EE IHHIIIHINHIWIIHlmlmllIIHIHHHIITlllHllHllfillll

   

.THS_

 

m LIBRARY
\ . Michigan State
it“? University

This is to certify that the
thesis entitled

Evaluating the Impacts of Residential Use on the Lapeer State
Game Area

presented by

Eric M. Clark

has been accepted towards fulfillment

of the requirements for the
MS degree in Community, Agriculture.
Recreation and Resource
Studies

 

 

' 5%/%

Date

 

MSU Is on Mariam-action, equal-opportunity employer

.—.-.-----J.-u--u---o--o-I--o-e-e---c-u-e-a-n--e--o-.-.-.-.-.-.-.-.-.-.-

---c-c-----e---e-e-u--.—— — — —

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 KzlProilec3Pres/CIRC/Date0ue indd

 

EVALUATING THE IMPACTS OF RESIDENTIAL USE ON
THE LAPEER STATE GAME AREA

By

Eric M. Clark

A THESIS

Submitted to
Michigan State University
In partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE
Community Agriculture Recreation Resource Studies

2008

ABSTRACT

EVALUATING THE IMPACTS OF RESIDENTIAL USE ON THE LAPEER STATE
GAME AREA

By
Eric M. Clark

The state-owned portion of the Lapeer State Game Area (LSGA) is 8,246 acres.
It is located in Lapeer County in Southeastern Michigan. This thesis evaluated the
impacts current and future residential use inside the LSGA’S Dedicated Management
Boundary surrounding the LSGA that consists mostly of private lands. This was
accomplished by utilizing landscape ecological metrics to quantify changes in the patch
structure of non-built landcover types over the last 28 years (1978-2006) and for two
alternative future development scenarios; The first scenario was a build out based on the
full capacity allow by current zoning regulations in the townships surround the LSGA
and the second was a build out that implements strategic conservation measures to

protect ecologically significant non-built landcover patches.

This study showed that increased residential use from 1978 to 2006 has altered
the patch structure of non-built landcover types. Projected increases in residential use
under the capacity scenario showed further impacts to non-built landcover. Projected
residential use under the conservation scenario showed that the preservation of
ecologically significant patches could reduce the impact of residential use on non-built
landcover patch structure. This research was intended to demonstrate methods
incorporated biological information into both the planning efforts of the local

municipalities and those for the LSGA.

ACKNOWLEDGEMENTS

First I would like thank my graduate committee, Dr. Charles Nelson, Dr. Gary Roloff,
and Dr. Christine Vogt for all their guidance through the course of this process. Second
and by no means least; I thank my wife Robin for all her hours of effort.

iii

TABLE OF CONTENTS

INTRODUCTION ............................................................................................................... 1
OBJ ECTIVES ...................................................................................................................... 5
STUDY SITE AND SITUATION ....................................................................................... 7
Study Site .................................................................................................................. 7
Study Situation .......................................................................................................... 8
LITERATURE REVIEW .................................................................................................. 11
METHODS ........................................................................................................................ 22
Land Use Mapping .................................................................................................. 23
Alternative Development Scenarios ........................................................................ 25
Landscape Structure ................................................................................................ 30
Limitations of the Study .......................................................................................... 34
RESULTS .......................................................................................................................... 36
Increased Residential Land Use from 1978 to 2006 and Alternative Development
Scenarios ................................................................................................................. 36
1978 to 2006 ........................................................................................................ 36
Full Capacity Alternative Development Scenario ............................................... 39
Conservation Alternative Development Scenario ............................................... 42
Change in the Landscape Metrics for Non-Built Land Cover Classes ................... 47
Number of Patches .............................................................................................. 47
Total Class Area .................................................................................................. 49
Mean Patch Size .................................................................................................. 50
Total Core Area ................................................................................................... 51
Mean Proximity Index ......................................................................................... 52
DISCUSSION .................................................................................................................... 54
Urbanization and Landcover Change from 1978 to 2006 ....................................... 54
Alternative Development Scenarios ........................................................................ 57
Capacity Scenario Potential Land Use Change and Impacts to Non-built Landcover
....................................................................................................................................... 58
Conservation Scenario Potential Land Use Change and Impacts to Non-Built
Landcover ................................................................................................................ 60

iv

Alternative Development Scenario Comparison ..................................................... 62

MANAGEMENT CONSIDERATIONS AND FUTURE RESEARCH ........................... 64

LITERATURE CITED ...................................................................................................... 68

LIST OF TABLES

Table 1. Classes used for land use mapping and comparison. ................................ 25
Table 2. Scoring criteria for the conservation constraints data layer. ....................... 28

Table 3. Zoning types with minimum lots sizes and total area of each zoning within the
Lapeer State Game Area Dedicated Management Boundary. ................................ 34

Table 4. Number of potential residential dwellings calculated at firll capacity build out
per zoning class and minimum lot size expressed in acres. ................................... 40

Table 5. Number of potential residential dwellings calculated at build out for the
conservation development scenario per zoning class and minimum lot size, expressed in
acres. .................................................................................................. 44

Table 6. Generalized behavior of landscape metrics from 1978 to 2006 for non-built
landcover classes inside the Lapeer State Game Area Dedicated Management Boundary.
......................................................................................................... 55

Table 7. Generalized behavior of landscape metrics from 2006 to full capacity alternative
development scenario for non-built landcover types within the Lapeer State Game Area
Dedicated Management Boundary. ............................................................... 58

Table 8. Generalized behavior of landscape metrics from 2006 to conservation
development scenario for non-built landcover classes within the Lapeer State Game Area
Dedicated Management Boundary. ............................................................... 60

vi

LIST OF FIGURES

Figure 1. Regional location of the Lapeer State Game Area in Southeast Michigan. . .3

Figure 2. Spatial Processes in Land Transformation. The above graph depicts the relative
importance of each spatial process in relation to the amount of existing original land
cover type (adapted from Forman 1995a). ...................................................... 20

Figure 3. 2006 land use dataset displayed with 40% transparency with the 2006 1 meter
leaf on photography at a scale of 125000. ........................................................ 24

Figure 4. Forest and grassland conservation rankings. Four is the highest conservation
value based on total class area, core area, and presence of water bodies. The top image
displays grassland features and the bottom image displays forest features; both are
displayed with the GLS Greenlinks Green Infrastructure Network boundaries. ........... 29

Figure 5. Conservation constraint dataset shown with the GLS Greenlinks Green
Infrastructure Network. ............................................................................. 30

Figure 6. One acre square buffers used to create the land use datasets for the alternative
development scenarios. ............................................................................ 30

Figure 7. Zoning classes for the five townships that fall within the Lapeer State Game
Area Dedicated Management Boundary. ........................................................ 32

Figure 8. Constraints to the build out analysis. The image displays roads, existing "built"
land use polygons, and utility corridors. ......................................................... 33

Figure 9. 1978 Land use map for the Lapeer State Game Area Dedicated Management
Boundary. ............................................................................................ 37

Figure 10. 2006 Land use map for the Lapeer State Game Area Dedicated Management
Boundary. Existing residential structures are displayed. ...................................... 38

Figure 11. Full capacity alternative development scenario land use map for the Lapeer
State Game Area Dedicated Management Boundary. ......................................... 41

Figure 12. Conservation alternative development scenario land use map for the Lapeer
State Game Area Dedicated Management Boundary. ......................................... 43

Figure 13. Change in the number of patches for built land use classes inside the Lapeer
State Game Area Dedicated Management Boundary from 1978 to 2006 and for the full
Capacity and Conservation alternative development scenarios. .............................. 45

vii

Figure 14. Change in the total class area for built land use classes inside the Lapeer State
Game Area Dedicated Management Boundary from 1978 to 2006 and for the Capacity
and Conservation alternative development scenarios. ......................................... 45

Figure 15. Change in the mean patch size for built land use classes inside the Lapeer
State Game Area Dedicated Management Boundary for 1978 to 2006 and for the firll
Capacity and Conservation alternative development scenarios. .............................. 46

Figure 16. Change in the mean proximity index for built land use classes inside the
Lapeer State Game Area Dedicated Management Boundary for 1978 to 2006 and for the
full Capacity and Conservation alternative development scenarios. The bar representing
the mean proximity index for the residential land use class is not to scale. . . . . . . . . . . ....46

Figure 17. Change in the number of patches for non-built landcover types inside the
Lapeer State Game Area Dedicated Management Boundary for 1978, 2006, and the full
Capacity and Conservation alternative development scenarios. .............................. 48

Figure 18. Change in the total class area for non-built landcover classes inside the Lapeer
State Game Area Dedicated Management Boundary for 1978, 2006, and the full Capacity
and Conservation alternative development scenarios. ......................................... 50

Figure 19. Change in the mean patch size for non-built landcover types inside the Lapeer
State Game Area Dedicated Management Boundary for 1978, 2006, and the full Capacity
and Conservation alternative development scenarios. ......................................... 51

Figure 20. Change in the total core area for non-built landcover types inside the Lapeer
State Game Area’s Dedicated Management Boundary for 1978, 2006, and the full
Capacity and Conservation alternative development scenarios. .............................. 52

Figure 21. Change in the mean proximity index for non-built landcover types inside the
Lapeer State Game Area’s Dedicated Management Boundary for 197 8, 2006, and the full
Capacity and Conservation alternative development scenarios. Data point for the
Capacity Scenario is not shown. .................................................................. 53

IMAGES IN THIS THESIS ARE PRESENTED IN COLOR

viii

INTRODUCTION

The Lapeer State Game Area (LSGA) has 8,246 acres of state owned as part of a
400,000 acre network of State game areas maintained by Michigan’s Department of
Natural Resources (MDNR) Wildlife Division. The LSGA is located in Lapeer County
in the southeastern portion of Michigan. The LSGA maintains an important role as a
social and ecological resource to the surrounding region. It forms the backbone of the
surrounding municipalities’ green infrastructure network. The LSGA performs valuable
ecological services, such as water quality protection, nonstructural flood protection, and
provision of wildlife habitat. Social services include recreational access, increased
property values, decreased cost of service provision, and the attraction of tourists. The
LSGA’s significance in the region gives land planners and managers of the surrounding
municipalities incentive to ensure the integrity of the LSGA by maintaining the
grasslands, forest, and wetlands of the surrounding landscape.

