
var
iati
on
. C
ano
py
, u
nd
erst
ory
and
gro
un
dla
yer
veg
etat
ion
ty
pic
al f
or
par
ticu
lar
com
mu
nit
y t
yp
e.
Go
od
pat
chin
ess,
can
op
y g
aps
pre
sen
t.
On
e (b
ut
no
t m
ore
) o
f th
e fo
llo
win
g i
s
tru
e: c
ano
py
tre
es a
re o
f th
e sa
me
size
clas
s, u
nd
erst
ory
is
abse
nt
or
aty
pic
al,
gro
un
dla
yer
is
dep
aup
erat
e o
r at
yp
ical
,
or
site
lac
ks
can
op
y g
aps
and
pat
chin
ess.
Tw
o o
f th
e at
trib
ute
s d
escr
ibed
fo
r
Gra
de
2 s
ites
are
tru
e.
May
dis
pla
y
po
or
pat
chin
ess
or
low
pre
sen
ce o
f
can
op
y g
aps.
Veg
etat
ive
stru
ctu
re s
ign
ific
antl
y
alte
red
at
all
thre
e le
vel
s.
Tre
es
imm
atu
re o
r o
f a
sin
gle
siz
e cl
ass.
Un
der
sto
ry p
oo
rly
dev
elo
ped
.
Gro
un
dla
yer
veg
etat
ion
sig
nif
ican
tly
red
uce
d o
r ab
sen
t.
Pre
sen
ce o
f ro
ttin
g l
og
s a
nd
sn
ag
sH
igh
ly d
eco
mp
ose
d r
ott
ing
lo
gs
com
mo
n o
n f
ore
st f
loo
r.
Man
y l
og
s
mo
ss-c
ov
ered
, m
ois
t.
Sn
ags
com
mo
n.
Hig
hly
dec
om
po
sed
ro
ttin
g l
og
s
com
mo
n o
n f
ore
st f
loo
r.
Man
y l
og
s
mo
ss-c
ov
ered
, m
ois
t.
Sn
ags
rare
.
Ro
ttin
g l
og
s in
ad
van
ced
sta
ges
of
dec
om
po
siti
on
rar
e.
Lo
gs
in l
ess
adv
ance
d s
tag
es o
f d
eco
mp
osi
tio
n (
litt
le
to n
o m
oss
co
ver
, et
c.)
occ
asio
nal
to
com
mo
n.
Fo
rest
flo
or
lack
s d
eco
mp
osi
ng
wo
od
.
Ro
ttin
g l
og
s ab
sen
t o
r ra
re.
Wo
od
y
sub
stra
te i
s re
stri
cted
to
sm
all
fall
en
bra
nch
es,
if p
rese
nt.
Pit
-an
d-m
ou
nd
to
po
gra
ph
yF
ore
st f
loo
r is
ch
arac
teri
zed
by
div
erse
pit
-an
d-m
ou
nd
to
po
gra
ph
y,
wit
h p
its
and
mo
un
ds
of
var
iou
s si
zes.
Maj
ori
ty (
>7
5%
) o
f fo
rest
flo
or
is
char
acte
rize
d b
y d
iver
se p
it-a
nd
-mo
un
d
top
og
rap
hy
, w
ith
pit
s an
d m
ou
nd
s o
f
var
iou
s si
zes.
Pit
-an
d-m
ou
nd
to
po
gra
ph
y p
rese
nt,
bu
t
no
t co
mm
on
, an
d l
ikel
y o
f re
cen
t o
rig
in.
Lac
k o
f p
its
and
mo
un
ds
ind
icat
e
form
er s
avan
na
con
dit
ion
s o
r p
low
ing
.
Pit
-an
d-m
ou
nd
to
po
gra
ph
y a
bse
nt.
S
ite
was
lik
ely
pre
-set
tlem
ent
sav
ann
a, o
r
was
plo
wed
.
Co
mm
un
ity
stru
ctu
re a
nd
fu
nct
ion
Fie
ld M
etri
c
Met
ric
rati
ng
cri
teri
a
La
nd
sca
pe
con
tex
t**
Sit
e in
tact
nes
s
Tabl
e 7. S
umm
ary
of cr
iteri
a fo
r fie
ld m
etri
cs fo
r eva
luat
ing
the e
colo
gica
l qua
lity
or co
nditi
on o
f sam
ple c
ells
in th
e upl
and
mes
ic fo
rest
natu
ral c
omm
unity
fiel
d cl
ass.
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008 Page-13
4 =
Exce
llen
t3 =
Good
2 =
Fair
1 =
Poor
Fie
ld M
etri
c
Met
ric
rati
ng c
rite
ria
Ev
iden
ce o
f lo
gg
ing
No
ev
iden
ce o
f lo
gg
ing
, o
r o
nly
ver
y
min
or
cutt
ing
wit
ho
ut
spec
ies
elim
inat
ion
.
Sel
ecti
ve
cutt
ing
, b
ut
larg
e, o
ld t
rees
of
all
typ
ical
sp
ecie
s re
mai
n.
Wid
esp
read
cu
ttin
g >
50
yea
rs b
efo
re
pre
sen
t.
Can
op
y t
rees
mat
ure
.
Wid
esp
read
cu
ttin
g <
50
yea
rs b
efo
re
pre
sen
t.
Can
op
y t
rees
im
mat
ure
an
d/o
r
of
smal
l d
iam
eter
.
Her
biv
ore
im
pa
cts:
dee
r a
nd
oth
er
larg
e m
am
ma
lia
n h
erb
ivo
res
Sig
ns
of
dee
r p
rese
nce
(b
row
sin
g,
trai
ls,
scat
) ra
re.
Occ
asio
nal
bro
wsi
ng
—fe
wer
th
an 1
0%
of
stem
s o
f p
refe
rred
sp
ecie
s (b
ud
s o
n
wo
od
y p
lan
ts o
r an
yth
ing
lil
iace
ou
s)
clip
ped
, an
d p
refe
rred
sp
ecie
s st
ill
abu
nd
ant.
Dee
r tr
ails
an
d s
cat
no
tice
able
.
Mo
der
ate
lev
el o
f b
row
sin
g—
10
to
25
% o
f st
ems
of
pre
ferr
ed s
pec
ies
clip
ped
. D
eer
trai
ls a
nd
sca
t co
mm
on
.
Sev
ere
bro
wsi
ng
. B
row
se l
ines
vis
ible
on
sh
rub
s an
d t
rees
(th
at t
op
iary
lo
ok
on
Jun
iper
us)
; p
refe
rred
sp
ecie
s
pro
min
entl
y b
row
sed
an
d/o
r ra
re o
r
abse
nt
fro
m s
ite,
an
d l
ess
pre
ferr
ed
spec
ies
also
bro
wse
d.
Dee
r tr
ails
an
d
scat
ab
un
dan
t.
Sp
ecie
s-sp
ecif
ic t
ree
mo
rta
lity
(d
ue
to
inse
cts
or
dis
ease
ra
ther
th
an
sh
ad
e
sup
pre
ssio
n,
win
dth
row
, et
c.)
No
ev
iden
ce o
f co
nif
er o
r h
ard
wo
od
die
-
off
du
e to
pes
ts (
hem
lock
wo
oll
y
adel
gid
, e
mer
ald
ash
bo
rer,
bee
ch b
ark
dis
ease
, an
thra
cno
se,
etc.
).
Pas
t d
ie-o
ff
of
larg
e A
mer
ican
elm
is
exp
ecte
d a
nd
no
t sc
ore
d.
Ver
y l
ittl
e d
ie-o
ff n
ote
d.
Les
s th
an 1
0%
of
ind
ivid
ual
s o
f an
y t
ree
spp
. af
fect
ed
by
in
sect
or
fun
gal
pes
ts.
Th
ese
tree
s
wid
ely
sca
tter
ed o
r in
iso
late
d p
atch
es.
Tre
e d
ie-o
ff s
ign
ific
ant.
U
p t
o 2
5%
of
the
can
op
y a
ffec
ted
by
pes
ts o
r
pat
ho
gen
s.
Ex
ten
siv
e d
ie-o
ff o
f ca
no
py
tre
es n
ote
d.
Gre
ater
th
an 2
5%
of
can
op
y t
rees
are
affe
cted
by
pes
ts o
r p
ath
og
ens.
Pre
sen
ce o
f in
va
siv
e n
on
-na
tiv
e
an
d/o
r n
ati
ve
va
scu
lar
pla
nt
spp
.
Inv
asiv
e sp
ecie
s o
f si
gn
ific
ant
imp
act
abse
nt.
O
ther
no
n-n
ativ
e sp
p.
may
be
pre
sen
t, i
n l
ow
nu
mb
ers.
Inv
asiv
e sp
p.
of
sig
nif
ican
t im
pac
t
pre
sen
t, b
ut
hig
hly
lo
cali
zed
(li
mit
ed t
o
edg
es o
r sm
all,
wid
ely
sp
aced
co
lon
ies)
.
Inv
asiv
e sp
p.
of
sig
nif
ican
t im
pac
t
pre
sen
t, o
ccas
ion
al.
May
lo
call
y
do
min
ate
com
mu
nit
y (
<2
5%
).
Inv
asiv
e sp
p.
of
sig
nif
ican
t im
pac
t
com
mo
n o
r d
om
inat
ing
>2
5%
of
occ
urr
ence
.
* I
ncl
ud
es m
esic
no
rth
ern
fo
rest
, m
esic
so
uth
ern
fo
rest
** M
ET
RIC
AS
SE
SS
ED
OU
TS
IDE
TH
E B
OU
ND
AR
IES
OF
TH
E 1
0 H
A S
AM
PL
E C
EL
L.
Co
mm
un
ity
stru
ctu
re a
nd
fu
nct
ion
(co
nti
nu
ed)
Tabl
e 7. C
ontin
ued.
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008 Page-14
4 =
Exce
llen
t3 =
Good
2 =
Fair
1 =
Poor
Bu
ffer
wid
thO
ccu
rren
ce i
s b
uff
ered
by
+/-
nat
ura
l la
nd
>1
00
m o
n e
nti
re
per
iph
ery
.
Occ
urr
ence
is
wel
l-b
uff
ered
(5
0-
10
0m
) o
n >
75
% o
f p
erip
her
y.
Occ
urr
ence
is
wel
l-b
uff
ered
on
>5
0%
of
per
iph
ery
or
nar
row
ly
bu
ffer
ed (
25
to
50
m)
on
>7
5%
of
per
iph
ery
.
Occ
urr
ence
po
orl
y b
uff
ered
(<2
5m
on
en
tire
per
iph
ery
), o
r n
ot
bu
ffer
ed.
Bro
ad
er l
an
d u
seN
atu
ral
(mat
rix
) la
nd
cov
er
do
min
ates
(>
75
% o
f la
nd
scap
e).
Lan
dsc
ape
is p
arti
ally
nat
ura
l (5
0-
75
%),
rem
ain
der
ag
ricu
ltu
ral.
Ag
ricu
ltu
ral
area
s ar
e is
ola
ted
by
nat
ura
l co
ver
.
Lan
dsc
ape
pre
do
min
antl
y
agri
cult
ura
l, w
ith
so
me
nat
ura
l
cov
er (
33
-50
%).
N
atu
ral
lan
ds
are
iso
late
d b
y a
gri
cult
ura
l fi
eld
s.
Lan
dsc
ape
urb
an o
r su
bu
rban
, o
r
alm
ost
co
mp
lete
ly a
gri
cult
ura
l
(<3
3%
nat
ura
l, n
atu
ral
lan
ds
bei
ng
iso
late
d,
smal
l p
ock
ets)
.
Pre
sen
ce o
f ro
ad
s, R
R
tra
cks,
or
oth
er
(res
iden
tia
l, i
nd
ust
ria
l,
com
mer
cia
l) d
evel
op
men
t
Occ
urr
ence
is
un
frag
men
ted
by
road
s o
r o
ther
dev
elo
pm
ent.
