Science of Watershed Health


BIOLOGY - Terrestrial Habitat Quality

What is the quality of terrestrial habitat in each watershed based on its size, configuration and cover type?

Why is this important for biology?

Terrestrial Habitat Quality Health Scores

Biology Index - Terrestrial Habitat Quality (catchment scale)

key for map

Areas with native and natural vegetation may provide habitat for a number of species, but habitat should also be large enough to support plant and animal interactions over time.  The assessment of habitat quality is complex because the ability to provide the necessary components for a species is dependent upon which species is being investigated. For this reason, the habitat quality index uses the different habitat needs of representative forest, wetland and grassland species to determine whether remaining habitat configurations meet the needs of those particular species. The species selected and the habitat configurations used for this index are based on the methods developed for the Regionally Significant Ecological Areas modeling done by the DNR.

Areas with native and semi-natural vegetation provide important habitat for a number of species. The size, shape and cover type of terrestrial habitat influence the productivity, accessibility and sustainability of the plant and animal communities dependent upon those habitats.

Quality habitat should be large enough to support plant and animal interactions over time. Larger areas that are round rather than long and narrow contain more habitats and may be more likely to support diverse plant and animal populations. However, the definition and assessment of habitat quality is complex because the ability of a particular habitat to provide the necessary components for a species is dependent upon which species is being investigated. For example, small fragments of habitat with unique conditions may harbor plant microhabitats. Thus, small habitat may contain rare plant habitat, which may maintain important genetic material.

Conversely, some organisms need larger areas, such as Ovenbirds and other ground-nesting forest songbirds, which require large, intact forest tracts (Mattsson and Niemi 2006). Fragmentation of vegetation may reduce nest success of Ovenbirds. Small, irregular shaped areas have a greater proportion of edge area than interior area. More birds may be forced to nest in the edges where there is a greater risk of nest predators (Chalfoun et al. 2002). In other words, it is hard to determine if it is “good” habitat unless it is known what species live and thrive there.

For this reason, the habitat quality index uses the different habitat needs of representative forest, wetland and grassland species to determine whether remaining habitat configurations meet the needs of those particular species. The species selected and the habitat configurations used for this index are based on the methods developed for the Regionally Significant Ecological Areas modeling done by the DNR in their Metro Region. The species used in the forest model included: the Red-Eyed Vireo, Wood Thrush, Scarlet Tanager, Ovenbird, Eastern Wood Pewee, Cerulean Warbler, Louisiana Waterthrush, and the Red-Shouldered Hawk. The species used for the grassland model included: bull snake, pocket mouse, and Blanding's turtle.






Input Data


Calculating the index

Interpreting the results


Relationship to other health components

Scientific literature support


Confidence in index

Future enhancements


Input Data:


Calculating the index

The Habitat Quality Index uses models based on land cover from the National Land Cover Data (2001) and National Agricultural Statistic Service (NASS) together with the Minnesota County Biologic Survey (MCBS) Biodiversity Significance data to identify and rank potential terrestrial habitat. Computer models were used to depict forest, wetland and grassland habitat quality, which were combined into a composite habitat score. This index does not consider aquatic (stream and lake) habitat types, but does score wetland habitats.

Natural areas identified through the individual forest, wetland, and grassland models are integrated with Minnesota County Biological Survey sites to identify the final habitat areas. These areas were then assigned a final score of 3, 2, or 1 (3 being the highest) based on how well the area meets standards for size, shape, connectivity, adjacent land use/cover, and species diversity.

Score 3 - Areas larger in size, and/or with few adjacent land cover types or land uses that could adversely affect the area; greater diversity of vegetation types; or an isolated native plant community mapped and given a score of outstanding biodiversity significance by the Minnesota County Biological Survey.

Score 2 - Areas moderate in size and/or with less suitable adjacent land cover types or land uses; areas with less diversity of vegetation cover types; or an isolated native plant community mapped and given a score of high biodiversity significance by the Minnesota County Biological Survey.