The MDNR Wildlife Division is charged with managing the LSGA and all other
State game areas “to provide benefits to hunters and trappers and to conserve wildlife
resources for the benefit of all citizens and visitors to the State” (MDNR, 2003). The
MDNR Wildlife Division is staffed primarily with biologists, whose work is focused
on managing the natural resources of the State. It differs from the MDNR Parks and
Recreation Management Division, which is primarily staffed with employees
specializing in recreation, whose work emphasizes the role of people in recreation and
management of the State Parks and Recreation Areas. The Wildlife Division is
mandated to “develop annual operational work plans and budgets, develop long-range
management plans for game areas and State forests, and participate in other landscape

level or basin planning efforts” (MDNR, 2003). The LSGA 10-year Management Plan

on-file is twenty years outdated (Douville, 1977). In 2001 , the MDNR began a two year
strategic planning process to create the Statewide Wildlife Action Plan, as part of their
broader planning and management efforts (Michigan Department of Natural Resources,
2008). This planning process was “aimed at identifying and prioritizing the
conservation needs of wildlife and their habitats, especially those species with low or
declining populations” (Michigan Department of Natural Resources, 2008). During this
process, the MDNR identified priority threats to wildlife in southern Michigan. Three
of these threats were considered most significant: expansion of industrial, residential,
and recreational development; invasive species; and fragmentation. The MDNR
included recommendations for dealing with fragmentation and other issues associated
with development, including working with local and regional land managers.

The LSGA is located in close proximity to two major metropolitan areas: Detroit
and Flint (Figure 1). Due to its position at the urban fringe, the landscape surrounding
the LSGA has been and continues to be impacted by residential development. The most
visible impacts of residential development is the alteration of the landscape structure,
including change in the pattern of grassland, forest, and wetland patches; yet social
impacts are also apparent, including a reduction in the area of the LSGA open to
hunting due to safety zones eliminating hunting opportunities within 450 feet of a

residential structure.

 

MDNR Parcel
Boundaries

 

 

 

 

Figure 1. Regional location of the Lapeer State Game Area in Southeast Michigan.

The maintenance of the LSGA as a social and ecologic resource for the region
requires active participation by the surrounding municipalities and the MDNR. Each
entity has tools for guiding residential development in the surrounding landscape.
Municipalities have the ability to guide the overall pattern of development with land use
planning and zoning and Park, Recreation, and Greenspace Planning processes. The
MDNR has fee simple ownership of 8,246 acres. One available tool is the fee simple
acquisition of land through available funding sources such as the Michigan Natural
Resource Trust Fund. The MDNR can also work through the Landowner Incentive
Program to conserve and provide public access to land. Collaborative efforts between
the MDNR and the surrounding municipalities could produce innovative strategies to
control growth with the possibility of greater success. These strategies include

incorporating ecological criteria into zoning ordinances and developing tax incentives

for the conservation of ecologically important lands, among other possible strategies. In
order for surrounding municipalities and the MDNR to employ their tools effectively, it
is important that they understand the impacts of residential development on the LSGA
and potential conservation lands.

The over arching goal of this thesis is to demonstrate one possible method of
linking the Michigan Department of Natural Resources’ planning process for the Lapeer
State Game Area to local-level land use planning processes for the five townships that
intersect the Dedicated Management Boundary: Arcadia, Deerfield, Marathon,
Mayfield, and Oregon Townships. The author will draw on the planning phases of
Ecosystem Management, the paradigm on which the MDNR Wildlife Action Plan is
based, as presented by Botequilha Leitao and Ahem (2002) citing McGarigal (1998).
The focus lies on identifying areas where landscape ecology can be used to facilitate the
inclusion of biological and/or ecological information in the MDNR and local municipal
planning processes. This research is mainly concerned with two of McGarigal’s
planning phases: performing a needs assessment and evaluating alternative development
plans by utilizing landscape ecology. This paper assumes a common goal of the MDNR
and local municipalities of protecting the ecological function of the landscape in the
face of expanding residential development.

This research furthers understanding of the impacts of past and future residential
development on the Lapeer State Game Area. The author developed strategies for
prioritizing important and at-risk areas for future conservation efforts. Michigan
Department of Natural Resources and local municipality tools were identified for use in
future conservation of prioritized areas. This work can serve as a model for utilizing

landscape ecology in firture MDNR and local municipality Master Plans.

OBJECTIVES

1. Evaluate past and firture residential development trends inside the Lapeer
State Game Area Dedicated Management Boundary using landscape metrics to:

a. Quantify the change in the patterns of residential use from 1978 to 2006.

b. Quantify potential change in residential land use patterns from 2006 land
use data for two alternative development scenarios:

i. Build out to full capacity allowed by local zoning ordinances.

ii. Build out guided by strategic conservation measures.
2. Evaluate the impact of past and firture residential development patterns
inside the Lapeer State Game Area Dedicated Management Boundary on the
landscape structure of non-built landcover patches (grassland, forest, and
wetland) using landscape metrics.

a. Quantify the change in non-built landcover patches from 1978 to 2006 to
evaluate the following hypotheses:

i. increases in residential landcover patches from 1978 to
2006 altered non-built landcover patch structure by:
1. loss of grassland and wetland landcover patches.
2. gain of forest landcover patches.

b. Quantify the change in non-built landcover patches for two
alternative development scenarios to evaluate the following
hypotheses:

i. projected increases in residential land use alter non-built

landcover patch structure by decreasing grassland, forest,

ii.

and wetland landcover patches.

the implementation of strategic land conservation
measures results in reduced impacts of future residential
development by protecting landcover patches of

ecological significance.

STUDY SITE AND SITUATION
Study Site
The Lapeer State Game Area (LSGA) is located in Lapeer County, Southeast

Michigan. It is composed of 8,246 acres owned by the State of Michigan. The LSGA
Dedicated Management Boundary encompasses 12,956 acres and includes property
owned by the State of Michigan (SOM) and property under private ownership. The
SOM property is surrounded by private property in most areas of the Dedicated
Management Boundary; yet in some areas the Dedicated Management Boundary ends at
the SOM property line. In such areas, the SOM property is directly exposed to current
and future development and consequent safety zones. A 450 foot safety zone exists
around all residential structures, inside of which it is unlawful to discharge a firearm.
The author buffered the LGSA Dedicated Management Boundary by 450 feet to
account for these areas directly exposed to adjacent private property development. The
site for this study is the area that lies inside of this 450 foot buffer of the Dedicated
Management Boundary, which totals 14,907 acres.

The state-owned portion of the LSGA is owned by the SOM and managed by
the MDNR Wildlife Division to “provide benefits to hunters and trappers and to
conserve wildlife resources for the benefit of all citizens and visitors to the State”
(MDNR, 2003). The majority of this area is open to public hunting with the exception
of the field dog trial area, the waterfowl refuge, and the areas that fall inside the safety
zones for adjacent residential structures. The field dog trial area is 474 acres of land set
aside for the training of field dogs and competitive field dog trials. It is unlawful to
possess a firearm loaded with ammunition other than blanks for the purpose of training
dogs. The waterfowl refuge is 727 acres of land closed to hunting year-round and

closed to entry during posted periods. Many other forms of recreation take place in the

7

LSGA, such as fishing, bird watching, and hiking (Nelson et.a1., 2007).

The LSGA lies in the Flint River Watershed. The North and South Branches of
the Flint River converge to form the Main Branch of the Flint River inside of the
LSGA. The study site is located in a physiographic region composed of thick glacial
deposits of the Wisconsinan glacial epoch which reached its maximum extent 18,000
years before present. These deposits are associated with ice proximal positions: end
moraines of coarse textured till, outwash sands and gravels, and glacio-lacustrine clays
and silts. Vegetation prior to European settlement consisted of open forests and
savannas of black and white oak on moraines, beech-maple forests on loamy soils, and
white pine and eastern hemlock scattered along the southern portions of the study site.
Northern white cedar, tamarack, black ash, white pine, and eastern hemlock were
common species found along kettle lakes and swampy depressions associated with

outwash (Leonardi, 2001).

Study Situation

Lapeer County is a rapidly changing landscape. The once dominant agricultural
and forest lands are being converted to residential, commercial, and industrial uses.
Lapeer County is positioned in close proximity to both the City of Flint and the Detroit
Metropolitan areas. In 2000, Lapeer County was designated part of the Detroit
Metropolitan Statistical Area. Lapeer County’s connection to these metropolitan areas
is facilitated by a well developed transportation system. Interstate 69 provides easy
access to the City of Flint and Michigan Highway 24 provides access to the Detroit
Metropolitan Area. As built land uses sprawl across southeast Michigan’s landscape,

moving from urban centers to the exurban fiinges, the population of Lapeer County has

increased. According to the US Census Bureau, from 2000 to 2006 there has been over
a six percent increase in population, from 87,904 to 93,761 people. This population
increase is disproportionately small when compared to the expansion of built land.
While explicit statistics describing this change in Lapeer County are not currently
available, the Land Resource Project did this analysis in 2000 for the Detroit and Flint
Metropolitan Statistical Areas.