Occ
urr
ence
is
ver
y l
oca
lly
imp
acte
d b
y d
evel
op
men
t (e
st.
<1
0%
are
a im
pac
ted
).
Occ
urr
ence
mo
der
atel
y i
mp
acte
d
by
dev
elo
pm
ent
(est
. <
25
% o
f
area
im
pac
ted
).
Ro
ad(s
) b
uil
t
thro
ug
h w
etla
nd
.
Occ
urr
ence
is
sig
nif
ican
tly
imp
acte
d b
y r
oad
s an
d/o
r o
ther
dev
elo
pm
ent
(est
. >
25
% o
f ar
ea
imp
acte
d).
R
oad
s, t
rail
s,
pip
elin
es,
etc.
sig
nif
ican
tly
alt
er
wet
lan
d h
yd
rolo
gy
.
So
il c
om
pa
ctio
n/e
rosi
on
S
oil
ess
enti
ally
un
dis
turb
ed,
exce
pt
alo
ng
tra
ils
or
nea
rby
road
s.
So
il l
oca
lly
dis
turb
ed,
bu
t th
at
po
rtio
n o
ccu
py
ing
<1
0%
of
area
.
So
il m
od
erat
ely
dis
turb
ed,
bu
t
un
plo
wed
.
So
il p
low
ed,
or
sig
nif
ican
tly
com
pac
ted
by
lo
gg
ing
eq
uip
men
t,
or
leaf
lit
ter
rem
ov
ed.
Veg
eta
tiv
e st
ruct
ure
(esp
ecia
lly
pre
sen
ce o
f
oa
k a
nd
pin
e
reg
ener
ati
on
as
seed
lin
gs,
sap
lin
gs,
an
d u
nd
erst
ory
tree
s)
Can
op
y,
un
der
sto
ry a
nd
gro
un
dla
yer
veg
etat
ion
ty
pic
al f
or
par
ticu
lar
com
mu
nit
y t
yp
e.
Go
od
het
ero
gen
eity
. O
aks
and
/or
pin
es
sho
w a
mp
le r
egen
erat
ion
in
all
size
cla
sses
.
On
e (b
ut
no
t m
ore
) o
f th
e
foll
ow
ing
is
tru
e: c
ano
py
tre
es a
re
of
the
sam
e si
ze c
lass
, u
nd
erst
ory
is a
bse
nt
or
aty
pic
al,
gro
un
dla
yer
is d
epau
per
ate
or
aty
pic
al,
or
site
is h
om
og
eneo
us.
Tw
o o
f th
e fo
llo
win
g a
re t
rue:
can
op
y t
rees
are
of
the
sam
e si
ze
clas
s, u
nd
erst
ory
is
abse
nt
or
aty
pic
al,
gro
un
dla
yer
is
dep
aup
erat
e o
r at
yp
ical
. P
oo
r
het
ero
gen
eity
.
Veg
etat
ive
stru
ctu
re s
ign
ific
antl
y
alte
red
at
all
thre
e le
vel
s.
Tre
es
imm
atu
re o
r o
f a
sin
gle
siz
e cl
ass.
Un
der
sto
ry p
oo
rly
dev
elo
ped
.
Gro
un
dla
yer
veg
etat
ion
sig
nif
ican
tly
red
uce
d o
r ab
sen
t.
Pre
sen
ce o
f re
d m
ap
le
(Ace
r ru
bru
m)
Red
map
le u
nco
mm
on
or
occ
asio
nal
, n
ot
a si
gn
ific
ant
com
po
nen
t (c
ov
er >
25
%)
in a
ny
stra
tum
. A
ll s
ize
clas
ses
rep
rese
nte
d.
Mo
no
do
min
ant
stan
ds
abse
nt.
Red
map
le o
ccas
ion
al.
May
be
a
sig
nif
ican
t co
mp
on
ent
(co
ver
>2
5%
) o
f o
ne
stra
tum
.
Red
map
le o
ccas
ion
al t
o c
om
mo
n,
com
pri
sin
g >
25
% c
ov
er i
n t
wo
stra
ta.
Red
map
le c
om
mo
n t
o a
bu
nd
ant,
com
pri
sin
g >
25
% c
ov
er i
n t
he
can
op
y,
sub
can
op
y,
and
sh
rub
lay
ers.
Fie
ld M
etri
c
Met
ric
rati
ng c
rite
ria
Com
mu
nit
y st
ruct
ure
an
d f
un
ctio
n
Lan
dsc
ape
con
text*
*
Sit
e in
tact
nes
s
Tabl
e 8. S
umm
ary
of cr
iteri
a fo
r fie
ld m
etri
cs fo
r eva
luat
ing
the e
colo
gica
l qua
lity
or co
nditi
on o
f sam
ple c
ells
in th
e upl
and
oak
and
oak-
pine
fore
st n
atur
al co
mm
unity
fiel
d cl
ass.
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008 Page-15
4 =
Ex
cell
ent
3 =
Go
od
2 =
Fa
ir1
= P
oo
r F
ield
Met
ric
Met
ric
rati
ng
cri
teri
a
Co
ars
e w
oo
dy
deb
ris
Co
arse
wo
od
y d
ebri
s p
rese
nt
and
abu
nd
ant
in a
ll e
xp
ecte
d s
ize
and
dec
om
po
siti
on
cla
sses
. S
nag
s an
d
cav
ity
tre
es c
om
mo
n.
See
ex
cell
ent
rati
ng
, b
ut
lack
ing
snag
s an
d/o
r ca
vit
y t
rees
.
Co
arse
wo
od
y d
ebri
s ab
sen
t o
ver
a si
gn
ific
ant
po
rtio
n o
f th
e fo
rest
(>2
5%
).
Siz
e cl
ass
and
dec
om
po
siti
on
cla
sses
po
orl
y
rep
rese
nte
d.
Sn
ags
and
cav
ity
tree
s , i
f p
rese
nt,
are
rar
e.
Co
arse
wo
od
y d
ebri
s n
earl
y
abse
nt.
O
nly
sm
all
bra
nch
es a
nd
rece
nt
tree
fall
pre
sen
t.
Dec
om
po
siti
on
min
imal
. S
nag
s
and
cav
ity
tre
es a
bse
nt.
Ev
iden
ce o
f lo
gg
ing
No
ev
iden
ce o
f lo
gg
ing
, o
r o
nly
ver
y m
ino
r cu
ttin
g w
ith
ou
t sp
ecie
s
elim
inat
ion
.
Sel
ecti
ve
cutt
ing
, b
ut
larg
e, o
ld
tree
s o
f al
l ty
pic
al s
pec
ies
rem
ain
.
Wid
esp
read
cu
ttin
g >
50
yea
rs
bef
ore
pre
sen
t.
Can
op
y t
rees
mat
ure
.
Wid
esp
read
cu
ttin
g <
50
yea
rs
bef
ore
pre
sen
t.
Can
op
y t
rees
imm
atu
re a
nd
/or
of
smal
l
dia
met
er.
Her
biv
ore
im
pa
cts:
dee
r
an
d o
ther
la
rge
ma
mm
ali
an
her
biv
ore
s
Sig
ns
of
dee
r p
rese
nce
(b
row
sin
g,
trai
ls,
scat
) ra
re.
Occ
asio
nal
bro
wsi
ng
—fe
wer
th
an
10
% o
f st
ems
of
pre
ferr
ed s
pec
ies
(bu
ds
on
wo
od
y p
lan
ts o
r an
yth
ing
lili
aceo
us)
cli
pp
ed,
and
pre
ferr
ed
spec
ies
stil
l ab
un
dan
t. D
eer
trai
ls
and
sca
t n
oti
ceab
le.
Mo
der
ate
lev
el o
f b
row
sin
g—
10
to 2
5%
of
stem
s o
f p
refe
rred
spec
ies
clip
ped
. D
eer
trai
ls a
nd
scat
co
mm
on
.
Sev
ere
bro
wsi
ng
. B
row
se l
ines
vis
ible
on
sh
rub
s an
d t
rees
(th
at
top
iary
lo
ok
on
Ju
nip
eru
s);
pre
ferr
ed s
pec
ies
pro
min
entl
y
bro
wse
d a
nd
/or
rare
or
abse
nt
fro
m s
ite,
an
d l
ess
pre
ferr
ed
spec
ies
also
bro
wse
d.
Dee
r tr
ails
and
sca
t ab
un
dan
t.
Sp
ecie
s-sp
ecif
ic t
ree
mo
rta
lity
(d
ue
to i
nse
ct
pes
ts,
fun
gu
s, o
r d
isea
se)
No
ev
iden
ce o
f tr
ee d
ie-o
ff d
ue
to
pes
ts (
gy
psy
mo
th,
oak
wil
t, p
ine
rust
, et
c.).
Ver
y l
ittl
e d
ie-o
ff n
ote
d.
Les
s
than
10
% o
f in
div
idu
als
of
any
tree
sp
p.
affe
cted
by
in
sect
or
fun
gal
pes
ts.
Th
ese
tree
s w
idel
y
scat
tere
d o
r in
iso
late
d p
atch
es.
Tre
e d
ie-o
ff s
ign
ific
ant.
U
p t
o
25
% o
f th
e ca
no
py
aff
ecte
d b
y
pes
ts o
r p
ath
og
ens.
Ex
ten
siv
e d
ie-o
ff o
f ca
no
py
tre
es
no
ted
. G
reat
er t
han
25
% o
f
can
op
y t
rees
are
aff
ecte
d b
y p
ests
or
pat
ho
gen
s.
Pre
sen
ce o
f in
va
siv
e n
on
-
na
tiv
e a
nd
/or
na
tiv
e
va
scu
lar
pla
nt
spp
.
Inv
asiv
e sp
ecie
s o
f si
gn
ific
ant
imp
act
abse
nt.
O
ther
no
n-n
ativ
e
spec
ies
may
be
pre
sen
t in
lo
w
nu
mb
ers.
Inv
asiv
e sp
p.
of
sig
nif
ican
t im
pac
t
pre
sen
t, b
ut
hig
hly
lo
cali
zed
(lim
ited
to
ed
ges
or
smal
l, w
idel
y
s pac
ed c
olo
nie
s).
Inv
asiv
e sp
p.
of
sig
nif
ican
t im
pac
t
pre
sen
t, o
ccas
ion
al.
May
lo
call
y
do
min
ate
com
mu
nit
y (
<2
5%
).
Inv
asiv
e sp
p.
of
sig
nif
ican
t im
pac
t
com
mo
n o
r d
om
inat
ing
>2
5%
of
occ
urr
ence
.
* I
ncl
ud
es d
ry s
ou
ther
n f
ore
st,
dry
-mes
ic s
ou
ther
n f
ore
st,
dry
-mes
ic n
ort
her
n f
ore
st
**
ME
TR
IC A
SS
ES
SE
D O
UT
SID
E T
HE
BO
UN
DA
RIE
S O
F T
HE
10
HA
SA
MP
LE
CE
LL
.
Co
mm
un
ity
stru
ctu
re a
nd
fu
nct
ion
(co
nti
nu
ed)
Tabl
e 8. C
ontin
ued.
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008 Page-16
Botanical Field Metrics
Species presence/composition
Species richness (total number of species)
Floristic Quality Index (FQI) - for each species and sample cell
Mean coefficient of conservatism
Mean coefficient of wetness
Summary of physiognomic classes for native and non-native species
Zoological Field Metrics
Bird species presence/composition
Number of birds
Density of birds
Species diversity
Partners in Flight conservation score - for each species and sample cell
Species guilds*
Aquatic Community Field Metrics
Mussel species presence/composition
Mussel species richness
Mussel species diversity (i.e., Shannon-Wiener score)
Mussel abundance
Presence/absence of exotic bivalves
Intensity of zebra mussel infestation on native mussels
Macroinvertebrate taxa composition - for each sample and sum for site
Macroinvertebrate taxa richness - for each sample and sum for site
Macroinvertebrate biotic indices (e.g., Hilsenhoff Biotic Index (HBI))
Mean tolerance value (MTV)
Relative abundance of benthic intolerant taxa
*Potential field metric
Table 9. Summary of botanical, zoological, and aquatic community field metrics used to assess thequality or condition of the sample cells in cover type patches across a gradient of quality based on theGIS patch analysis model.