Score 1 - Areas smaller in size but meeting minimum size requirements for regional significance; may have less diversity of vegetation cover types; may have less suitable adjacent cover types or land uses; or the area may be an isolated native plant community mapped and given a score of moderate biodiversity significance by the Minnesota County Biological Survey.

Habitat ranked as “high” or “very high” for biodiversity significance received an additional value of 2 or 3. These created habitat patches with aggregate rankings between 1 and 9. Although a wetland score can be combined with a forest or grassland score, the forest and grassland land cover types are mutually exclusive limiting the maximum habitat quality score to 9. (See Appendix A: “Methodology for Conducting a Statewide GIS Ecological Assessment” for further detail regarding the data models.)

After identifying and weighting the value of the habitat patches across the state, these values were summarized to the watershed boundary into an area weighted score.

Scoring Method:
The weighted habitat value for each watershed was divided by the mean habitat quality value multiplied (x) by the watershed land area:

(Patch Scores x Patch Areas) / (mean Habitat Quality Score possible x Watershed Land Area)

The watershed score is based on a comparison between current habitat quality to an estimated “best attainable” habitat quality. A mean habitat quality value (5) was selected rather than the highest possible habitat value because even a pristine landscape would not generate the highest value across an entire watershed landscape, given that optimal conditions for multiple habitat types would not be present at the same location.

The final score was created by multiplying this watershed value by 100.

   0 -20 = very poor habitat quality
  20-40 = poor habitat quality
  40-60 = moderate habitat quality
  60-80 = good habitat quality
80-100 = excellent habitat quality


Interpretation of results

There were more terrestrial habitat patches in the northeast, with moderate values in small areas in the southeast and east central and north. Most of southern and western watersheds have low terrestrial habitat quality scores.

Although there are areas of Minnesota that have significant pockets of high quality habitat, in most areas of the state, the configuration of that habitat does not meet the requirements of the representative species. Due to the large amount of fragmentation from agriculture, urban development, forest harvest, mining and other uses, the habitat value of remaining patches has been reduced. There is a pattern of poor habitat quality throughout the southern half of Minnesota and the Red River Valley.

There are some low scores in the north and northeast where large expanses of forested land cover, wetlands, and lakes dominate the landscape. Some of the low scores are due to forest harvest and “blow down” of trees. Additionally, some low scores may be the result of fragmentation of terrestrial habitat by open water. The fragmentation by open water reduces the value of the terrestrial habitat in the modeling process.


Relationship to other health components

Water quality

High water quality is correlated with large tracts of perennial cover and wide riparian areas. The same configuration of land cover that scores as high value habitat also serves to protect adjacent water bodies by filtering contaminants, reducing erosion and moderating temperature.


The pattern of remaining quality habitat across the landscape often follows Minnesota’s streams and rivers. These habitat corridors contribute to the health of both the terrestrial and aquatic systems (see the Riparian Connectivity and Terrestrial Connectivity indexes). The Terrestrial Habitat Quality Index is used as a base for the Terrestrial Habitat Connectivity Index that ranks potential connections between quality habitats. Increasing the size of these existing natural networks would enhance and connect the movement of organisms, nutrients and energy through Minnesota’s watersheds.


The patterns of historic land cover correspond directly to the soil types and climate patterns of Minnesota. The geology and glacial history of the state greatly influence the location of expected habitat types and species. The ecological setting is key to understanding the expected habitat and the degree to which the current landscape deviates from past conditions.


Terrestrial habitat quality is directly related to perennial land cover. The low scoring areas in most cases have much less perennial cover and are most likely to experience high runoff, lowered rates of groundwater recharge and less transpiration and more erosion, whereas areas with high quality habitat likely experience less runoff, more groundwater recharge, and less erosion, and more evapotranspiration.