The Land Resource Project, conducted by Public Sector Consultants and a
collection of academic institutions in Michigan, worked to understand how future land
use change will impact the States land based economy. The project calculated the land
to population growth rates for the period from 1960 to 1990 for the Flint and Detroit
MSAs as 7:1 and 13:1, respectively. This expansion of built land comes at the cost of
agricultural, forest, grassland, and wetland areas. The Land Resource Project used the
Land Transformation Model, developed at Michigan State University. This model
simulated land use change from 1980 to 2040 based on detailed information
characterizing land use change in 16 counties in Michigan. The Land Transformation
Model projected a drastic increase of built lands throughout the State. The Land
Resource Project showed that past and future changes to Michigan’s landscape could
have serious impacts on the land-based economy that is already struggling.

Two land based industries are especially important in Lapeer County:
agriculture and natural resource-based recreation and tourism. The Land Resource
Project suggests that the number of working farms in Michigan is likely to decrease
from about 42,000 in 2000 to 24,000 in 2040. The total area of farm land in Michigan’s
metropolitan areas is projected to decrease by 25 percent in that same time period. The

medium size farms (50-500 acres) are projected to experience the largest decrease. Of

the land adjacent to the LSGA, 12% is in agricultural use. Over 10% of adjacent
landowners are likely to split and over 23% are likely to sell their land (Nelson et. al.,
2007). This suggests that the findings of the Land Resource Project have merit with
regard to the area surrounding the LSGA.

While the economic impacts of recreation and tourism are difficult to directly
measure, these two industries make a significant contribution to Lapeer County’s
economy and to the quality of life of area residents. The LSGA is the single largest
piece of public land to facilitate natural resource-based recreation in Lapeer County. It
receives heavy use by those who live inside and outside of the County. According to
Nelson et. a1. (2007), in the year 2006 over 67,659 user hours were spent by those who
travel from surrounding areas and over 21,653 user hours were spent by adjacent land
owners recreating in the LSGA.

The LSGA is also an important entity that defines a sense of place for local
residents. The Lapeer State Game Area was one of the top green and open spaces of
importance according the Genesee, Lapeer, and Shiawassee Counties (GLS) Greenlinks
Conservation Needs Assessment conducted in 2006 (Vogt et. al., 2006). It also forms
the backbone of the most ecologically significant potential conservation area in the GLS

region (Hyde, Paskus, & Enander, 2006).

10

LITERATURE REVIEW

This literature review will focus first on the need to control residential
development taking place on the urban fringe in order to maintain ecosystem services
and natural resources provided ecological processes. Second, it will focus on_the
importance of including biological and ecological information in the planning process
in order to direct development patterns away from ecologically significant areas.
Finally, it will conclude with a discussion on the use of landscape ecology to evaluate
changes in landscape structure for both past and future residential development.

What is urbanization? Where does the urban area become the suburban, the
suburban become the exurban, and the exurban become rural? These questions have
been answered differently by nearly all disciplines. Pickett et. a1. (2001) suggests that,
from an ecological perspective, there is little use in distinguishing urban areas from
abutting “wild” lands. They write that a systemic approach, as in urban ecology, is
most valuable to science and resource management. Urban ecology defines an urban
system as one that accounts for the interactions and exchanges between cities and their
surrounding landscapes (Pickett, et. al., 2001). Pickett et. a1. (2001) cites the United
Nations (1993) as estimating that in the next 30 years, over 60% of the world’s
population will live in urban areas. In developed countries around the globe, 75% of
the population lives in urban areas and these areas are physically expanding
(Bierwagen, 2007).

Urbanization is generally defined as “the process by which cities grow or by
which societies become more urban” (The American Heritage, 2005). Suburbanization
is the process by which cities and their surrounding footprints expand to form

multinucleated edge cities (Stem & Marsh, 1997). Edge cities are characterized by

11

isolated areas of residential, commercial centers, and corporate or industrial campuses
(Kaufinan & Marsh, 1997). Forman (1995) identifies what he calls ‘a few common
mosaic sequences’ associated with suburbanization: concentric rings spreading outward
from an urban center; grth along an exurban transportation corridor; and spread from
satellite towns, plus infilling. Exurbanization, is defined as low density residential
development outside the urban service boundary (Theobald, 2007a). Distinguishing
these types of landscape change elucidates the complexity of patterns created by urban
expansion. For the purposes of this paper, urbanization refers to all of these processes
of urban expansion, including urbanization, suburbanization, and exurbanization. In
southern Michigan, urbanization is typically the transformation of forested and
agricultural lands to uses associated with urban areas (i.e. residential, commercial, or
industrial). Urban sprawl can be thought of as consumptive urbanization or the
migration of the urban complex beyond what is needed for current use.

Theobald (2007a) cites Benedict and McMahon (2001), who suggest that land
resources in the United States have been use inefficiently. From 1982 to 1997, Benedict
and McMahon estimate that the rate of land conversions due to urbanization has
outpaced population growth by 47.1% to 17%, respectively. Exurbanization is
occurring at an even higher rate than urbanization and the overall footprint of this is 10
times larger than the high density development: 77.6 vs. 6.7 million acres (Theobald,
2007a)

Human social systems depend on goods and services provided by ecosystems.
Without natural resources such as oil, trees, and water, current social and economic
systems would collapse. Ecosystem alterations and stresses caused by human and

natural actions decrease the ecosystem’s ability to provide natural resources and

12

services. Land use change is the major driving force behind habitat loss and ecosystem
degradation (Merenlender, 2007). Land use changes often result in non-built habitat
fragmentation, isolation, and degradation, simplification and homogenization of species
composition, disruption of hydrological systems, and modification of energy flow and
nutrient cycling (Alberti & Marzluff, 2004). The extent to which non-built areas are
impacted by land use change depends on the spatial configuration of land uses that
results from the physical planning process employed by the agencies responsible for
land management.

As with many States, land management decisions are made at the local level in
Michigan. Local municipalities have the authority to control land use patterns through
the master planning and zoning processes. Typically, these processes tend to focus on
the social impacts of land use patterns, giving less consideration to biology and ecology.
Insufficient consideration of biology and ecology in land use planning is inappropriate
because humans depend on the ecosystem services and natural resources provided by
the landscape. Inherent social and ecological connections necessitate the consideration
of biological and/or ecological information in the local-level land management process.
The challenge, however, is deciding what biological and/or ecological information
needs to be incorporated and how to incorporate it. There are two sides to this issue: 1)
how can local land managers incorporate biological and/or ecological concepts in to the
management process, and 2) how can biologists and ecologists help to facilitate the use
of this information in the land management process.

There are several ways that local land managers can alter the local land management
process so that ecological and biological information is incorporated in to decision-

making land use. Many researchers agree that it is important that land use decisions are

13

made using a landscape context (Theobald, 2007a; Merenlender, 2007; Perlman, 2007).
There are spatial mismatches between the scale at which land is managed, the
boundaries of local municipalities, and the scale of ecological processes. Landscapes
that would be contiguous ecological units are divided amongst tens or even hundreds of
local municipalities. The result is often disconnected land management with small un-
contiguous blocks of protected lands. Opportunities to protect corridors that cross
jurisdictional lines are seldom realized (Berke, 2007). Coordinating planning across
multiple jurisdictions is essential to address the exurban development that occurs at
scales larger than one jurisdiction (Merenlender, 2007). Perlman (2007) writes that it is
important for local land managers to understand the spatial scale of ecological
processes. Perlman asserts that, as regulations and ordinances are developed without an
understanding of the scale of the ecological process they seek to protect, they can miss
the mark. It is not only important to focus on the spatial scale of processes but also on
the temporal scale. Theobald (2007b) writes that alternative development scenarios
should be used to augment planning in order to understand the temporal effects of land
management decisions.

Theobald (2007b) identifies two aspects of an information gap that exists
between the implementation of sustainable goals for landscape planning and the
possible pathways to achieve these goals through land use planning. One aspect is the
need to explore alternative development scenarios. A scenario is defined as a
“hypothetical sequence of events constructed for the purpose of focusing attention on
causal processes and decision points” (Theobald, 2007b). Scenario planning is a
systemic method of exploring the uncertainties of possible futures. The focus is not on

accurate predictions, rather, the focus is on facilitating discussion about planning and

14

management outcomes.

The other aspect of the information gap asserted by Theobald (2007b) is the need to
develop useful and creditable forecasts of future development. One way of creating
future development forecasts is through the use of spatially explicit land use change
models. These models have been developed using several modeling approaches,
including cellular automata (White, Engelen, & Uljee, 1997; Verburga, Koninga, Koka,
Veldkampb, & Bouma, 1999; and Jenerette & Wu, 2001), spatial transition
probablilities (Theobald, 2003), regression based relaionships (Landis, 1995; Bradshaw
and Mueller, 1998), and “studio-based” planning scenarios (Stienz et. al., 1996). One
type of simulation model has been increasingly employed in land use planning: build
out analysis. This simulation projects end states of planning assumptions, such as
zoning ordinances (Theobald, 2007b).

Ecologists and biologists have an important role in facilitating the use of biological
and ecological information in the local land management process. Theobald (2000)
asserts that two of the reasons that biological information is rarely included in the local
land use planning process are: 1) biological information is difficult to access and, 2) it is
unclear to planners how the information can be used. In tending to these two issues, it
is first important for the research agenda of ecologists and biologists to focus on the
development of more reliable indices, models, and tools regarding the impacts of land
cover change on natural resources and ecosystem services. These indices, models, and
tools must be developed in a manner that is usable by local level officials and planners.
This would help overcome the problem of limited access or availability as well as the
problem of uncertainty in use by local level planners.