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008 Page-17
the same sample cells or plots in which botanical andecological surveys were conducted in 2008. Breedingbirds were surveyed in the sample cells using pointcounts. In 2006, nine 10-minute point count surveyswere conducted in each sample cell. In 2007 and 2008,a different sampling protocol was used in that only one20-minute point count was conducted in each samplecell. The point count was randomly placed within thesample cell. Point counts began 15 minutes beforesunrise and continued until 1030 hours. No data werecollected during periods of inclement weather such asrain, fog, or a wind speed greater 20 km per hour.After arriving at the point count site, observersallowed a one-minute acclimation period beforeconducting a 20-minute point count. Although manyother studies have used a 5 – 10 minute point count,the 20-minute point count allowed us to observe moreavian diversity with the most efficiency given thedifficulties of traveling to each point count site (Huff etal. 2000). All birds observed (aurally and visually)were recorded within the 50-m fixed radius pointcount, however birds beyond this distance wererecorded as well with detailed information recorded ontheir distance from the observer (variable circular plotmethod). Gender of the birds was recorded wheneverpossible. Birds flying over the site were noted as suchand may not be used in further analysis. No playbacksor sounds were used to attract birds into the pointcount location.
The field metrics that were collected to assessthe quality or condition of the breeding bird communityinclude bird species composition, number of birds,density of birds, and species diversity (Table 9).Species guilds, indicator species, and/or conservationvalue of individual bird species also may be used toevaluate the bird community in the sample cells (Table9). We also can compare the quality of the cover typepatches based on the conservation value of theindividual species and the bird communities using thePartner’s in Flight (PIF) ranking system (Nuttle et al.2003). This classification allows bird species to beranked by their demography in many categories andthe level of conservation concern (Carter et al. 2000).Each bird species has been assigned PIF scoresincluding a population trend score, threats to breedingregional score, and regional combined score.Considering that a simple measure of the aviandiversity index of a site does not consider theconservation ranking of each species, the use of PIFranking system will add a component of conservationpriority.
Aquatic Field ComponentThe aquatic areas of high ecological value or
high site quality as identified by the field metrics willbe referred to as high quality aquatic sites or stream/river natural community classes. The aquatic fieldcomponent from 2006-2008 was conducted in theRogue River and Flat River in Kent, Montcalm, Ionia,and Newaygo counties. Aquatic field sampling from2006 to 2008 was conducted at 50 sites. To adequatelytest the GIS model developed by Wang et al. (in press)and the efficacy of the stream impact rankingsidentified in the model for accurately assessing thequality or condition of aquatic sites, the goal was toobtain field metrics from a sufficient number ofstream/river reaches across a gradient of impactrankings. The aquatic field sites that were sampledfrom 2006 to 2008 were fairly evenly distributedacross stream/river reaches of only three of the sixexpected impact rankings assigned by the GIS model,with 19 sites in stream reaches with the moderateimpact ranking category, 15 in the detectable impactranking, and 16 in the no impact ranking.
The field metrics to assess the quality of theaquatic site or stream/river natural community havefocused on the macroinvertebrate and musselcommunities found at the sample sites. Specific fieldmetrics that were examined and used to assess thequality or condition of aquatic sites have includedmussel species composition, mussel species richnessand diversity (i.e., Shannon-Wiener score), musselabundance, presence/absence of exotic bivalves,intensity of zebra mussel infestation on native mussels,macroinvertebrate taxa composition and diversity (i.e.,Shannon-Wiener score), macroinvertebrate taxarichness, macroinvertebrate biotic indices (e.g.,Hilsenhoff Biotic Index), macroinvertebrate meantolerance value (MTV), and relative abundance ofbenthic intolerant taxa (Table 9).
Survey sites were primarily determined byaccess (i.e., accessible road and/or bridge crossing).Survey sites in first order stream/river reaches wereomitted from the field portion of the study. A pool ofpotential survey sites was generated by identifyingpoints 100 m upstream and 100 m downstream of allroad-stream crossings. Survey sites were randomlyselected from this pool of potential sites, and werelocated either 100-m upstream or downstream fromselect road/stream intersections in the targeted
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008 Page-18
watershed. Some randomly selected sites were notsurveyed because of access problems (i.e., when wewere unable to contact the landowner for permission).At each sample site, we collected four Surber samplesand one D-net sample for a total of five samples forinvertebrates. D-net sweeps were made inmicrohabitats not covered by Surber samples,
especially near the water’s edge and emergent/submergent vegetation. We also surveyed for musselsin a 128-m2 area at each sample site. Freshwatermussel communities were sampled using glass bottombuckets and tactile searches. Substrate compositionalso was described by estimating the proportion ofeach particle size classes.
Aerial Photo Interpretation Component
In 2007, an aerial photo interpretation componentwas added to the project to improve our ability toaccurately identify areas to a particular land cover ornatural community type on the landscape and evaluatetheir quality remotely. The aerial photo interpretationcomponent delineated photo-interpreted (PI) polygonswhich are spatially defined areas of a single land covertype or class, as defined by the terrestrial GIS analysis,based on aerial photo interpretation of 1998 air photos.Polygons were delineated based on cover type,condition, roads, and other barriers or gaps in thevegetation. These polygons were delineatedindependently of the cover type patches identified bythe GIS model although significant overlap with thepatches likely occurred. Polygons comprised an entirecover type patch if it was small and contiguous. As aresult, each cover type patch identified in the GISmodel that was photo-interpreted could include a singlePI polygon or multiple PI polygons.
Each PI polygon was typed as one of twobroadly-defined MNFI natural community types:upland deciduous or mixed forest (e.g., dry-mesic and
mesic northern forest) and lowland mixed orconiferous forest (e.g., hardwood-conifer swamp andrich conifer swamp). Two additional naturalcommunity types, bog and intermittent wetland, alsowere used to classify PI polygons in 2007. Thesedesignations are based on a modified Anderson level 3or level 4 land cover type which are used for landcover mapping at MNFI. The modified Anderson levelclassifications are generally similar to MNFI naturalcommunity classifications but not exactly the same.Each PI polygon was evaluated for ecological qualitybased on a number of criteria related to size/area,landscape context, and ecological condition using theair photos (Table 10). Specific aerial photointerpretation metrics or criteria were developed fordifferent natural communities (Appendix 2). Based onthis analysis, this component identified high qualityphoto-interpreted polygons. These results can be usedto evaluate the GIS model results, and depending onthe results, this component may be used tocomplement or enhance the GIS model. The fieldmetrics also can be used to test and evaluate theaccuracy of the PI polygons.
Conservation Gazetteer
This project also will develop a “conservationgazetteer” which will identify high ecological valueareas as priorities for conservation in an ecoregionaland statewide framework based on currently availabledata for the state. This product will provideinformation and analysis that will enable informeddecisionmaking by federal, state, and localgovernments, non-governmental organizations, andother entities in management, land use andconservation planning decisions. The conservationgazetteer was initially proposed to be a printeddocument or book comprised of a series of mapsdepicting priority conservation areas by county. Thedata layers used in the analysis to identify the priorityconservation areas also would be provided withassociated background information or descriptions ofthe data layers including the data sources, limitations,and continuing needs. However, upon consultation withthe MDNR this year, it was decided that the
conservation gazetteer would consist of a web-basedproduct instead of a printed document. Theconservation gazetteer at this time will primarilyprovide the results of the GIS patch analysis andassociated data layers that were developed as part ofthe GIS component of this project. Additional datalayers that can be used to help identify and prioritizeareas of high ecological value for conservation alsomay be provided as part of the conservation gazetteer.These may include important areas for rare speciesand high quality natural communities based onMichigan’s Natural Heritage Database, areas thathave been designated as important for birds, otherwildlife, and biodiversity in general (e.g., important birdareas, old growth forests, GAP analysis, etc.), areasimportant for watershed function or watershedprotection areas, proximity to public lands, and landsunder threat (i.e., threat analysis).
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008 Page-19
Ran
k fa
cto
rK
ey E
colo
gic
al
Att
rib
ute
Met
ric
Ran
kin
g c
rite
ria
LANDSCAPE CONTEXT
Land
scap
e P
atte
rn [a
sses
sed
@ 1
:24,
000
scal
e ce
nter
ed o
n po
lygo
n of
inte
rest
]B
road
er la
nd u
se
4 =
Exc
elle
nt
Land
scap
e ch
arac
teriz
ed b
y >9
0% n
atur
al c
over
(i.e
., cl
assi
fiabl
e to
nat
ural
com
mun
ity ty
pe),
with
the
sum
tota
l of
bui
lt ar
eas,
act
ive
agric
ultu
re, o
ld fi
elds
, and
pin
e pl
anta
tions
com
pris
ing
<10%
of a
rea.
3 =
Go
od
Land
scap
e ch
arac
teriz
ed b
y >6
0% n
atur
al c
over
, with
str
ong
conn
ectiv
ity a
mon
g pa
tche
s. A
ctiv
e ag
ricul
tura
l fie
lds,
old
fiel
ds, a
nd/o
r pi
ne p
lant
atio
ns c
ompr
ise
0-40
% o
f lan
dsca
pe.
Bui
lt ar
eas
cann
ot e
xcee
d 25
%.
2 =
Fai
rLa
ndsc
ape
char
acte
rized
by
30-6
0% n
atur
al c
over
, eith
er a
s re
lativ
ely
few
isol
ated
blo
cks
or a
s sc
atte
red
patc
hes
with
poo
r co
nnec
tivity
. T
he r
emai
nder
of v
ario
us ty
pes
(see
abo
ve).
Bui
lt ar
eas
cann
ot e
xcee
d 50
%.
1 =
Po
or
Land
scap
e ur
ban
or s
ubur
ban,
or
dom
inat
ed b
y ac
tive
agric
ultu
re.
<30%
of l
ands
cape
cha
ract
eriz
ed b
y na
tura
l co
ver.
Land
scap
e C
ompo
sitio
nB
uffe
r w
idth
4 =
Exc
elle
nt
Occ
urre
nce
is b
uffe
red
by +
/- n
atur
al c
over
of >
100
m w
idth
on
entir
e pe
riphe
ry.
3 =
Go
od
Occ
urre
nce
is b
uffe
red
by +
/- n
atur
al c
over
of >
100
m w
idth
on
>75%
of p
erip
hery
. U
nbuf
fere
d po
rtio
n no
t in
activ
e ag
ricul
ture
or
built
.
2 =
Fai
rO
ccur
renc
e is
buf
fere
d by
+/-
nat
ural
cov
er o
f >10
0 m
wid
th o
n >5
0% o
f per
iphe
ry o
r by
>50
m w
idth
on
>75%
of
perip
hery
. U
nbuf
fere
d po
rtio
n ca
n be
in a
ctiv
e ag
ricul
ture
or
built
. Inc
lude
s si
tes
buffe
red
by p
ine
plan
tatio
n.
1 =
Po
or
Occ
urre
nce
buffe
red
by +
/- n
atur
al c
over
of <
50 m
wid
th o
n en
tire
perip
hery
, or
not b
uffe
red.
ABIOTIC CONDITION
Hyd
rolo
gy
Pre
senc
e of
dr
ains
, ditc
hes,
ch
anne
lizat
ion,
ro
ads
4 =
Exc
elle
nt
Hyd
rolo
gy u
nim
pact
ed b
y dr
ains
, ditc
hes,
roa
ds, o
r ch
anne
ls.
No
anth
ropo
geni
c hy
drol
ogic
man
ipul
atio
n no
ted.