Scientific literature support

An extensive literature review was part of the Regionally Significant Ecological Areas (RSEA) Modeling Project done by the DNR in their Metro Region.  The same basic approach was applied to this statewide analysis, so the literature from that effort serves to also enforce the validity and importance of this index.  A major focus of the RSEA was to identify, and when possible, maintain the remaining large habitat areas in the landscape.

A number of studies in Minnesota and other states have evaluated the size of habitat patches relative to productivity, accessibility, and sustainability of plants and animals.  All things being equal, large areas of habitat support more diverse species than smaller areas, especially for birds or mammals.  For example, large tracts of public forest land and private industrial forest land are both important in managing for a broad-ranging animal, such as the wolf (Mladenoff et al. 1995).  However, habitat size can be relatively small and still maintain many plants and small, non-migratory animals.  In the Midwest, Temple and Cary (1988) recommend areas of forest habitat should be large enough to allow for a 200 m edge buffer for interior forest bird populations.  Keller et al. (1993) reported that forest tracts wider than 100 m had more neotropical migrant species and the birds continued to increase in numbers but much more gradually in forests wider than 200 m.  Robbins et al. (1989) studied birds in the Mid-Atlantic United States and found that several species were not detected in forests less than 10 ha in area.  Friesen et al. (1995) studied woodland sizes relative to neotropical bird abundance and found 4 ha woodlots without any nearby houses had on average more bird community abundance than 25 ha woodlots in urban areas.


Small areas may be sufficient to maintain vegetation communities.  Levenson (1981:38) estimated that dry, midwestern forest areas must be larger than 3.8 ha to sustain forest-interior vegetation communities.  Watkins et al. (2003) examined the distribution of understory plants and relationships to unpaved forest roads in a northern hardwood forest landscape in the Chequamegon National Forest, Wisconsin and found that species richness of vegetation was similar to forest interior levels within 5 m from the road.  Chalfoun et al. (2002) reported that forest edges in agricultural landscapes may attract a greater numbers of nest predator species than forest interiors, but found that forest vegetation structure was very similar between edge and interior, suggesting that differences in predator abundance and species richness were not driven by variation in habitat structure.


Lehtinen et al. (1999) reported the species richness of 10 amphibians was lower with greater wetland isolation and road density in urban and agricultural regions of central and southwestern Minnesota.  The reduction in amphibians was likely because the majority of adult amphibians travel no further than 200 or 300 m from a home pond to foraging or over-wintering sites (Rothermel et al., 2002).  Gibbs (2000) suggests wetlands greater than 1 acre (0.4 ha) are likely the minimum size to sustain the wetland organisms.  In addition to maintaining a wetland of a minimum size the habitat that surrounds a wetland is important. Shisler et al. (1987) suggests vegetation widths of 30 and 45 meters around wetlands is important for wildlife, especially in urban environments.


The RSEA used 16 hectares as the minimum size for tall grasslands with a minimum width of 90 meters, because large patches of grassland habitat have been recommended for grassland birds in North America (Winter et al. 2006).  Density of three grassland bird species in North Dakota and Minnesota was influenced by size of the habitat, landscape, region, and local vegetation structure (Winter et al. 2006).



Confidence in index

The basic relationship between habitat size, shape and distribution and quality of habitat for representative species is well documented. This relationship has been tested in many settings and the variables selected for defining the different habitat needs based on type were investigated thoroughly for the original RSEA modeling effort.


Future enhancements

The quality of the data is good with consistent coverage across the state. The index could be improved by including the Biodiversity rankings for the sections of Minnesota that are incomplete. Greater detailed habitat cover could produce more accurate results, but given the size and complexity of the computations, would only be beneficial as greater computer power becomes available.

The scoring process could be improved with a baseline value for habitat quality in each watershed. A potential baseline value could be generated using circa 1895 land cover dataset (Marschner data) to determine a more accurate maximum quality value for comparison. This value would be used to more accurately score the current condition based on an attainable historical condition.

An aquatic habitat quality component could be added to this index to offset the terrestrial fragmentation in the current index, but data are not available at present to calculate an aquatic habitat quality index for lakes.