Ecologists and biologists can help deal with these issues by being directly involved

15

in the land management process. There are several vehicles for ecologists and
biologists to enter and contribute to the land use planning process. They can sit as
members of the planning board or work with advocacy groups, provide educational
opportunities for local decision makers, testify in public hearings, confer with local
municipality staff, and help revise ordinances as part of citizen review panels (Broberg,
2003)

While these measures will help local level planners incorporate biological and/or
ecological information in to the land management process, they could also benefit the
MDNR in carrying out State management objectives presented in their Wildlife Action
Plan (WAP). This plan repeatedly identifies fragmentation as a threat of the highest
priority to Michigan’s wildlife. The identified conservation needs for land, water, and
species management in the WAP includes the incorporation of “identified areas of high
biological significance into local, regional, and statewide planning and management
efforts” (Michigan Department of Natural Resources, 2008). Potential Natural Areas
Assessments are one method of identifying these areas, which has been used
increasingly in Michigan by the Michigan Natural Features Inventory. The Genesee,
Lapeer, Shiawassee GreenLinks Project used this method to identify important
biological resources at the regional scale (Hyde, Paskus, & Enander, 2006). It is the
author’s belief that this method is applicable at a finer-scale, given the use of
corresponding finer-scale land use data.

The MDNR’s WAP was developed based on the principals of Ecosystem
Management presented by Christensen et. a1. (1996). The planning stages in Ecosystem
Management (McGarigal, 1998) share similarities with other ecological-based planning

methodologies advanced as solutions for creating ecologically functional landscapes:

l6

Landscape Planning (Fabos, 1985), Rural Planning (Golley and Bellot, 1999), and
Sustainable Land Planning (Botequilha Leitao and Ahem, 2002). One method
particularly relevant to this research is the evaluation of alternative development
scenarios to understand the impact of proposed management actions on the structure of
the landscape.

As a discipline, landscape ecology is useful for evaluating proposed alternative
development scenarios. It is concerned with the relationship between pattern and
process or structure and function. Landscape ecology can provide the tools needed to
understand the ecological implications of landscape structure. There are two
components of landscape structure: composition and configuration. Composition is not
spatially explicit and does not deal with patch geometry or geographic location.
Composition deals with measures of richness, diversity, evenness, and dominance
(Botequilha Leitao and Ahem, 2002). Configuration addresses patch geometry and the
spatial distribution of landscape elements. Landscape ecologists have developed a
series of pattern metrics to quantify landscape configuration and composition. This
study will focus on five of these metrics: total class area, mean patch size, total core
area, number of patches, and mean proximity index.

The following set of landscape metrics are designed to calculate the composition
of the landscape. The total class area is the sum of all patches of a corresponding patch
type in square meters divided by 4046.9 to convert to acres. The formula is as follows,
where the ag- is the area in square meters of patch ij:

1
TCA= a
1.21 l1(40469]

The mean patch size is the sum of the areas, in square meters, of all the patches

17

of corresponding patch type, divided by the number of patches of the same type, divided

by 4046.9 to convert to acres.

at“

MPS c=j 1 £ 1 j
_ nl- 4046.9

II M:

 

The total core area is the sum of the core areas, in square meters, of all patches

of the corresponding patch type, divided by 4046.9 to convert to acres.

 

n 1
TCA c= a--C
— 121 ’1 (4046.9)

As previously mentioned, landscape configuration deals with the spatially
explicit aspects of the landscape. The important measures of configuration are
landcover patch location and shape. The landscape context in which a landcover patch
resides may have a greater affect on its function and sustainability than does its own
characteristics (Forman, 1995b). The distance between landcover patches and the
nature of the landcover matrix between similar landcover patches will influence species
survival (Ruggiero et. al., 1994; Ander, 1997 cited in Environmental Law Institute,
2003). In this study, the mean proximity index was used to quantify landscape
configuration.

The mean proximity index measures the degree of isolation and fragmentation
of corresponding patch types. The mean proximity index is the sum of patch area in
square meters divided by the nearest edge-to-edge distance squared (square meters)
between the patch and the focal patch of all patches of the corresponding patch type

whose edges are within a specified distance in meters of the focal patch.

18

n a--
MP1: 2 ”S

s=lhijS2

 

The mean proximity index is represented in the calculation above. Here, aijs is
equal to the area of patch ijs within specified neighborhood of patch ij and hi), is equal
to the distance between patch ijs and patch ijs, based on patch edge-to-edge distance,
computed fi'om cell center to cell center. The mean proximity index is 0 when there are
no patches of corresponding patch type within the defined search radius.

Using these metrics, it is possible to quantify changes in landcover that cause
shifts from one mosaic pattern to another. Some landcover classes increase and some
decrease through human and natural processes. In this study, the amount of landcover
associated with residential uses increases and causes changes in the pattern of non-built
landcover including forest, wetland, and grassland patches. In recent decades,
fragmentation of these natural landcover patches has been widely examined as a source
of habitat loss and degradation. In this case, fragmentation is defined as the division of
large landcover patches in to smaller ones. Fragmentation is but one of many spatial
processes that occur when mosaics shift from one pattern to another. Forman (1995)
provides a useful framework for understanding the spatial processes that occur during
the transformation of mosaic patterns due to decreases in undeveloped landcover. This
framework places fragmentation in the broader context of landscape change. Forman
suggests that there are five spatial processes in land transformation that are associated
with loss of landcover patches: perforation, dissection, fragmentation, shrinkage, and

attrition (Figure 2).

l9

Perforation

 

 

 

Dissection am)

an

O

9

n.
Fragmentation g

0

C

(D

t

o

O.
Shrinkage g

100 50 o

Attrition % Of original land type

 

Figure 2. Spatial Processes in Land Transformation. The above graph depicts the relative
importance of each spatial process in relation to the amount of existing original land cover type
(adapted from Forman 1995a).

Perforation and dissection are important spatial processes when the amounts of
original landcover are high, as is typically found in exurban environments (Figure 2).
An example of perforation is low density residential development in a landscape
dominated by forests. Dissection occurs as new roads are built in a forested landscape.
Perforation and dissection decrease the average patch size and total core area of a
landcover type. These two processes impact species with a high sensitivity to edge
conditions or with large area requirements. Fragmentation is the division of landcover
patches in to smaller pieces. Fragmentation is important when half or more of the
original landcover still exists (Figure 2). It is a common spatial process associated with
suburban landscapes. Fragmentation can result in an increased number of generalist

and multi-habitat species (Forman, 1995). Attrition and shrinkage are important spatial
20

processes when the amount original landcover type is low (Figure 2). An example of
attrition is the loss of small patches of woodlots between agricultural fields. Attrition
and shrinkage can cause the loss of the last few remaining species with large area
requirements. However, the actual ecological impacts of the first three spatial processes
are greater. Forman (1995) suggests that the most rapid ecological changes take place
in the first 40% of land transformation. In this period, land planning and conservation
are most important and effective. Landscape ecology can help in understanding the
impacts of the spatial distribution of landscape elements on the function of the
landscape; however, the responsibility of making decisions regarding spatial

distributions falls on land use planners and local municipalities.

21

METHODS
This study documents landscape-level changes in the structure of grassland,

forest, and wetland patches inside the LSGA Dedicated Management Boundary as a
result of increased residential use. The analysis strategy consists of three sections. The
first section is an analysis of the change in landscape-level patch structure from 1978 to
2006. The second section includes build out analysis and construction of two
alternative development scenarios for the LSGA Dedicated Management Boundary: one
based on development at the full capacity allowed by current township zoning
ordinances and the second based on the targeted conservation of ecologically significant
landcover patches in areas where the LSGA Dedicated Management Boundary and the
GLS Greenlinks green infrastructure network intersect. The third section is an analysis
of the potential impacts of the alternative development scenarios on the landscape

structure of grassland, forest, and wetland patches.

22

Land Use Mapping

To analyze the change in landscape structure, a 2006 land use dataset was
compared to the 1978 Michigan Resource Information System (MIRIS) land use
dataset. The 2006 land use dataset was interpreted from digital orthophotos obtained
from the United States Department of Agriculture (USDA). The digital orthophotos
were l-meter resolution leaf-on aerial photography taken during the growing season.
Land use polygons were digitized on screen in a geodatabase using ArcGIS 9 at a scale
of 1 to 5000 and were classified using the MIRIS 2000 scheme. A 2.5 acre minimum
mapping unit was used (Figure 3). In order to delineate wetland polygons, the author
used two datasets: the updated 1996 National Wetlands Inventory obtained from the
Great Lakes Regional Office of Ducks Unlimited and the hydric soils information from
the Soil Survey Geographic (SSURGO) Database obtained from the Michigan Center
for Geographic Information (MCGI). The 1978 MIRIS dataset was interpreted from
National Aerial Photography color-infrared photographs and obtained from MCGI.

The MIRIS classification scheme was developed in 1980 based on the Anderson
Classification system. There were minor changes in the land use classification schemes
between 1978 and 2000, “including the restructuring of the forest groups, new names
for some group headings and categories, the addition of new descriptions of feature
assignments” (MDNR, 1998). While landuse data for both the 1978 and 2006 data were
mapped using more than 30 classes, the differences in classification schemes
necessitated that they be aggregated into 11 common land use categories. The changes
in wetland and forest groups proved to be particularly problematic in this case. Table 1

contains the land use classes that were used in mapping and comparison.