3 =
Go
od
Min
or d
isru
ptio
n of
hyd
rolo
gic
regi
me.
Dra
ins,
ditc
hes,
cha
nnel
s, o
r ro
ads,
if p
rese
nt, i
mpa
ct <
10%
of w
etla
nd.
Or,
for
larg
e w
etla
nd c
ompl
exes
, hyd
rolo
gy is
impa
cted
>10
0 m
from
per
iphe
ry o
f are
a id
entif
ied
as b
eing
of
pote
ntia
l nat
ural
qua
lity.
Incl
udes
are
as a
ssoc
iate
d w
ith o
ther
wis
e in
tact
str
eam
s cr
osse
d by
roa
ds.
2 =
Fai
r
Mod
erat
e di
srup
tion
of h
ydro
logi
c re
gim
e. D
rain
s, d
itche
s, c
hann
els,
or
road
s im
pact
<25
% o
f occ
urre
nce.
Or,
fo
r la
rge
wet
land
com
plex
es, h
ydro
logy
is im
pact
ed <
100
m fr
om p
erip
hery
of a
rea
iden
tifie
d as
bei
ng o
f pot
entia
l na
tura
l qua
lity.
1 =
Po
or
Hyd
rolo
gic
regi
me
sign
ifica
ntly
alte
red.
Dra
ins,
ditc
hes,
cha
nnel
s, o
r ro
ads
impa
ct >
25%
of o
ccur
renc
e.
Site
Inta
ctne
ss
With
in-p
atch
fr
agm
enta
tion
and
dist
urba
nce,
ca
used
by
logg
ing,
plo
win
g,
hum
an
deve
lopm
ent,
etc.
4 =
Exc
elle
nt
<10%
of p
atch
impa
cted
by
frag
men
tatio
n an
d/or
hum
an d
istu
rban
ce.
Logg
ing
road
s, tw
o-tr
acks
, and
trai
ls
visi
ble
on a
eria
l pho
togr
aphs
are
loca
lized
. R
ecen
t cut
ting,
oil
pad
deve
lopm
ent,
hum
an-c
reat
ed s
oil d
istu
rban
ce,
etc.
loca
l.
3 =
Go
od
<25%
of p
atch
impa
cted
by
frag
men
tatio
n an
d/or
hum
an d
istu
rban
ce.
Logg
ing
road
s, tw
o-tr
acks
, and
trai
ls
visi
ble
on a
eria
l pho
togr
aphs
loca
l to
occa
sion
al.
Rec
ent c
uttin
g, o
il pa
d de
velo
pmen
t, hu
man
-cre
ated
soi
l di
stur
banc
e, e
tc. l
ocal
to o
ccas
iona
l.
2 =
Fai
r
<40%
of p
atch
impa
cted
by
frag
men
tatio
n an
d/or
hum
an d
istu
rban
ce.
Logg
ing
road
s, tw
o-tr
acks
, and
trai
ls
visi
ble
on a
eria
l pho
togr
aphs
occ
asio
nal t
o co
mm
on.
Rec
ent c
uttin
g, o
il pa
d de
velo
pmen
t, hu
man
-cre
ated
soi
l di
stur
banc
e, e
tc. o
ccas
iona
l to
com
mon
.
1 =
Po
or
Occ
urre
nce
is s
igni
fican
tly im
pact
ed b
y fr
agm
enta
tion
and/
or h
uman
dis
turb
ance
(es
t. >4
0% o
f are
a im
pact
ed).
R
ecen
t cut
ting,
oil
pad
deve
lopm
ent,
hum
an-c
reat
ed s
oil d
istu
rban
ce, e
tc. c
omm
on to
wid
espr
ead.
Tabl
e 10.
Sum
mar
y of
rank
fact
ors,
key
ecol
ogic
al a
ttrib
utes
, met
rics
, and
rank
ing
crite
ria
for a
sses
sing
the e
colo
gica
l qua
lity
orco
nditi
on of
pho
to-in
terp
rete
d po
lygo
ns in
gene
ral.
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008 Page-20
BIOTIC CONDITION
Com
mun
ity S
truc
ture
and
F
unct
ion
Veg
etat
ive
stru
ctur
e
4 =
Exc
elle
nt
Spe
cific
to n
atur
al c
omm
unity
type
3 =
Go
od
Spe
cific
to n
atur
al c
omm
unity
type
2 =
Fai
rS
peci
fic to
nat
ural
com
mun
ity ty
pe
1 =
Po
or
Spe
cific
to n
atur
al c
omm
unity
type
SIZE
Com
mun
ity s
ize
Are
a in
acr
es
4 =
Exc
elle
nt
Spe
cific
to n
atur
al c
omm
unity
type
, bas
ed o
n ex
istin
g M
NF
I EO
spe
cific
atio
ns.
3 =
Go
od
Spe
cific
to n
atur
al c
omm
unity
type
, bas
ed o
n ex
istin
g M
NF
I EO
spe
cific
atio
ns.
2 =
Fai
rS
peci
fic to
nat
ural
com
mun
ity ty
pe, b
ased
on
exis
ting
MN
FI E
O s
peci
ficat
ions
.
1 =
Po
or
Spe
cific
to n
atur
al c
omm
unity
type
, bas
ed o
n ex
istin
g M
NF
I EO
spe
cific
atio
ns.
Tabl
e 10.
Con
tinue
d.
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008 Page-21
The GIS patch analysis modeling has beencompleted except for testing the model against thefield metrics and aerial photo interpretation results. Atotal of 2,181,975 patches of land cover was delineatedand included in the analysis. When scores for all 13regions of the state were summed together, less than5% of the patches were assigned a score of seven orhigher. Only 5,751 patches of the total 2,181,975patches included in the analysis scored ten or higher(i.e., 0.3%) (Figure 3). Figure 4 provides an example
RESULTS AND DISCUSSION
GIS Analysis/Component
of the patch analysis results. Generally, both thenumber of high scoring patches and the size of highscoring patches increased with latitude (Figures 5 and6). Also, the cover type patches, especially the patcheshaving high quality scores, were generally very large.Sampling only one 10-ha sample cell or plot in verylarge cover type patches may need to be examined forthe analysis. The data analysis will test whether theGIS patch scores correlate with the field metricscores.
Figure 3. Distribution of quality scores for all cover type patches from GIS patch analysis.
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008 Page-22
Highest ecological value
Lowest ecological value
Figure 4. Example of GIS patch analysis results from Newaygo County.
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008 Page-23
High value patches
Moderate value patches
Figure 5. Map showing general distribution of high and moderately scoring cover type patches inthe Lower Peninsula based on results from the GIS patch analysis.
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008 Page-24
High value patches
Moderate value patches
Figure 6. Map showing general distribution of high and moderately scoring cover type patches inthe Upper Peninsula based on results from the GIS patch analysis.
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008 Page-25
Field Component
Terrestrial Field ComponentIn 2008, ecological, botanical, and zoological field
metrics or sampling were conducted in 10 sample cellsin upland deciduous forest cover type patches rankedas moderate to high quality (UDM) and 10 samplecells in upland deciduous forest cover type patchesranked as low to moderate quality (UDL) based on theGIS patch analysis scores. The same sample cellswere surveyed by all three disciplines. Data fromthese sample cells will result in field metrics data from2006 to 2008 from at least 30 upland deciduous forestcover type patches across a gradient of quality, with atleast 10 sample cells/cover type patches in each of thethree quality rankings (i.e., high, moderate, and lowquality).
Also, the field sampling scheme was revised in2008 so that cover type patches were randomlyselected for field sampling instead of randomlyselecting sample cells, and that each sample cell waslocated within a single cover type patch instead ofmultiple cover type patches potentially as in 2006 and2007. For data collected in 2006 and 2007, if thesample cells fall within a single cover type patch, thedata will be used in the current analysis. If the samplecell crosses multiple patch boundaries, the data will becensored and not included in the current analysis atthis time. We should be able to keep field data from anumber of upland forest sample cells from 2006 and2007, but this needs to be verified. This will result in aGIS score and a field metric score or scores for eachsample cell and cover type patch, rather than onesample cell/field metric score to multiple GIS patches/patch scores.
Ecological field surveys/samplingThe majority of sample cells or plots that were
surveyed in 2008 were characterized by dry-mesicnorthern forest, dominated by white oak (Quercusalba) and black oak (Q. velutina), often associatedwith white pine (Pinus strobus). Additional plots werecharacterized by aspen (Populus grandidentata andP. tremuloides), mesic northern forest, and oak-pinebarrens. Several field metrics assessing landscapecontext, present condition, and threats were scored inthe field based on meander surveys of the entire plot.GPS waypoints were taken to record locations ofsignificant features, disturbances, changes in covertype, and soil samples. All but one plot fell within coverpatches exhibiting significantly altered vegetativestructure associated with timber removal and otherhabitat management. High deer browse and an
abundance of red maple were identified as significantthreats to the majority of sampled forest patches.Invasive species were uncommon and concentrated onroadsides, along well-used paths, and in clearings.Field scores were similar for habitat patches predictedas being of medium or low quality in the GIS-basedpatch analysis. Further data analysis will be conductedin winter/spring 2009. Part of the data analysis will bedetermining if any of the field metrics or variablescorrelate with one another and how to aggregate thefield metric scores to develop a cumulative score foreach sample cell.
Botanical field surveys/floristic samplingPlant species presence/composition was
recorded during meander/timed-meander surveys in alltargeted sample cells or plots in 2008. No listed plantspecies were encountered during plot inventories.Several sites noted during aerial photo review aspotential coastal plain marshes (in proximity to samplecells/plots) were subsequently visited in the field enroute to sample cells/plots (and in one case, during arainy day when it was not possible to use the handheldcomputer for plot sampling). Several high qualitycoastal plain marshes were thus identified, includingseveral rare plant species, although there was notsufficient time to conduct thorough surveys or collectdata adequate for formal processing. One statethreatened species, Eleocharis tricostata, wasidentified and collected in coastal plain marsh habitatwithin a powerline right-of-way between UDM 8 andUDM 9 in northern Newaygo County, constituting asignificant record (data were obtained for this site). Atotal of 103 plant specimens were collected during thefield sampling. These specimens have been identifiedand verified. Data entry and analysis will be conductedin the winter/spring of 2009. As with the ecologicaldata, part of the data analysis will be determining howto aggregate the field metric scores to generate acumulative score for the plant community for eachsample cell.
Zoological field samplingBreeding bird point count surveys were
conducted in May and June of 2008. These surveyswere conducted in the same sample cells or plots inwhich botanical and ecological surveys wereconducted in 2008. Several rare or noteworthy birdspecies that were detected during breeding birdsurveys in 2008 include the state special concernCerulean Warbler (Dendroica cerulea), statethreatened Red-shouldered Hawk (Buteo lineatus),
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008 Page-26
Golden-winged Warbler (Vermivora chrysoptera),and Bell’s Vireo (Vireo bellii). The Cerulean Warbler,Red-shouldered Hawk, and Golden-winged Warbleralso have been identified as species of greatestconservation need by Michigan’s Wildlife Action Plan(Eagle et al. 2005). The breeding bird data will beentered and analyzed during the winter/spring of 2009.Part of the data analysis will need to determine how todeal with the different bird sampling methodologyemployed in 2007 and 2008 compared to that used in2006. The exact field metrics that will be utilized in theanalysis and how to aggregate or combine the fieldmetric scores into a cumulative score for the birdcommunity for each sample cell also will need to bedetermined.
Aquatic Field ComponentAquatic field sampling was conducted primarily
in 2006 and 2007. Limited aquatic field sampling wasconducted in 2008. The aquatic component of theproject in 2008 focused primarily on processing thenumerous macroinvertebrate samples that werecollected during field sampling from 2006 to 2008. Five
macroinvertebrate samples were collected in the fieldfrom each site, but only a subset of the samples will beprocessed, identified and included in the data analysisat this time due to time and funding constraints. Foursamples from 16 sites (i.e., 64 samples) have beenprocessed and identified. Two samples from anadditional 14 sites (i.e., 28 samples) also have beenprocessed and identified to date. This will result in 2-4samples from 30 sites total, with 10 sites in each of thethree predicted impact rankings, that will be processed,identified, and included in the data analysis at this time.Aquatic data processing/entry and analysis will beconducted in 2009. As with the terrestrial field metrics,part of the aquatic data analysis will be determininghow to aggregate or combine field metrics into acumulative score for each site to compare with thepredicted stream impact rankings for each site fromthe GIS model. The field metric data will be used tosee if they can be correlated to the predicted stream/river impact levels or rankings to test whether theseimpact rankings and the GIS model can be used toinfer or assess the ecological quality or condition ofstream/river reaches remotely.
Aerial Photo Interpretation Component
In 2007, all upland forests 10 ha or greater in 8of the 24 townships in Newaygo County were photointerpreted. Only upland forest and lowland mixedforest polygons meeting C-rank size criteria in theseeight townships were identified and photo interpreted.In addition, potential high quality occurrences of theseand other natural communities were identified anddelineated in the remaining 16 townships in NewaygoCounty to provide a more complete conservation layerof the county. A total of 833 sites were identified anddelineated throughout the county. In 2008, no additionalaerial photo interpretation or identification of additionalphoto-interpreted polygons was conducted other thanan aerial photo review of the sample cells that wereselected for sampling.
In 2007, all upland forest sample cells or plotswere located in photo-interpreted polygons. In 2008,two of the 20 sampled cells or plots were located inphoto-interpreted polygons. Results from the aerial
photo interpretation and evaluation of polygon qualityconducted in 2007 will be compiled and potentiallyanalyzed during winter/spring of 2009. Each polygonhas a score for each criterion, but we still need todetermine the specific criteria that would be includedin the analysis and develop an algorithm or formula forcombining or aggregating the criteria into one score/ranking indicating the quality or condition of eachphoto-interpreted polygon. Also, additional photointerpretation would need to be conducted to delineateand evaluate the polygons in which 18 of the 20 cellssampled in 2008 were located if the PI polygons aregoing to be included in the data analysis. The photo-interpreted polygon scores can be compared with theGIS patch analysis scores and field metric scores toevaluate the efficacy and contribution of this approachto assessing the ecological quality of these areasremotely. The photo-interpreted polygons couldpotentially be used to enhance or complement the GISmodels.
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008 Page-27
Conservation GazetteerDevelopment of the web-based conservation
gazetteer continued in 2008. The overall architectureof the gazetteer has been established. After substantialexperimentation with displaying different data types,the format that will be used to present the data hasbeen determined. This will entail converting the patchanalysis models from the ESRI raster format to avector format from which the models will be converted
to a different raster format. We are currentlyinvestigating different ways for displaying the GISmodels and resulting data layers in a user-friendlyformat. Development of the web-based conservationgazetteer will be completed in 2009. The web-basedconservation gazetteer also will be updated and revisedbased on feedback from initial users.
REFERENCESCarter, M., W. Hunter, D. Pashley, and K. Rosenberg.
2000. Setting conservation priorities forlandbirds in the United States: The Partners inFlight Approach. Auk 117:541-548.
Eagle, A. C., E. M. Hay-Chmielewski, K. T.Cleveland, A. L. Derosier, M. E. Herbert, andR. A. Rustem, eds. 2005. Michigan’s WildlifeAction Plan. Michigan Department of NaturalResources. Lansing, Michigan. 1592 pp. http://www.michigan.gov/dnrwildlifeactionplan
Herman, K. D., L. A. Masters, M. R. Penskar, A. A.Reznicek, G. S. Wilhelm, W. W. Brodovich, andK. P. Gardiner. 2001. Floristic QualityAssessment with Wetland Categories andExamples of Computer Applications for theState of Michigan – Revised, 2nd Edition.Michigan Department of Natural Resources,Wildlife, Natural Heritage Program. Lansing,MI. 19 pp. + Appendicies.
Huff, M., K. Bettinger, H. Ferguson, M. Brown, andB. Altman. 2000. A habitat-based point-countprotocol for terrestrial birds, emphasizingWashington and Oregon. U. S. Department ofAgriculture, Forest Service General TechnicalReport PNW-GTR-501.
Lindenmayer, D. B., C. R. Margules, and D. B.Botkin. 2000. Indicators of biodiversity forecologically sustainable forest management.Conservation Biology 14: 941-950.
Lindenmayer, D. B., J. F. Franklin, and J. Fischer.2006. General management principles and achecklist of strategies to guide forest biodiversityconservation. Biological Conservation 131: 433-445.
MacArthur, R. and J. MacArthur. 1961. On birdspecies diversity. Ecology 42:594-598.
Maurer, B. 1993. Biological diversity, ecologicalintegrity, and neotropical migrants: Newperspectives for wildlife management. Pages24-31 in Status and Management of NeotropicalMigratory Birds (D. Finch and P. Stangel, Eds.).U.S. Department of Agriculture, Forest ServiceGeneral Technical Report RM-229.
McElhinny, C., Gibbons, P., Brack, C. and J. Bauhus.2005. Forest and woodland stand structuralcomplexity: Its definition and management.Forest Ecology and Management 218: 1-24.
Michigan Department of Natural Resources (MDNR).2001. Integrated forest monitoring assessmentand prescription (IFMAP) / GAP LowerPeninsula and Upper Peninsula land cover.Remote sensing image. Content 1997-2001.
Noss, R. F. 1990. Indicators for monitoringbiodiversity: A hierarchical approach.Conservation Biology 4: 355-364.
Noss, R. F. 1999. Assessing and monitoring forestbiodiversity: A suggested framework andindicators. Forest Ecology and Management 115:135-146.
Nuttle, T., A. Leidolf, and L. Burger. Jr. 2003.Assessing conservation value of birdcommunities with partners in flight-based ranks.Auk 120:541-549.
Rappole, J. and M. McDonald. 1994. Cause andeffect in population declines of migratory birds.Auk 111:652-660.
Wang, L., T. Brenden, P. W. Seelbach, A. Cooper, D.Allan, R. Clark, Jr., and M. Wiley. In press.Landscape based identification of humandisturbance gradients and references forstreams in Michigan. Environmental Monitoringand Assessment.
Statewide Analysis and Surveys to Identify Priority Conservation Areas 2008 Page-28
ACKNOWLEDGEMENTS
We would like to gratefully acknowledge theMichigan Department of Natural Resources (MDNR)Wildlife Division for providing the funding for thisproject. We would like to especially acknowledge Dr.Patrick Lederle, Michael Donovan, and SarahMayhew with the MDNR Wildlife Division for theirguidance and assistance with this project. We alsowould like to thank Dr. Gary Roloff with the MSUDepartment of Fisheries and Wildlife for providingassistance with the study design. We also would like toacknowledge Dr. Patrick Brown, former MNFIDirector, for his efforts initiating and leading thisproject from 2006 to 2007. Several MNFI staff alsoprovided assistance with this project. Kraig Korroch,
MNFI Information Technologist, provided invaluableassistance with the development of the web-basedconservation gazetteer. Beverly Walters, former MNFIBotanist, conducted the botanical sampling in 2006 and2007. Helen Enander, MNFI GIS Analyst, providedtechnical assistance with the GIS components of thestudy. We would especially like to acknowledgeRebecca Rogers, MNFI Information Technologist,Kraig Korroch, and David Cuthrell, MNFI Zoologist,for their invaluable assistance with the processing,sorting, picking and identification of the aquaticmacroinvertebrate samples. Finally, Nancy Toben, SueRidge, and Connie Brinson with MNFI providedadministrative assistance on this project.
A-1
APPENDICES
A-2
Appendix I. Summary of field metrics to assess the ecological quality or condition of the range of naturalcommunity field classes associated with the GIS/IFMAP land cover types or classes and MNFI natural commu-nities.
Bo
g
Po
or
fen
Po
or
con
ifer
sw
amp
Mu
skeg
Co
asta
l p
lain
mar
sh
Inte
rmit
ten
t w
etla
nd
Metrics
Landscape context**
Buffer width X X X X X X
Broader land use X X X X X X
Site intactness
Presence of roads, RR tracks, or other (residential,
industrial, commercial) development
X X X X X X
Hydrology
Presence of drains, ditches, channelization X X X X X X
Soil erosion
ORV damage X X X X X X
Stream morphology
Community structure and function
Vegetative structure and composition X X X X X X
Evidence of logging
Evidence of plowing
Evidence of fire
Presence of invasive non-native vascular plant spp. X X X X X X
Presence of rotting logs and snags
Tree regeneration (oak, conifer, etc. as appropriate)
Evidence of deer browse
Species-specific die-offs
Condition of sphagnum peat X X X X
Cat-tail invasion (indicator of nutrient enrichment) X X X X X X
Presence of red maple (Acer rubrum )
Pit-and-mound topography
Dominance by native grasses and/or sedges
Non-forested wetland Lowland forest Non-forested wetland
Acidic peatlands Intermittent wetlands
A-3
Appendix I (continued). Summary of field metrics to assess the ecological quality or condition of the rangeof natural community field classes associated with the GIS/IFMAP land cover types or classes and MNFInatural communities.
Metrics
Landscape context**
Buffer width
Broader land use
Site intactness
Presence of roads, RR tracks, or other (residential,
industrial, commercial) development
Hydrology
Presence of drains, ditches, channelization
Soil erosion
ORV damage
Stream morphology
Community structure and function
Vegetative structure and composition
Evidence of logging
Evidence of plowing
Evidence of fire
Presence of invasive non-native vascular plant spp.
Presence of rotting logs and snags
Tree regeneration (oak, conifer, etc. as appropriate)
Evidence of deer browse
Species-specific die-offs
Condition of sphagnum peat
Cat-tail invasion (indicator of nutrient enrichment)
Presence of red maple (Acer rubrum )
Pit-and-mound topography
Dominance by native grasses and/or sedges
Dry
San
d P
rair
ie
Mes
ic s
and
pra
irie
Oak
bar
ren
s
Oak
op
enin
gs
Oak
-pin
e b
arre
ns
Pin
e b
arre
ns
X
X X X X X X
X X X X X X
X X X X X X
X X X X X X
X X X X X X
X X X X X X
X X X X X X
X X X X X X
X X X X X X
X X X X X X
X X X X X X
Grassland/shrub
Upland prairies and savanna
A-4
Appendix I (continued). Summary of field metrics to assess the ecological quality or condition of the range ofnatural community field classes associated with the GIS/IFMAP land cover types or classes and MNFI naturalcommunities.
Metrics
Landscape context**
Buffer width
Broader land use
Site intactness
Presence of roads, RR tracks, or other (residential,
industrial, commercial) development
Hydrology
Presence of drains, ditches, channelization
Soil erosion
ORV damage
Stream morphology
Community structure and function
Vegetative structure and composition
Evidence of logging
Evidence of plowing
Evidence of fire
Presence of invasive non-native vascular plant spp.
Presence of rotting logs and snags
Tree regeneration (oak, conifer, etc. as appropriate)
Evidence of deer browse
Species-specific die-offs
Condition of sphagnum peat
Cat-tail invasion (indicator of nutrient enrichment)
Presence of red maple (Acer rubrum )
Pit-and-mound topography
Dominance by native grasses and/or sedges
Dry
so
uth
ern
fo
rest
Dry
-mes
ic n
ort
her
n f
ore
st
Dry
-mes
ic s
ou
ther
n f
ore
st
Mes
ic n
ort
her
n f
ore
st
Mes
ic s
ou
ther
n f
ore
st
X X X X X
X X X X X
X X X X X
X X X X X
X X X X X
X X X X X
X X X X X
X X X
X X X X X
X X X X X
X X X
X X X X X
X X X X X
X X X
X X
Upland forest
Upland oak and oak-pine
forests
Upland mesic
forests
A-5
Appendix I (continued). Summary of field metrics to assess the ecological quality or condition of the rangeof natural community field classes associated with the GIS/IFMAP land cover types or classes and MNFInatural communities.