23

3 Residential

. Commecial
- Utilities / Roads
Eff} Extractive

‘ Public

Agricultural

I7 F I - l Grasses / Shrubs
- Forest

- Open Water
- Wetland

l l Baren

 

 

Figure 3. 2006 land use dataset displayed with 40% transparency with the 2006 l-meter
leaf on photography at a scale of 1:5000

24

Table 1. Classes used for land use mapping and comparison

 

 

 

 

 

 

 

 

 

 

 

 

Land Use Class Class Description

Areas maintained as part of a residential parcel
Residential (lawns and garages)

Areas maintained as part of a commercial
Commercial parcel (gas stations and stores)

Areas maintained for use as a outdoor
Community/Outdoor recreation or public building (DNR offices and

Recreation Outdoor Youth Camps)

Areas maintained for utilities (roads and gas
Utilities line corridors)

Areas used for mineral Extraction (gravel pits
Extractive and oil fields)

Areas used for agricultural production (corn
Agricultural fields, cow pastures, and orchards)
Forested Areas with over 25% tree cover

Areas with less than 25% tree cover, also
Grassland includes shrub lands
Wetland
Open Water
Barren

 

 

 

 

Alternative Development Scenarios

In creating the two alternative development scenarios, the author used the Orton
Foundation’s Community Viz 360 for ArcGIS 9. The “Build Out Wizard” module of
Community Viz calculated the spatial arrangements of buildings based on user inputted
zoning parameters, including minimum lot size and setbacks, and constraints based on
spatial data. Three spatial datasets were produced prior to running the “Build Out
Wizardz” zoning boundaries, minimmn lot size, and constraint layers datasets. The
zoning boundaries dataset represented the spatial extent of the zoning classes for each of
the five townships intersecting the LSGA Dedicated Management Boundary: Arcadia,
Deerfield, Marathon, Mayfield, and Oregon Townships (figure 7). This dataset was
created by scanning township zoning maps into jpegs at 600 dots per inch. The jpegs

were geore ferenced to the Framework Version Six Transportation Dataset

25

from the MCGI using the georeferencing toolbar in ArcGIS 9. The resulting zoning
polygons were digitized in ArcGIS 9 on-screen in a geodatabase.

The second spatial dataset was developed to contain the minimum lot sizes. The
third dataset needed for the alternative development scenarios was a constraint layer, or
a spatial dataset containing all areas not open to development. In the full capacity
development scenario, the build out was constrained by areas presently occupied by
built uses identified in the 2006 land use dataset, i.e. residential, commercial, industrial,
and road right-of—way polygons (Figure 9). The road right-of-way polygons were
calculated by buffering the road centerlines in the Framework Version Six
Transportation Dataset from MCGI by 66 feet.

The conservation alternative development scenario necessitated an extra set of
constraints to the build out. This constraint layer addressed the non-built landcover
patches within the LSGA Dedicated Management Boundary that lie outside of the
MDNR owned parcels and inside the GLS Greenlinks green infrastructure network
boundaries. These landcover patches were prioritized based on three criteria: total area,
core area, and presence of water. To prioritize these landcover patches, grassland,
forest, and wetland datasets were extracted from the 2006 land use data and clipped by
the green infrastructure network boundaries in ArcGIS 9 (Figure 4). The grassland and
forest datasets were filtered to include only patches larger than one acre. Because zero
tolerance wetland loss ordinances are common to the townships, all wetland patches
were included in the constraint layer. Grassland and forest patches were prioritized
using an additive scale based on four natural breaks in the in the landscape metrics.
This way, grassland patches were only compared with other grassland patches and

forest patches with forest. Table 2 presents the scoring criteria for grassland and forest

26

patches. A final score for each patch was calculated by adding the scores for each of
the three criteria together. Based on four natural breaks in the scores, the highest
ranking patches for each landscape element was selected and merged, along with all of
the wetland polygons, into one final conservation constraint layer (Figure 4). This final
conservation constraint layer covered 1,880 acres and included seven forest patches
totaling 1,541 acres, six grassland patches totaling 56 acres, and 71 wetland patches

totaling over 282 acres (Figure 5).

27

Table 2. Scoring criteria for the conservation constraints data layer.

Class Metric Score

 

Grassland

3 (3.6-7.9 acres)
(< 1.8 acres)
3 (1.5-2.9 acres)

1 (< O 3 acres)
0 (not present)

 

 

Forest

0 (not present)

 

 

 

28

I»‘

_ Dedicated Management Boundary

' MDNR Parcels ‘

 

"*1 ewe?

A
Conservation Rank ‘
iii—l .
N h a
-

Figure 4. Forest and grassland conservation rankings. Four is the highest conservation
value based on total class area, core area, and presence of water bodies. The top image
displays grassland features and the bottom image displays forest features; both are
displayed with the GLS Greenlinks Green Infrastructure Network boundaries.

k “b ’1’

29

 
 
 
 

1.)”? #1

I:] Dedimted Management Boundary
MDNR Parcels

-yr‘

  
  
 

     
     

  

Forest

-
1 _Wetlands _ a.
2' Mi'e Grassland .

Figure 5. Conservation constraint dataset shown with the GLS Greenlinks Green
Infrastructure Network.

  

Landscape Structure
To calculate the potential change in landscape structure, firture land use maps

were produced for each alternative development scenario. Future land use maps were
created by buffering the build out analysis, a point shape file, by 35.9 meters in ArcGIS.
This was the approximate radius needed to produce a circle with an area of one acre, the
assumed area maintained per residential dwelling, these buffers were then converted to
squares. The square buffers were dissolved, based on common fields, to create two
shapefiles representing potential patterns of residential land use. Using the editor and
topology toolbars in ArcGIS, the new square buffer shapefiles were incorporated into

the 2006 land use dataset as residential polygons for each scenario (Figure 6).

 

 

 

 

 

Figure 6. One acre square buffers used to create the land use datasets for the
alternative development scenarios.

30

The change in landscape structure for 1978, 2006, and the two alternative
development scenarios was analyzed using ArcGIS 9 and the Vector-Based Landscape
Analysis Tools Extension (V-LATE 1.1) developed by the Landscape Analysis and
Resource Management Research Group. Change in the total class area was assessed for
all eleven land use classes. Change in the total class area, mean patch size, total number
of patches, total core area, and mean proximity index were analyzed for non-built
landcover: wetland, forest, and grassland classes. For residential and agricultural land
classes, total class area, number of patches, the mean patch size and mean proximity
index were calculated.

Core areas were calculated using an inside buffer of 100 meters. This parameter was
used in effort to be consistent with the GLS Greenlinks Potential Conservation Areas
Assessment (Hyde, Paskus, & Enander, 2006). For the mean proximity index, a radius
of 15,960 meters was used to specify the neighborhood. This number was calculated

based on the extent of the LSGA’s Dedicated Management Boundary.

31

I;

I a

I an

I; 2%

mango:
081.55 I m; 26

         
             

dzu

Z5825:
.720

 
 

 

>8m§m

£81.55 «in .26

D 42596 wocsamzmm

H 08.8an Zmammmama wocsamq
g :02.» page

  

mam—=6 q. Near—m «Fauna n2. :8 :5 «osmium 3:: as: 5.5:: :5 F382. m3? Oan >3: Gonna—:2— guE—NaBaE vegan—9Q.

32

WU 9:5 0.: 00:98.36
H— anmomfima zmzmmoaoa mocsamq
§\\ :02.» p285

sV\l.
an»! ,

 

    

     

        

 

39.3 a. 0252.153 3 :5 c5:— oi ans—vim. .5; ”Emma Emu—36 2.9%. 8:335 5.5:: .3... =3 vac—«mean.

can :25! 83.33.?

33

Table 3. Zoning types with minimum lots sizes and total area of each zoning within the
Lapeer State Game Area Dedicated Management Boundary

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Minimum Lot Total Area
Township Zoning Class Size (acres) (acres)
Arcadia Recreational (REC) 10.00 2,148.53
Single FamilLResidential (R-l) 0.69 4.86
Agriculture- Residential (AR) 2.00 462.39
Deerfield Single Family Residential (R-2) 1.00 102.03
Residential-Agriculture (R-l) 4.00 995.20
Marathon Residentialflriculture (R-l) 2.00 2,382.66
Mayfield Residential-Agriculture (R-l) 4.00 7,054.68
Single Family Residential (R-2) 0.22 478.55
Oregon Agriculture- Residential (AR) 2.50 1,274.78
Single Family Residential (R-l) 0.75 4.22

 

Limitations of the Study

Discrepancies between the 1978 and 2006 land use dataset serve to limit the
accuracy of the land use change comparison. The 1978 land use dataset was interpreted
from a different aerial photography base than was the 2006 dataset. The 2006 dataset
was digitized using a higher resolution aerial photography at a finer scale. The
differences in scale and resolution can impact some of the landscape metrics.

Limitations in the build out scenarios resulted in over-simplified development
patterns. The full capacity development and conservation development scenarios
assumed zero land transformation outside of the expansion of residential land. This
assumption was necessary because it was not within the scope of this study to produce
models of ecologic processes (i.e. forest succession, old field succession) or to model
changes in the spatial configuration of land use types (i.e., rezoning residential areas as
commercial or industrial). The conservation scenario also did not take in to account
important social criteria necessary to conduct conservation on privately owned lands,
such as property ownership patterns, property values, or risk of development. Finally,

the land use patterns that resulted from the full capacity development scenario assumed

34

standard one acre square footprints associated with each residential dwelling. This
assumption may have over-simplified the configuration of residential lots on the

landscape.

35

RESULTS
Increased Residential Land Use from 19 78 to 2006 and Alternative Development

Scenarios
1978 to 2006
The Lapeer State Game Area (LSGA) is 8,246 acres of land under the ownership

of the State of Michigan. The LGSA Dedicated Management Boundary encompasses a
larger area of 14,907 acres. Inside the LSGA property boundaries owned by the State
of Michigan, landcover is mostly grassland, upland and lowland forest, and wetlands.
Fifty-five percent of the area inside the LSGA Dedicated Management Boundary was in
private ownership in 2006. The majority of land use change occurred on these private
lands during the 28 years from 1978 to 2006. The amount of LSGA Dedicated
Management Boundary lands classified as residential increased from about two percent
(345 acres) in 1978 to about six percent (935 acres) in 2006 (Figure 14). During this
same time period, the number of residential land use patches increased from 98 to 270
(Figure 13) and the mean patch size remained the same at 3.5 acres (Figure 15). The
mean proximity index for residential patches decreased from 1,756.8 to 205.4 (Figure
16). The total class area of lands used for agriculture decreased from 14 percent of the
LSGA Dedicated Management Boundary in 1978 (2,091 acres) to four percent (623
acres) in 2006 (Figure 14). The number of agricultural patches decreased from 133 to
62 (Figure 13) and the mean patch size decreased from 15.5 to 10.1 acres (Figure 15).
The mean proximity index for agricultural lands increased from 14.3 in 1978 to 66.4 in

2006 (Figure 16).