Metrics
Landscape context**
Buffer width
Broader land use
Site intactness
Presence of roads, RR tracks, or other (residential,
industrial, commercial) development
Hydrology
Presence of drains, ditches, channelization
Soil erosion
ORV damage
Stream morphology
Community structure and function
Vegetative structure and composition
Evidence of logging
Evidence of plowing
Evidence of fire
Presence of invasive non-native vascular plant spp.
Presence of rotting logs and snags
Tree regeneration (oak, conifer, etc. as appropriate)
Evidence of deer browse
Species-specific die-offs
Condition of sphagnum peat
Cat-tail invasion (indicator of nutrient enrichment)
Presence of red maple (Acer rubrum )
Pit-and-mound topography
Dominance by native grasses and/or sedges
Har
dw
oo
d-c
on
ifer
sw
amp
Rel
ict
con
ifer
sw
amp
Ric
h c
on
ifer
sw
amp
So
uth
ern
sw
amp
No
rth
ern
sw
amp
So
uth
ern
flo
od
pla
in f
ore
st
Floodplain
forest
X X X X X X
X X X X X X
X X X X X X
X X X X X
X
X X X X X X
X X X X X
X X X X X X
X X X X X X
X X X X X
X X X X X X
X X X X X X
X X X
Lowland forest
Lowland mixed and
coniferous forests
Lowland
deciduous forests
A-6
Appendix I (continued). Summary of field metrics to assess the ecological quality or condition of the rangeof natural community field classes associated with the GIS/IFMAP land cover types or classes and MNFInatural communities.
Met
rics
La
nd
sca
pe
con
tex
t**
Bu
ffer
wid
th
Bro
ader
lan
d u
se
Sit
e in
tact
nes
s
Pre
sen
ce o
f ro
ads,
RR
tra
cks,
or
oth
er (
resi
den
tial
,
ind
ust
rial
, co
mm
erci
al)
dev
elo
pm
ent
Hyd
rolo
gy
Pre
sen
ce o
f d
rain
s, d
itch
es,
chan
nel
izat
ion
So
il e
rosi
on
OR
V d
amag
e
Str
eam
mo
rph
olo
gy
Co
mm
un
ity
stru
ctu
re a
nd
fu
nct
ion
Veg
etat
ive
stru
ctu
re a
nd
co
mp
osi
tio
n
Ev
iden
ce o
f lo
gg
ing
Ev
iden
ce o
f p
low
ing
Ev
iden
ce o
f fi
re
Pre
sen
ce o
f in
vas
ive
no
n-n
ativ
e v
ascu
lar
pla
nt
spp
.
Pre
sen
ce o
f ro
ttin
g l
og
s an
d s
nag
s
Tre
e re
gen
erat
ion
(o
ak,
con
ifer
, et
c. a
s ap
pro
pri
ate)
Ev
iden
ce o
f d
eer
bro
wse
Sp
ecie
s-sp
ecif
ic d
ie-o
ffs
Co
nd
itio
n o
f sp
hag
nu
m p
eat
Cat
-tai
l in
vas
ion
(in
dic
ato
r o
f n
utr
ien
t en
rich
men
t)
Pre
sen
ce o
f re
d m
aple
(Ace
r ru
bru
m)
Pit
-an
d-m
ou
nd
to
po
gra
ph
y
Do
min
ance
by
nat
ive
gra
sses
an
d/o
r se
dg
es
Inundated Shrub Swamp
Northern Shrub Thicket
Southern shrub-carr
Emergent marsh
Inland salt marsh
Submergent marsh
Lakeplain wet-mesic prairie
Northern fen
Northern wet meadow
Northern Wet-Mesic Prairie
Southern wet meadow
Wet prairie
Wet-mesic prairie
XX
XX
XX
XX
XX
XX
X
XX
XX
XX
XX
XX
XX
X
XX
XX
XX
XX
XX
XX
X
XX
XX
XX
XX
XX
XX
X
XX
XX
XX
X
XX
XX
XX
XX
XX
XX
X
XX
XX
XX
X
XX
XX
XX
XX
XX
XX
X
XX
XNo
n-f
ore
sted
wet
lan
d
Wet
mea
do
ws,
wet
an
d w
et-m
esic
pra
irie
s, a
nd
ca
lca
reo
us
fen
Sh
rub
-do
min
ate
d w
etla
nd
sE
mer
gen
t a
nd
su
bm
erg
ent
ma
rsh
es
A-7
Appendix II. Summary of metrics and criteria used to evaluate the quality or condition of photo-inter-preted polygons by natural community type.
CO
MM
UN
ITY
TY
PE
Ran
k fa
cto
rK
ey E
colo
gic
al
Att
rib
ute
Met
ric
Ran
kin
g c
rite
ria
BOG
BIOTIC CONDITIONC
omm
unity
Str
uctu
re a
nd
Fun
ctio
n
Veg
etat
ive
stru
ctur
e an
d co
mpo
sitio
n
4 =
Exc
elle
nt
N/A
. R
equi
res
field
sur
vey.
3 =
Go
od
Veg
etat
ion
with
in e
xpec
ted
rang
e of
var
iatio
n. S
phag
num
mat
inta
ct.
Zon
atio
n va
rious
de
pend
ing
on m
atur
ity.
2 =
Fai
rP
artia
lly d
egra
ded
bog
mat
(i.e
., a
port
ion
of th
e sy
stem
is in
tact
, ano
ther
por
tion
is
degr
aded
).
1 =
Po
or
Veg
etat
ion
outs
ide
expe
cted
ran
ge o
f var
iatio
n. O
R S
phag
num
mat
deg
rade
d. O
R fl
oode
d (in
clud
ing
beav
er fl
oodi
ng)
form
er b
og n
ow la
ckin
g ty
pica
l str
uctu
re.
SIZE
Com
mun
ity s
ize
Are
a in
acr
es
4 =
Exc
elle
nt
70+
acre
s
3 =
Go
od
40-6
9 ac
res
2 =
Fai
r10
-39
acre
s
1 =
Po
or
<10
acre
s
POOR CONIFER SWAMP
BIOTIC CONDITION
Com
mun
ity S
truc
ture
and
F
unct
ion
Veg
etat
ive
stru
ctur
e an
d co
mpo
sitio
n
4 =
Exc
elle
nt
N/A
. R
equi
res
field
sur
vey.
3 =
Go
od
Veg
etat
ion
with
in e
xpec
ted
rang
e of
var
iatio
n. S
phag
num
mat
inta
ct.
Zon
atio
n va
rious
de
pend
ing
on m
atur
ity.
2 =
Fai
rN
/A.
Req
uire
s fie
ld s
urve
y.
1 =
Po
or
Veg
etat
ion
outs
ide
expe
cted
ran
ge o
f var
iatio
n. O
R S
phag
num
mat
deg
rade
d. O
R fl
oode
d (in
clud
ing
beav
er fl
oodi
ng),
now
lack
ing
typi
cal s
truc
ture
.
SIZE
Com
mun
ity s
ize
Are
a in
acr
es
4 =
Exc
elle
nt
120+
acr
es
3 =
Go
od
60-1
19 a
cres
2 =
Fai
r30
-59
acre
s
1 =
Po
or
<30
acre
s
A-8
Appendix II (continued). Summary of metrics and criteria used to evaluate the quality or condition ofphoto-interpreted polygons by natural community type.
RICH CONIFER SWAMP
BIOTIC CONDITION
Com
mun
ity S
truc
ture
and
F
unct
ion
Veg
etat
ive
stru
ctur
e an
d co
mpo
sitio
n
4 =
Exc
elle
nt
N/A
. Old
-gro
wth
site
s ap
pear
to h
ave
rela
tivel
y la
rge
crow
n di
amet
ers.
May
be
scor
ed if
we
use
circ
a 19
38 a
eria
l pho
tos.
3 =
Go
od
Con
ifers
dom
inan
t (>6
6% c
over
). S
houl
d be
mat
ure,
but
pla
ce h
ere
if un
cert
ain.
2 =
Fai
rC
onife
rs >
50%
cov
er, o
r ce
dars
imm
atur
e, if
dis
cern
ible
on
aeria
l pho
tos.
1 =
Po
or
Har
dwoo
ds d
omin
ant (
>50%
cov
er).
Typ
ical
ly h
eavi
ly d
istu
rbed
sta
nds
that
hav
e co
nver
ted
to r
ed m
aple
dom
inan
ce.
SIZE
Com
mun
ity s
ize
Are
a in
acr
es
4 =
Exc
elle
nt
120+
acr
es
3 =
Go
od
60-1
19 a
cres
2 =
Fai
r30
-59
acre
s
1 =
Po
or
<30
acre
s
RELICT CONIFER SWAMP
BIOTIC CONDITION
Com
mun
ity S
truc
ture
and
F
unct
ion
Veg
etat
ive
stru
ctur
e an
d co
mpo
sitio
n
4 =
Exc
elle
nt
N/A
. R
equi
res
field
sur
vey.
3 =
Go
od
Con
ifers
(ta
mar
ack
and
othe
rs o
ccas
iona
lly)
>60%
cov
er.
Dec
iduo
us tr
ees
<33%
cov
er.
Zon
atio
n va
rious
.
2 =
Fai
rC
onife
rs 3
3-60
% c
over
. Z
onat
ion
vario
us.
Rep
rese
nts
site
s w
ith s
ome
degr
ee o
f co
nver
sion
to h
ardw
ood
(typ
ical
ly r
ed m
aple
) do
min
ance
.
1 =
Po
or
Con
ifers
<33
% c
over
. T
ypic
ally
hea
vily
dis
turb
ed s
tand
s th
at h
ave
conv
erte
d to
red
map
le
dom
inan
ce.
SIZE
Com
mun
ity s
ize
Are
a in
acr
es
4 =
Exc
elle
nt
70+
acre
s
3 =
Go
od
45-6
9 ac
res
2 =
Fai
r15
-45
acre
s
1 =
Po
or
<15
acre
s
A-9
Appendix II (continued). Summary of metrics and criteria used to evaluate the quality or condition ofphoto-interpreted polygons by natural community type.
HARDWOOD-CONIFER SWAMP
BIOTIC CONDITION
Com
mun
ity S
truc
ture
and
F
unct
ion
Veg
etat
ive
stru
ctur
e an
d co
mpo
sitio
n
4 =
Exc
elle
nt
N/A
. Old
-gro
wth
site
s ap
pear
to h
ave
rela
tivel
y la
rge
crow
n di
amet
ers.
May
be
scor
ed if
we
use
circ
a 19
38 a
eria
l pho
tos.
3 =
Go
od
Con
ifers
>33
% c
over
, or
dist
ribut
ed th
roug
hout
site
but
com
pris
ing
<33%
cov
er. S
houl
d be
m
atur
e, b
ut p
lace
her
e if
unce
rtai
n.
2 =
Fai
rC
onife
rs p
rese
nt, b
ut lo
caliz
ed o
r co
mpr
isin
g <3
3% c
over
. OR
, con
ifers
hav
e gr
eate
r co
ver,
but
are
imm
atur
e (if
dis
cern
ible
on
aeria
l pho
tos)
.
1 =
Po
or
Con
ifers
rar
e. T
ypic
ally
hea
vily
dis
turb
ed s
tand
s th
at h
ave
conv
erte
d to
red
map
le
dom
inan
ce. S
tand
s on
ly m
appe
d if
coni
fer
com
pone
nt id
entif
iabl
e.