36

ll; _ mmmamsca — ; Gammon \ wracm
003385. n ._ moan

 

I c258 \ 7.0QO I Cue: <<m$_.
I £385 flu <<m=m=a
I page i mean
HH— 318.38.

 

_ H 08.8.23 Zmzmcmama mocnumq
”U 2E0 g :02.» p285

Enema e. 3.; FE:— =mo Eat new :8 H.352. maze Oman >3: 62:2.qu 39552.33 awe—=53?

   

37

a ”enigma. WM 0833 \ myaam
Us 003385. E moafl

I 5.3% \ womam I Ono: <<m6q

I qumogé n <<m=m=a

I 36:0 D mmqm:

D >oq.oc_E_.m_

    

  

. minim xmmaosza 92086.0.
_.IIiIil_ anfimflon stmmmama mecaamQ

\ g 2.02.» p285
A
”3.5

Ens—.0 3. no: been :8 5.6 m3. :8 562:. was? 03:» >3» 62:232. guano—Boa web—395.. minnow
«2.32.2». 3325.8 :6 Emu—58...

38

Full Capacity Alternative Development Scenario

The full capacity alternative development scenario calculated the potential for
1,929 new residential dwellings within the LSGA Dedicated Management Boundary
according to township zoning allowances. The Residential 2 (R2) zoning in Marathon
Township accounted for 42% (812) of the projected residential dwellings. This zoning
class only occupied three percent of the total area of the LSGA Dedicated Management
Boundary; yet, it provided a higher zoning density than any other zoning category in all
five townships: approximately 4.5 dwellings per acre or a .22 acre minimum lot size
(Table 4). The Residential 1 (R1) zoning class in Mayfield Township, the
Agriculture/Residential (AR) zoning class in Oregon Township, and the AR zoning
class in Marathon Township were all similar in their contributions to the final build out
total, with 350, 340, and 323 potential new dwellings, respectively (Table 4). Of all
zoning classes, these three comprised the largest portion of the LSGA Dedicated
Management Boundary. With the 1,929 potential new residential units added to the
existing 892 in the LSGA Dedicated Management Boundary, the potential for a total of

2,821 residential units exists within the LSGA Dedicated Management Boundary.

39

Table 4. Number of potential residential dwellings calculated at full capacity build out per
zoning class and minimum lot size expressed in acres.

 

 

 

 

 

 

 

 

 

 

 

 

Minimum Build Out
Township Zoning Class Lot Size Dwellings
Arcadia Recreational (REC) 10 0
Residential 1 (R1) 0.69 0
Agricultural Residential (AR) 2 94
Deerfield Residential 1 (R1) 4 0
Residential 2 (R2) 1 10
Mayfield Residential 1 (R1) 4 350
Residential 1 (R2) 0.22 812
Marathon Agricultural Residential (AR) 2 323
Oregon Residential 1 (R1) 0.75 0
Agricultural Residential (AR) 2.5 340
Total 1,929

 

 

 

 

 

 

40

D $8.335. _ 1H mammom \ madam

I 5.58 \ womam I Ono: <<m84
I @9885 m. ”52.25
I page i mean

320538.

 
 

  

   

“— 08.022. Zmzmcmama onamQ
.. 2.022 3285

 

Ens—d :. 3:— 25:23. 383376 ago—evaga mangle .25 :8 53. men :5 E352. maze 09:; >3» 62:232—
Z—auwNoEoun 9.53an

41

The full capacity alternative development scenario build out analysis produced a
potential land use map with 2,359 acres of residential land, based on a buffer of 1.4
acres (Figure 1 1). The number of residential land patches increased by 316 to total 586
patches (Figure 13) and the mean patch size increased from 0.5 to 4.0 acres (Figure 14).
The mean proximity index increased substantially from 205.4 to 4,653l.1 (Figure 16).
Agricultural lands within the LSGA Dedicated Management Boundary decreased by
157 acres to 467 acres (Figure 14) while the number of agricultural landcover patches
increased from 62 to 95 (Figure 13) and mean 467 patch size decreased from 10.1 to 4.9
acres (Figure 15). The mean proximity index for agricultural landcover patches

increased from 66.4 to 91.5 (Figure 16).

Conservation Alternative Development Scenario

The conservation alternative development scenario revealed the potential for
1,430 new residential dwellings in the LSGA Dedicated Management Boundary. The
R2 zoning class in Mayfield Township accounted for 710 residential dwellings, the
largest increase of all zoning classes inside the LSGA Dedicated Management
Boundary. The build out analysis in this scenario calculated a potential of 263 new
dwellings in the AR zoning class in Marathon Township and 212 new dwellings in the
R1 zoning class in Mayfield Township. Other zoning classes revealed less capacity for
new dwellings (Table 5). The projected 1,430 new residential units, in addition to the
existing 892, create the potential for 2,322 residential units inside the LSGA Dedicated

Management Boundary in the conservation alternative development scenario.

42

flu momaoszm. NH 0838 \ mzacw
003385. .11; .382

I csaom \ momam I. one: $22
I mxchZo
I 35.6 ill mean
_ . _>onoc=ca_

      

 

 
  

d I 083228: Oosmgim
D 08828 Zmzmmmama onamQ

2.02.» 3485

_L

 

H i=0

5%.: S. Gonna—£33: 192.5375 99.23.52: 82.2.8 .23 :8 BE. 3.. :3 F352. was; 0.qu >3» 02:389.
assume—Bonn wax—5.34.

43

The conservation alternative development scenario build out analysis produced a
potential land use map with 1,946 acres of residential, land based on the one acre square
buffer (Figure 12). There was an increase of 1,011 acres of residential land from the
2006 dataset (Figure 14). Residential land patches increased by 300 to total 570 patches
(Figure 13) with a mean patch size that increased from 0.01 to 3.6 acres (Figure 15).
The mean proximity index increased substantially from 205.4 to 1,385.1 (Figure 16).
The amount of agricultural lands inside the LSGA Dedicated Management Boundary
decreased by over 159 acres to 464 acres (Figure 14) while the number of agricultural
land patches increased from 62 to 99 (Figure 13) and the mean patch size decreased
from 10.1 to 4.7 acres (Figure 15). The mean proximity index for agricultural land

patches increased from 66.4 to 517.2 (Figure 16).

Table 5. Number of potential residential dwellings calculated at build out for the
conservation development scenario per zoning class and minimum lot size, expressed in

 

 

 

 

 

 

 

 

 

 

 

 

acres.
Minimum Build Out
Zoninfilass Lot Size Dwellingi
Arcadia Recreational (REC) 10 0
Residential 1 (R1) 0.69 0
Agricultural Residential (AR) 2 83
Deerfield Residential 1 (R1) 4 0
Residential 2 (R2) 1 10
Mayfield Residential 1 (R1) 4 212
Residential 1 (R2) 0.22 710
Marathon Agricultural Residential (AR) 2 263
Oregon Residential 1 (R1) 0.75 0
Agricultural Residential (AR) 2.5 90
Total 1,430

 

 

 

 

 

 

44

Number of Patches

 

 

570 586
——I
270
98
’4 —I
133
62 95 99
1978 2006 Capacity Conservation
+Residential —I—Agriculture

Figure 13. Change in the number of patches for built land use classes inside the Lapeer State
Game Area Dedicated Management Boundary from 1978 to 2006 and for the full Capacity and
Conservation alternative development scenarios.

Total Class Area

 

 

2359
1946

2091

935
345

‘I I

623 467 _ 464

1978 2006 Capacity Conservation
+Residential +Agriculture

Figure 14. Change in the total class area for built land use classes inside the Lapeer State
Game Area Dedicated Management Boundary from 1978 to 2006 and for the Capacity and
Conservation alternative development scenarios.

45

Mean Patch Size

 

 

15. 5
10.1
4-9 4.7
I I—f + :-
3.5 3.5 4 73.6
1978 2006 Capacity Conservation
—l—Residential +Agriculture

Figure 15. Change in the mean patch size for built land use classes inside the Lapeer State
Game Area Dedicated Management Boundary for 1978 to 2006 and for the full Capacity and
Conservation alternative development scenarios.

Mean Proximity Index

  

517.2

‘iti‘ws- an:

 

 

 

 

 

 

Conservation l:;._—.._,...__....___-_ “mug 1385 , 1
i
Capacity :91'5 J
2006 -6é35.4
1978lli'1’l:3 . .__ __ l1756.8
l Residential I Agriculture

Figure 16. Change in the mean proximity index for built land use classes inside the Lapeer
State Game Area Dedicated Management Boundary for 1978 to 2006 and for the full Capacity
and Conservation alternative development scenarios. The bar representing the mean
proximity index for the residential land use class is not to scale.

46

Change in the Landscape Metrics for Non-Built Land Cover Classes

Number of Patches

The number of patches is the count of all patches of a corresponding patch type.
From 1978 to 2006, the number of grassland patches decreased by 48: from 198 to 150
patches, respectively. From the 150 grassland patches identified for 2006, the full
capacity alternative development scenario produced an increase of 15 grassland patches
while the conservation alternative development scenario produced an increase of 2
grassland patches (Figure 17). Between 1978 and 2006, the number of forest landcover
patches increased by 79 to total 190 forest patches. From the 190 forest patches
identified for 2006, the full capacity alternative development scenario produced an
increase in the number of forest patches of 41. From the 2006 landcover dataset, the
conservation alternative development scenario produced a larger increase in the number
of forest patches, with 157 additional forest patches (Figure 17). From 1978 to 2006,
the number of wetland patches decreased by 23 to total 124 patches. From the 124
wetland patches identified for 2006, the full capacity alternative development scenario
produced an increase in the number of wetland patches of 20 patches. There was no
change in the number of wetland patches resulting from the conservation alternative

development scenario (Figure 17).

47

Number of Patches

400 -— — — _ - -_ -_ .- . _ . -_ __ __ .
l I
300 1 7-, .. -. .. . .__ 1
200 i‘ A" . ST, "’ _ : _. f . ‘r 7 i ,. ,
1 ‘ ’2 X .
100 »"'V i 7 7 7 7 , ,
O i i i i i ' 7 Z " ‘ '3
g 1978 % 2006 1 Capacity lConservation;
é 3 ~ -- - ~ -
0 Grassland: 198 ; 150 l 165 l 177
T .. Forest l 111 E 190 i 231 ! 347
: 1 ,
j A Wetland 147 - l 124 l 144 i 124

 

Figure 17. Change in the number of patches for non-built landcover types inside the
Lapeer State Game Area Dedicated Management Boundary for 1978, 2006, and the full
Capacity and Conservation alternative development scenarios.

48

Total Class Area

The Total Class Area (TClA) is a measurement of the total area of all patches of a
corresponding patch type in acres. From 1978 to 2006, the TClA of grassland
landcover patches decreased by 799 acres, to total 1,293.8 acres of grassland. From the
2006 TClA, the full capacity alternative development scenario decreased the grassland
TClA by 122 acres and the conservation alternative development scenario decreased the
grassland TClA by 108 acres (Figure 18). Between 1978 and 2006, the TClA for forest
landcover patches increased by 1,652 acres to total 9,737.0 forested acres. The full
capacity alternative development scenario reduced the 2006 forest TClA by 1,041 acres.
The conservation scenario reduced the 2006 forest TClA by 717 acres (Figure 18).
From 1978 to 2006, the TClA of wetland patches increased by 105 acres to total 1,168.1
acres. The full capacity alternative development scenario reduced the 2006 TClA for
wetland patches by 76 acres while the conservation alternative development scenario

produced no change in the TClA of wetland patches (Figure 18).

49

Total

Class Area

 

 

 

 

 

15,000.0 ‘

10,000.0 - _ .-.- I .

5,000.0

...
0.0 A t ------- a a
1978 2006 Capacity Conservation

1. Grassland 2,092.5 1,293.8 1,172.0 1,186.1
I- Forest 8,084.9 9,737.0 8,695.7 9,020.3
A wetland 1,063.3 1,168.1 1,091.9 1,163.6

 

 

 

Figure 18. Change in the total class area for non-built landcover classes inside the Lapeer
State Game Area Dedicated Management Boundary for 1978, 2006, and the full Capacity
and Conservation alternative development scenarios.

Mean Patch Size
Mean Patch Size (MPS) is the average size of all the patches of a corresponding

patch type. From 1978 to 2006, the MPS for grassland patches decreased by 1.9 acres
to produce a MPS of 8.6 acres (Figure 19). From the 2006 landcover data, the full
capacity alternative development scenario decreased the MPS for grassland patches by
1.5 acres while the conservation alternative development scenario decreased the MPS
for grassland patches by 1.9 acres (Figure 19). From 1978 to 2006, the MPS of forest
landcover patches decreased by 21.6 acres. From 2006, the full capacity alternative
development scenario decreased the MPS for forest landcover patches by 13.6 acres
while the conservation alternative development scenario decreased the MPS for forest
patches by 24.2 acres (Figure 19). The MPS for wetland landcover patches increased
by 2.2 acres from 1978 to 2006. From 2006, the full capacity alternative development
scenario decreased the MPS for wetland patches by 1.8 acres while the conservation

alternative development scenario produced no change in the MPS for wetland patches

50

(Figure 19).

Mean Patch Size

100.0 1 e e e _ i i i i 1 f i
i I-
50.0 1- e n
1 1 .1
n
O . O J; , ‘ i - Mr --—-A ~ —— ‘ . __.?.. _ ..._ ,r_.- .. .__-_ . . ,I _ Q -— E
l 1978 I 2006 1 Capacity lConservationl
; e Grassland; 10.6 i 8.6 1 7,1 l 6.7 l
* ‘ . 1
;- Forest 1 72.8 1 51.2 1 37.6 1 26.0 1
E A Wetland j 77.27fl J 79:4 i, 7:6 _, i ##9.2 i j

Figure 19. Change in the mean patch size for non-built landcover types inside the Lapeer
State Game Area Dedicated Management Boundary for 1978, 2006, and the full Capacity
and Conservation alternative development scenarios.

Total Care Area
The Total Core Area (TCA) is area greater than 100 meters fiom the edge of all

patches of a corresponding patch type. From 1978 to 2006, the TCA for grassland
patches inside the LSGA Dedicated Management Boundary decreased by 350 acres,
from 859.5 to 510.5, respectively. From the 2006 landcover data, the full capacity
alternative development scenario decreased the TCA by 61 acres and the conservation
alternative development scenarios decreased the TCA for grassland patches by 52 acres
(Figure 20). Between 1978 and 2006, the TCA for forest landcover patches increased
by over 1,444 acres, from 6,070.7 to 7,114.6 acres, respectively. From 2006, the full
capacity alternative development scenario decreased the TCA for forest patches by
1,444 acres and the conservation alternative development scenario decreased the TCA
for forest patches by 825 acres (Figure 20). From 1978 to 2006, the TCA for wetland
patches increased by 77 acres, from 274.5 to 352.3 acres, respectively. From the 2006

landcover data, the full capacity alternative development scenario decreased the TCA

51

for wetland patches by 39 acres and the conservation alternative development scenario

produced no change in TCA for wetland patches (Figure 20).

Total Core Area

10,000.0 3
1 , .7 I . I
I' ' .7 ,
5,000.0 '
0 _ O l, X 3 ' ‘ t' ' , -* W t . - :~ ~ g 3
3 1978 5 2006 3 Capacity 3Conservationl
l f f 1 7 ‘
i o Grassland? 859.5 510.5 3 449.5 3 458.5 3
;. Forest 5 6,070.7 3 7,114.6 j 5,670.3 l 6,289.3 '
3 A Wetland 1 274.5 3 3752.3 3 312.8 3 7352.3

1

Figure 20. Change in the total core area for non-built landcover types inside the Lapeer
State Game Area’s Dedicated Management Boundary for 1978, 2006, and the full Capacity
and Conservation alternative development scenarios.

Mean Proximity Index

The Mean Proximity Index (MP1) measures the degree of isolation and
fragmentation of the corresponding patch type. It equals zero when there are no patches
of corresponding patch type within the specified search radius (15,960 meters) and
increases as the number of patches within that specified radius increases. From 1978 to
2006, the MP1 for grassland landcover patches decreased from 203.4 to 106.1,
respectively. From the 2006 landcover data, the MP1 for grassland patches increased to
1,721.36 in the full capacity alternative development scenario. In the conservation
alternative development scenario, the MP1 for grassland patches increased to 255.51
from the 2006 landcover dataset (Figure 21). From 1978 to 2006, the MP1 for forested
landcover patches increased from 1,577.6 to 4,143.2, respectively. In the full capacity
alternative development scenario, MPl for forest patches increased to 145,703.6 and the
conservation alternative development scenario increased the MP1 for forest patches to

20,2078 (Figure 21). From 1978 to 2006, the MP1 for wetland patches decreased from
52

7,835.8 to 50.4, respectively. From 2006, the full capacity alternative development
scenario increased the MP1 for wetland patches to 9,062.8 while the conservation

alternative development scenario produced no change in the MP1 for wetland landcover

patches (Figure 21).
Mean Proximity Index
30000 3 3 3
20000 3 f u
3 3’
10000 3 ‘ 3'" g
3 d , -1- H 1
O f, ‘ _._._______ _ 3,A_3 . . _-_. .i.. -' -— -' ' ' . ~ ,_ —- __....__. ,_ ,3. , 3
3 1978 3 2006 3 Capacity Conservation3
3 . Grassland3 203.38 3 106.07 3 1721.36 3 255.51 :
? ‘: 3 i
- Forest 3 1577.6 1 4143.22 3 145703.59 3 20207.75 3
* l ' 1 : ‘ 3
3 . Wetland 3 7835.82 50.43 3 9062.08 3 50.43 3

Figure 21. Change in the mean proximity index for non-built landcover types inside the
Lapeer State Game Area’s Dedicated Management Boundary for 1978, 2006, and the full
Capacity and Conservation alternative development scenarios. Data point for the
Capacity Scenario is not shown

53

DISCUSSION
Urbanization and Landcover Change from 19 78 to 2006

From 1978 to 2006, the predominant land use change within the Lapeer State
Game Area Dedicated Management Boundary was the conversion of agricultural lands
to residential lands. This land use change not only altered the proportion of the LSGA
Dedicated Management Boundary occupied by these two landcover types; it changed
the spatial patterns exhibited by both land use classes. The drastic difference in the
mean proximity index for both of these classes indicates that there was a shift in the
type of residential development taking place. In 1978, there were 98 patches of
residential land in a dense pattern, indicated by the high mean proximity index
(1,756.8). As the number of residential land patches increased to 270 in 2006, the
residential foot print on the landscape increased and the pattern of residential land use
became more dispersed, indicated by the mean proximity index of 205.4. The increase
in residential area was not a result of large subdivision development; it was a result of
the development of individual homes.

Urbanization has produced marked changes in the built landscape patterns inside
the LSGA Dedicated Management Boundary. The first hypothesis of this research was
aimed at evaluating whether these changes affected the non-built landcover classes by
decreasing the numbers of grassland and wetland patches and increasing the number of
forest patches. The pattern of grassland patches inside the LSGA Dedicated
Management Boundary changed ways indicative of attrition and shrinkage. The
grassland landcover class exhibited a decrease in the number of patches, total class area,
mean patch size, and total core area (table 7). As Forman (1995a) suggests, attrition

and shrinkage are most important when there is less than half the original landcover

54

type remaining. In this case, the grassland patches on the periphery of the LSGA are
experiencing attrition as a result of residential development and forest succession.
Shrinkage is most often experienced by grassland patches in the interior of the LSGA as

forest succession occurs at their outer edges due to lack of management.