SIZE
Com
mun
ity s
ize
Are
a in
acr
es
4 =
Exc
elle
nt
100+
acr
es
3 =
Go
od
50-9
9 ac
res
2 =
Fai
r25
-49
acre
s
1 =
Po
or
<25
acre
s
SOUTHERN SWAMP
BIOTIC CONDITION
Com
mun
ity S
truc
ture
and
F
unct
ion
Veg
etat
ive
stru
ctur
e an
d co
mpo
sitio
n
4 =
Exc
elle
nt
N/A
. Old
-gro
wth
site
s ap
pear
to h
ave
rela
tivel
y la
rge
crow
n di
amet
ers.
May
be
scor
ed if
we
use
circ
a 19
38 a
eria
l pho
tos.
3 =
Go
od
Sta
nd d
omin
ated
by
wet
land
har
dwoo
ds.
Exc
lude
site
s co
nver
ted
from
con
ifer
dom
inan
ce.
<30%
asp
en o
r sh
rub
cove
r (s
ee IF
MA
P lu
_200
0 G
IS la
yer)
.
N/A
. Req
uire
s fie
ld s
urve
y.2
= F
air
1 =
Po
or
Site
con
tain
s >3
0% a
spen
or
shru
b co
ver
(see
IFM
AP
lu_2
000
GIS
laye
r).
Exc
lude
site
s co
nver
ted
from
con
ifer
dom
inan
ce.
SIZE
Com
mun
ity s
ize
Are
a in
acr
es
4 =
Exc
elle
nt
80+
acre
s
3 =
Go
od
50-7
9 ac
res
2 =
Fai
r
1 =
Po
or
25-4
9 ac
res
<25
acre
s
A-10
Appendix II (continued). Summary of metrics and criteria used to evaluate the quality or condition ofphoto-interpreted polygons by natural community type.
SOUTHERN FLOODPLAIN FOREST
BIOTIC CONDITION
Com
mun
ity S
truc
ture
and
F
unct
ion
Veg
etat
ive
stru
ctur
e an
d co
mpo
sitio
n
N/A
. Old
-gro
wth
site
s ap
pear
to h
ave
rela
tivel
y la
rge
crow
n di
amet
ers.
May
be
scor
ed if
we
4 =
Exc
elle
nt
use
circ
a 19
38 a
eria
l pho
tos.
3 =
Go
od
Inta
ct s
truc
ture
. Sho
uld
be m
atur
e, b
ut p
lace
her
e if
unce
rtai
n.
N/A
. P
lace
hold
er fo
r im
mat
ure
stan
ds. P
lace
her
e if
mat
urity
dis
cern
ible
on
aeria
l 2
= F
air
phot
ogra
phs.
Sig
nific
antly
deg
rade
d st
ruct
ure;
site
bar
ely
reco
gniz
able
as
flood
plai
n fo
rest
. Poo
r 1
= P
oo
rco
nditi
on ty
pica
lly in
dica
ted
by a
ltere
d hy
drol
ogy
(see
Abi
otic
Con
ditio
n).
SIZE
Com
mun
ity s
ize
Are
a in
acr
es
4 =
Exc
elle
nt
60+
acre
s
3 =
Go
od
30-5
9 ac
res
2 =
Fai
r15
-29
acre
s
1 =
Po
or
<15
acre
s
NORTHERN SHRUB THICKET
BIOTIC CONDITION
Com
mun
ity S
truc
ture
and
F
unct
ion
Veg
etat
ive
stru
ctur
e an
d co
mpo
sitio
n
4 =
Exc
elle
nt
N/A
. R
equi
res
field
sur
vey.
3 =
Go
od
Veg
etat
ion
with
in e
xpec
ted
rang
e of
var
iatio
n. Z
onat
ion
vario
us d
epen
ding
on
mat
urity
.
2 =
Fai
rN
/A.
Req
uire
s fie
ld s
urve
y.
Veg
etat
ion
outs
ide
expe
cted
ran
ge o
f var
iatio
n. P
oor
cond
ition
typi
cally
indi
cate
d by
1
= P
oo
ral
tere
d hy
drol
ogy
(see
Abi
otic
Con
ditio
n).
SIZE
Com
mun
ity s
ize
Are
a in
acr
es
4 =
Exc
elle
nt
60+
acre
s
3 =
Go
od
35-5
9 ac
res
2 =
Fai
r15
-34
acre
s
1 =
Po
or
<15
acre
s
A-11
Appendix II (continued). Summary of metrics and criteria used to evaluate the quality or condition of photo-interpreted polygons by natural community type.
SOUTHERN SHRUB-CARR
BIOTIC CONDITION
Com
mun
ity S
truc
ture
and
F
unct
ion
Veg
etat
ive
stru
ctur
e an
d co
mpo
sitio
n
4 =
Exc
elle
nt
N/A
. R
equi
res
field
sur
vey.
3 =
Go
od
Veg
etat
ion
with
in e
xpec
ted
rang
e of
var
iatio
n. Z
onat
ion
vario
us d
epen
ding
on
mat
urity
.
2 =
Fai
rN
/A.
Req
uire
s fie
ld s
urve
y.
Veg
etat
ion
outs
ide
expe
cted
ran
ge o
f var
iatio
n. P
oor
cond
ition
typi
cally
indi
cate
d by
1
= P
oo
ral
tere
d hy
drol
ogy
(see
Abi
otic
Con
ditio
n).
SIZE
Com
mun
ity s
ize
Are
a in
acr
es
4 =
Exc
elle
nt
60+
acre
s
3 =
Go
od
35-5
9 ac
res
2 =
Fai
r15
-34
acre
s
1 =
Po
or
<15
acre
s
INUNDATED SHRUB SWAMP
BIOTIC CONDITION
Com
mun
ity S
truc
ture
and
F
unct
ion
Veg
etat
ive
stru
ctur
e an
d co
mpo
sitio
n
4 =
Exc
elle
nt
N/A
. R
equi
res
field
sur
vey.
3 =
Go
od
Veg
etat
ion
with
in e
xpec
ted
rang
e of
var
iatio
n. Z
onat
ion
vario
us d
epen
ding
on
mat
urity
.
2 =
Fai
rN
/A.
Req
uire
s fie
ld s
urve
y.
Veg
etat
ion
outs
ide
expe
cted
ran
ge o
f var
iatio
n. P
oor
cond
ition
typi
cally
indi
cate
d by
1
= P
oo
ral
tere
d hy
drol
ogy
(see
Abi
otic
Con
ditio
n).
SIZE
Com
mun
ity s
ize
Are
a in
acr
es
4 =
Exc
elle
nt
60+
acre
s
3 =
Go
od
35-5
9 ac
res
2 =
Fai
r15
-34
acre
s
1 =
Po
or
<15
acre
s
A-12
Appendix II (continued). Summary of metrics and criteria used to evaluate the quality or condition of photo-interpreted polygons by natural community type.
MESIC SOUTHERN FOREST
BIOTIC CONDITION
Com
mun
ity S
truc
ture
and
F
unct
ion
Veg
etat
ive
stru
ctur
e an
d co
mpo
sitio
n
N/A
. Old
-gro
wth
site
s ap
pear
to h
ave
rela
tivel
y la
rge
crow
n di
amet
ers.
May
be
scor
ed if
we
4 =
Exc
elle
nt
use
circ
a 19
38 a
eria
l pho
tos.
Sta
nd d
omin
ated
by
nort
hern
har
dwoo
ds (
beec
h, s
ugar
map
le, e
tc.)
. Sho
uld
be m
atur
e,
3 =
Go
od
but p
lace
her
e if
unce
rtai
n. <
30%
asp
en (
see
IFM
AP
lu_2
000
GIS
laye
r).
Imm
atur
e st
ands
, or
stan
ds w
ith m
ixed
har
dwoo
d an
d as
pen
dom
inan
ce. I
nclu
des
thin
ned
2 =
Fai
rha
rdw
ood
fore
sts.
1 =
Po
or
>60%
asp
en-d
omin
ated
. O
r, y
oung
, suc
cess
iona
l sta
nd, a
s de
term
ined
on
aeria
l pho
tos.
SIZE
Com
mun
ity s
ize
Are
a in
acr
es
4 =
Exc
elle
nt
80+
acre
s
3 =
Go
od
40-7
9 ac
res
2 =
Fai
r15
-39
acre
s
1 =
Po
or
<15
acre
s
MESIC NORTHERN FOREST
BIOTIC CONDITION
Com
mun
ity S
truc
ture
and
F
unct
ion
Veg
etat
ive
stru
ctur
e an
d co
mpo
sitio
n
N/A
. Old
-gro
wth
site
s ap
pear
to h
ave
rela
tivel
y la
rge
crow
n di
amet
ers.
May
be
scor
ed if
we
4 =
Exc
elle
nt
use
circ
a 19
38 a
eria
l pho
tos.
Sta
nd d
omin
ated
by
nort
hern
har
dwoo
ds a
nd/o
r he
mlo
cks.
Sho
uld
be m
atur
e, b
ut p
lace
3
= G
oo
dhe
re if
unc
erta
in. <
30%
asp
en (
see
IFM
AP
lu_2
000
GIS
laye
r).
Imm
atur
e st
ands
, or
stan
ds w
ith m
ixed
har
dwoo
d an
d as
pen
dom
inan
ce. I
nclu
des
thin
ned
2 =
Fai
rha
rdw
ood
fore
sts.
1 =
Po
or
>60%
asp
en-d
omin
ated
. O
r, y
oung
, suc
cess
iona
l sta
nd, a
s de
term
ined
on
aeria
l pho
tos.
SIZE
Com
mun
ity s
ize
Are
a in
acr
es
4 =
Exc
elle
nt
100+
acr
es
3 =
Go
od
50-9
9 ac
res
2 =
Fai
r25
-49
acre
s
1 =
Po
or
<25
acre
s
A-13
Appendix II (continued). Summary of metrics and criteria used to evaluate the quality or condition of photo-interpreted polygons by natural community type.
DRY-MESIC NORTHERN FOREST
BIOTIC CONDITION
Com
mun
ity S
truc
ture
and
F
unct
ion
Veg
etat
ive
stru
ctur
e an
d co
mpo
sitio
n
N/A
. Old
-gro
wth
site
s ap
pear
to h
ave
rela
tivel
y la
rge
crow
n di
amet
ers.
May
be
scor
ed if
we
4 =
Exc
elle
nt
use
circ
a 19
38 a
eria
l pho
tos.
Oak
- or
pin
e-do
min
ated
sta
nd. S
houl
d be
mat
ure,
but
pla
ce h
ere
if un
cert
ain.
<30
% a
spen
3
= G
oo
d(s
ee IF
MA
P lu
_200
0 G
IS la
yer)
.
Imm
atur
e st
ands
, or
stan
ds w
ith m
ixed
oak
and
asp
en d
omin
ance
. Inc
lude
s th
inne
d oa
k 2
= F
air
fore
sts,
incl
udin
g th
ose
unde
rpla
nted
with
whi
te p
ine.
1 =
Po
or
>60%
asp
en-d
omin
ated
. O
r, y
oung
, suc
cess
iona
l sta
nd, a
s de
term
ined
on
aeria
l pho
tos.
SIZE
Com
mun
ity s
ize
Are
a in
acr
es
4 =
Exc
elle
nt
100+
acr
es
3 =
Go
od
50-9
9 ac
res
2 =
Fai
r25
-49
acre
s
1 =
Po
or
<25
acre
s
OAK-PINE BARRENS
BIOTIC CONDITION
Com
mun
ity S
truc
ture
and
F
unct
ion
Veg
etat
ive
stru
ctur
e an
d co
mpo
sitio
n
N/A
. R
equi
res
field
sur
vey,
unl
ess
nativ
e vs
. non
-nat
ive
grou
ndco
ver
can
be d
eter
min
ed
4 =
Exc
elle
nt
from
aer
ial p
hoto
s.
Mod
erat
e tr
ee a
nd s
hrub
enc
roac
hmen
t, to
geth
er c
ompr
isin
g <5
0% c
over
at t
he s
ite.
3 =
Go
od
Nat
ive
vs. n
on-n
ativ
e gr
ound
cove
r lik
ely
inde
term
inab
le o
n ae
rial p
hoto
s.
Tre
e an
d sh
rub
encr
oach
men
t sev
ere,
toge
ther
com
pris
ing
>50%
cov
er a
t the
site
, whi
ch
2 =
Fai
ris
stil
l rec
ogni
zabl
e as
bar
rens
.
Rep
rese
nts
form
er o
ak-p
ine
barr
ens
(bas
ed o
n ci
rca
1800
veg
etat
ion
map
) th
at h
ave
1 =
Po
or
conv
erte
d to
dry
-mes
ic n
orth
ern
fore
st d
ue to
pos
t-18
00 fi
re s
uppr
essi
on.
SIZE
Com
mun
ity s
ize
Are
a in
acr
es
4 =
Exc
elle
nt
100+
acr
es
3 =
Go
od
55-9
9 ac
res
2 =
Fai
r25
-54
acre
s
1 =
Po
or
<25
acre
s
A-14
Appendix II (continued). Summary of metrics and criteria used to evaluate the quality or condition of photo-interpreted polygons by natural community type.
DRY SAND PRAIRIE
BIOTIC CONDITION
Com
mun
ity S
truc
ture
and
F
unct
ion
Veg
etat
ive
stru
ctur
e an
d co
mpo
sitio
n
N/A
. R
equi
res
field
sur
vey.
Gro
undc
over
sho
uld
be >
90%
nat
ive
spp.
(w
arm
sea
son
4 =
Exc
elle
nt
gras
ses
show
as
gray
or
brow
n on
aer
ial p
hoto
s).
Mod
erat
e tr
ee a
nd s
hrub
enc
roac
hmen
t, to
geth
er c
ompr
isin
g <2
5% c
over
at t
he s
ite.
3 =
Go
od
Mos
tly n
ativ
e co
ver,
if d
eter
min
able
on
aeria
l pho
tos.
Tre
e an
d sh
rub
encr
oach
men
t sig
nific
ant,
toge
ther
com
pris
ing
>25%
cov
er a
t the
site
, 2
= F
air
whi
ch is
stil
l rec
ogni
zabl
e as
pra
irie.
Or,
sig
nific
ant c
over
by
non-
nativ
e sp
p.
Con
vert
ed o
ak-p
ine
barr
ens,
or
seve
rely
deg
rade
d st
ruct
ure.
Non
-nat
ive
spp.
dom
inat
e 1
= P
oo
rgr
ound
laye
r (c
ool s
easo
n gr
asse
s te
nd to
sho
w a
s pi
nk o
n ph
otos
).SIZE
Com
mun
ity s
ize
Are
a in
acr
es
4 =
Exc
elle
nt
10+
acre
s
3 =
Go
od
6-9.
9 ac
res
2 =
Fai
r3-
5.9
acre
s
1 =
Po
or
<3 a
cres
SUBMERGENT MARSH
BIOTIC CONDITION
Com
mun
ity S
truc
ture
and
F
unct
ion
Veg
etat
ive
stru
ctur
e an
d co
mpo
sitio
n
4 =
Exc
elle
nt
N/A
. R
equi
res
field
sur
vey.
3 =
Go
od
Veg
etat
ion
with
in e
xpec
ted
rang
e of
var
iatio
n. Z
onat
ion
vario
us d
epen
ding
on
mat
urity
.
2 =
Fai
rN
/A.
Req
uire
s fie
ld s
urve
y.
Veg
etat
ion
outs
ide
expe
cted
ran
ge o
f var
iatio
n. P
oor
cond
ition
typi
cally
indi
cate
d by
1
= P
oo
ral
tere
d hy
drol
ogy
(see
Abi
otic
Con
ditio
n).
SIZE
Com
mun
ity s
ize
Are
a in
acr
es
4 =
Exc
elle
nt
100+
acr
es
3 =
Go
od
60-9
9 ac
res
2 =
Fai
r30
-59
acre
s
1 =
Po
or
<30
acre
s
A-15
Appendix II (continued). Summary of metrics and criteria used to evaluate the quality or condition of photo-interpreted polygons by natural community type.
EMERGENT MARSH
BIOTIC CONDITION
Com
mun
ity S
truc
ture
and
F
unct
ion
Veg
etat
ive
stru
ctur
e an
d co
mpo
sitio
n
4 =
Exc
elle
nt
N/A
. R
equi
res
field
sur
vey.
3 =
Go
od
Veg
etat
ion
with
in e
xpec
ted
rang
e of
var
iatio
n. Z
onat
ion
vario
us d
epen
ding
on
mat
urity
.
2 =
Fai
rN
/A.
Req
uire
s fie
ld s
urve
y.
Veg
etat
ion
outs
ide
expe
cted
ran
ge o
f var
iatio
n. P
oor
cond
ition
typi
cally
indi
cate
d by
1
= P
oo
ral
tere
d hy
drol
ogy
(see
Abi
otic
Con
ditio
n).
SIZE
Com
mun
ity s
ize
Are
a in
acr
es
4 =
Exc
elle
nt
100+
acr
es
3 =
Go
od
60-9
9 ac
res
2 =
Fai
r30
-59
acre
s
1 =
Po
or
<30
acre
s
NORTHERN WET MEADOW
IOTIC CONDITION
Com
mun
ity S
truc
ture
and
F
unct
ion
Veg
etat
ive
stru
ctur
e an
d co
mpo
sitio
n
4 =
Exc
elle
nt
N/A
. R
equi
res
field
sur
vey.
Mod
erat
e sh
rub
encr
oach
men
t, co
mpr
isin
g <2
5% c
over
at t
he s
ite.
Nat
ive
vs. n
on-n
ativ
e 3
= G
oo
dgr
ound
cove
r lik
ely
inde
term
inab
le o
n ae
rial p
hoto
s.
Shr
ub e
ncro
achm
ent s
igni
fican
t, co
mpr
isin
g >2
5% c
over
at t
he s
ite, w
hich
is s
till
2 =
Fai
rre
cogn
izab
le a
s w
et m
eado
w.
Shr
ub e
ncro
achm
ent s
ever
e, c
ompr
isin
g >5
0% c
over
at t
he s
ite, w
hich
is s
till r
ecog
niza
ble
1 =
Po
or
B
as w
et m
eado
w.
SIZE
Com
mun
ity s
ize
Are
a in
acr
es
4 =
Exc
elle
nt
60+
acre
s
3 =
Go
od
40-5
9 ac
res
2 =
Fai
r20
-39
acre
s
1 =
Po
or
<20
acre
s
A-16
Appendix II (continued). Summary of metrics and criteria used to evaluate the quality or condition of photo-interpreted polygons by natural community type.
INTERMITTENT WETLAND
IOTIC CONDITION
Com
mun
ity S
truc
ture
and
F
unct
ion
Veg
etat
ive
stru
ctur
e an
d co
mpo
sitio
n
4 =
Exc
elle
nt
N/A
. R
equi
res
field
sur
vey.
3 =
Go
od
Veg
etat
ion
with
in e
xpec
ted
rang
e of
var
iatio
n. Z
onat
ion
vario
us d
epen
ding
on
mat
urity
.
2 =
Fai
rN
/A.
Req
uire
s fie
ld s
urve
y.
Veg
etat
ion
outs
ide
expe
cted
ran
ge o
f var
iatio
n. P
oor
cond
ition
typi
cally
indi
cate
d by
1
= P
oo
rB
alte
red
hydr
olog
y (s
ee A
biot
ic C
ondi
tion)
.SIZE
Com
mun
ity s
ize
Are
a in
acr
es
4 =
Exc
elle
nt
40+
acre
s
3 =
Go
od
20-3
9 ac
res
2 =
Fai
r5-
19 a
cres
1 =
Po
or
<5 a
cres
COASTAL PLAIN MARSH
IOTIC CONDITION
Com
mun
ity S
truc
ture
and
F
unct
ion
Veg
etat
ive
stru
ctur
e an
d co
mpo
sitio
n
4 =
Exc
elle
nt
N/A
. R
equi
res
field
sur
vey.
3 =
Go
od
Veg
etat
ion
with
in e
xpec
ted
rang
e of
var
iatio
n. Z
onat
ion
vario
us d
epen
ding
on
mat
urity
.
2 =
Fai
rN
/A.
Req
uire
s fie
ld s
urve
y.
B
Veg
etat
ion
outs
ide
expe
cted
ran
ge o
f var
iatio
n. P
oor
cond
ition
typi
cally
indi
cate
d by
SIZE
Com
mun
ity s
ize
Are
a in
acr
es
1 =
Po
or
alte
red
hydr
olog
y (s
ee A
biot
ic C
ondi
tion)
.
4 =
Exc
elle
nt
40+
acre
s
3 =
Go
od
20-3
9 ac
res
2 =
Fai
r5-
19 a
cres
1 =
Po
or
<5 a
cres
A-17
Appendix II (continued). Summary of metrics and criteria used to evaluate the quality or condition of photo-interpreted polygons by natural community type.
1 =
Po
or
<5 a
cres
LAKEPLAIN WET-MESIC PRAIRIE
BIOTIC CONDITION
Com
mun
ity S
truc
ture
and
F
unct
ion
Veg
etat
ive
stru
ctur
e an
d co
mpo
sitio
n
4 =
Exc
elle
nt
N/A
. R
equi
res
field
sur
vey,
unl
ess
nativ
e vs
. non
-nat
ive
grou
ndco
ver
can
be d
eter
min
ed
from
aer
ial p
hoto
s.
3 =
Go
od
Mod
erat
e tr
ee a
nd s
hrub
enc
roac
hmen
t, to
geth
er c
ompr
isin
g <2
5% c
over
at t
he s
ite.
Nat
ive
vs. n
on-n
ativ
e gr
ound
cove
r lik
ely
inde
term
inab
le o
n ae
rial p
hoto
s.
2 =
Fai
rT
ree
and
shru
b en
croa
chm
ent s
igni
fican
t, to
geth
er c
ompr
isin
g >2
5% c
over
at t
he s
ite,
whi
ch is
stil
l rec
ogni
zabl
e as
pra
irie.
Tre
e an
d sh
rub
encr
oach
men
t sev
ere,
toge
ther
com
pris
ing
>50%
cov
er a
t the
site
, whi
ch
SIZE
Com
mun
ity s
ize
Are
a in
acr
es
1 =
Po
or
is s
till r
ecog
niza
ble
as p
rairi
e.
4 =
Exc
elle
nt
5+ a
cres
3 =
Go
od
2-4.
9 ac
res
2 =
Fai
r1-
1.9
acre
s
NORTHERN FEN
BIOTIC CONDITION
Com
mun
ity S
truc
ture
and
F
unct
ion
Veg
etat
ive
stru
ctur
e an
d co
mpo
sitio
n
1 =
Po
or
<1 a
cre
4 =
Exc
elle
nt
N/A
. R
equi
res
field
sur
vey.
3 =
Go
od
Mod
erat
e sh
rub
encr
oach
men
t, co
mpr
isin
g <2
5% c
over
at t
he s
ite.
Nat
ive
vs. n
on-n
ativ
e gr
ound
cove
r lik
ely
inde
term
inab
le o
n ae
rial p
hoto
s.
2 =
Fai
rS
hrub
enc
roac
hmen
t sig
nific
ant,
com
pris
ing
>25%
cov
er a
t the
site
, whi
ch is
stil
l re
cogn
izab
le a
s fe
n.
Shr
ub e
ncro
achm
ent s
ever
e, c
ompr
isin
g >5
0% c
over
at t
he s
ite, w
hich
is s
till r
ecog
niza
ble
SIZE
Com
mun
ity s
ize
Are
a in
acr
es
1 =
Po
or
as fe
n.
4 =
Exc
elle
nt
35+
acre
s
3 =
Go
od
20-3
4 ac
res
2 =
Fai
r5-
19 a
cres