Table 6. Generalized behavior of landscape metrics from 1978 to 2006 for non-built
landcover classes inside the Lapeer State Game Area Dedicated Management Boundary.

 

 

 

 

 

Number Total Mean Total Mean
Class of Class Patch Core Proximity
Patches Area Size Area Index
Grassland - - - - -
Forest + -
Wetland - + + +

 

 

 

 

 

 

 

From 1978 to 2006, the patterns of forest and wetland landcover patches
changed, also: the numbers and extent of wetland and forest patches increased inside the
LSGA Dedicated Management Boundary. Forest patches increased by over 1,600 acres
with the addition of 76 new patches. The increases in forest and wetland patches were
coupled with decreases in mean patch size, which suggests that relatively small patches
were added. The expansion of forest landcover is related, in part, to the succession of
grassland patches to forest. Also, in agricultural areas being converted to residential
developments, portions of the residential lot not maintained as lawn typically succeed to
forest. The increase in wetland landcover patches may be related to wetland hydrology
returning to areas previously drained for agriculture. According to the MDNR
Management Plan for the LSGA, some increases in wetland area were the result of
management actions. A portion of this change may also be due to differences in data

interpretation and quality between land use datasets.

The gains in forest landcover create habitat for species that are not sensitive to

55

edge conditions and species with small area requirements. The landscape was
simplified and many patches of forest were perforated. Forest patch perforation alters
the patch structure by reducing the total area, increasing the edge effects and it also
introduces new disturbances. Many game species (i.e. white tail deer, ruffed grouse,
and pheasant) use edge habitats between forest and grassland. The edges between
residential lots and forest do not necessarily function as habitat for these game species.
From this study, it is difficult to quantify how much of the increase in forest landcover
actually occurred within the boundaries of the MDNR property and to what extent
MDNR management actions facilitated this increase in forest landcover. Based on
visual inspection of the maps depicting land use change, the majority of the landcover
change appears to have taken place outside of the LSGA State owned parcel boundaries.
Also, agricultural land has some habitat value for many of these game species. This
habitat value is significantly decreased or lost with the conversion of agricultural lands

to residential lands.

56

Alternative Development Scenarios

The two alternative development scenarios were intended to facilitate discussion
on the possible futures of the LSGA. Alternative scenarios can help land managers deal
with uncertainty in real-world natural resource management issues. The intent of this
section was to evaluate the usefirlness of employing landscape metrics to understand
and compare differences in landscape structure as a result of potential development
scenarios. The full capacity alternative development scenario was based on actual
planning documentation from the five townships surrounding the LSGA. The
conservation alternative development scenario was intended to demonstrate a future
where conservation measures are implemented.

The use of landscape metrics in the planning process is one way to incorporate
biological information into the planning processes of the MDNR and the local
municipalities. There were two objectives for this section. The first objective was to
test the hypothesis that projected increases in residential land use inside the LSGA
Dedicated Management Boundary under the full capacity alternative development
scenario would produce changes in the structure of non-built landcover patches
associated with decreasing non-built landcover. The second objective was to test the
hypothesis that implementing strategic conservation measures to limit increased
residential use on ecologically significant non-built landcover patches under the
conservation alternative development scenario would mitigate the impact of
urbanization on the landscape structure. The author will discuss each of the two

alternative development scenarios followed by a comparison of the two.

57

Capacity Scenario Potential Land Use Change and Impacts to Non-built Landcover
The full capacity alternative development scenario calculated over a 216%
increase in the number of residential dwellings within the Lapeer State Game Area
Dedicated Management Boundary, using the full capacity of adjacent township zoning
allowances. Assuming a standard area of residential dwellings to be one acre actively
maintained as a yard, there would be a 150% increase in the total area and 117%
increase in the number of patches of residential landcover in the full capacity
development scenario. The pattern of this projected development is more dispersed
than the pattern of residential use in 2006, indicated by an increase in the mean
proximity index. Under the fill] capacity alternative development scenario, agricultural
lands within the LSGA Dedicated Management Boundary would be effectively
eliminated with mean patch sizes dropping from 10.1 to 4.9 acres of agricultural lands.
The projected mean patch size of agricultural land is less than one acre larger than the
mean patch size of residential landcover patches. The projected pattern of residential
landcover patches does alter the structure of the non-built landcover patches within the

LSGA Dedicated Management Boundary.

Table 7 Generalized behavior of landscape metrics from 2006 to full capacity alternative
development scenario for non-built landcover types within the Lapeer State Game Area
Dedicated Management Boundary.

 

 

 

 

Number Total Mean Total Mean
of Class Patch Core Proximity
Class Patches Area Size Area Index
Grassland + - - - +
Forest + - - - +
Wetland + - - - +

 

 

 

 

 

 

 

 

The first stated objective of this section was proven: projected increases in

residential land use based on build out at full capacity do alter the structure of grassland,

58

forest, and wetland landcover. Projected changes in grassland patches exhibit the
characteristics of fragmentation, including an increase in the number of patches and
mean proximity index, and decrease in the mean patch size and core area (Table 7).
The number of grassland patches increased by 15, producing a loss in total area of 122
acres, a decrease in patch size of 1.5 acres, and a loss of 61 acres of core area. The
decrease in total area is equal to almost 10% of the total grassland area within the
LSGA Dedicated Management Boundary. Over 75% of the remaining grassland
patches lay within the LSGA parcel boundaries; those within the surrounding LSGA
Dedicated Management Boundary would be small and fragmented. These changes in
spatial patterns have serious implications for many species that use large patches of
grassland. Such decreases in suitable habitat in the landscape surrounding the LSGA
could thwart restoration efforts within the Lapeer State Game Area.

Forest patches also exhibited fragmentation in the full capacity alternative
development scenario. The scenario projected decreases in total class area, mean patch
size, and core area, along with increases in the number of patches, total edge, and mean
proximity index (Table 7). The majority of the land use change in the full capacity
development scenario was from forested landcover to residential landcover. There was
a 1,041 acre decline in total forest area and a 25% decrease (over 1,444 acres) in core
habitat. The mean proximity index for forest patches increased from 4,143.2 to
14,5706, indicating that the degree of isolation decreased with the increase in number
of small patches. These changes in landcover patterns can impact both game and non-
game wildlife species.

Fragmentation was the predominant spatial process occurring on the wetland

patches within the LSGA Dedicated Management Boundary in the full capacity

59

alternative development scenario. Land use change associated with wetland landcover
was similar to that exhibited by the grassland patches in the full capacity development
scenario. This similarity is most likely related to the small percentage of the actual
landscape occupied by both landcover classes.
Conservation Scenario Potential Land Use Change and Impacts to Non-Built
Landcover

The conservation alternative development scenario calculated the potential for a
116% increase in the number of residential dwellings within the Lapeer State Game
Area Dedicated Management Boundary, if large areas of ecological value were
conserved. The increase in number of residential dwellings would allow a 107%
increase in the total area and 111% increase in the number of patches of residential
landcover, assuming a standard residential dwelling area of one acre actively
maintained as yard. The pattern of this projected development is dispersed, as indicated
by the increase in the mean proximity index. Agricultural lands within the LSGA
Dedicated Management Boundary would be effectively eliminated under the
conservation alternative development scenario, with mean patch sizes dropping from
10.1 to 4.7 acres of agricultural landcover. In the conservation alternative development
scenario, the projected pattern of residential land use alters the structure of the
grassland, forest, and wetland patches inside the LSGA Dedicated Management

Boundary.

60

Table 8. Generalized behavior of landscape metrics from 2006 to conservation
development scenario for non-built landcover classes within the Lapeer State Game Area
Dedicated Management Boundary

 

 

 

 

 

Number Total Mean Total Mean
of Class Patch Core Proximity
Class Patches Area Size Area Index
Grassland + - - - +
Forest + - - _ +
Wetland + - - - +

 

 

 

 

 

 

 

The second objective of this section was to test whether targeted conservation
measures could mitigate potential alterations to the structure of grassland, forest, and
wetland patches. Clearly, based on the changes in the landscape metrics, this targeted
conservation scenario did not eliminate the projected impacts of residential
development. However, targeting wetland patches and non built landcover patches with
high total and core area did mitigate some impacts of residential development.
Projected changes in grassland patches exhibited the characteristics of fiagmentation
under the conservation alternative development scenario, with increases in the number
of patches and total edge, and decreases in the mean patch size and core area. The
number of grassland patches increased by 27 with a loss in total area of 108 acres, a
decrease in patch size of 1.9 acres, and a loss of 52 acres of core area.

Forest landcover was also fragmented in the conservation alternative
development scenario, with projected decreases in total class area, mean patch size, and
core area along with increases in the number of patches, total edge, and mean proximity
index (Table 8). As with the fill] capacity alternative development scenario, the
majority of the land use change in the conservation alternative development scenario

was from forested to residential landcover. As a result of the conservation constraints

in this scenario, the decline in total forest area was 717 acres and the decline in core

61

area was 825 acres. Forest patches were less isolated in this scenario due to the
fragmentation of large patches. The mean proximity index increased from 4,143.2 to
20,207.7. The conservation alternative development scenario produced no change in
wetland metrics because they were constrained from the build out.
Alternative Development Scenario Comparison

In order to demonstrate the impacts that conservation efforts could have within
the Lapeer State Game Area Dedicated Management Boundary, this section will
compare patterns of grassland, forest, and wetland patches produced by the two
alternative development scenarios along with those from 1978 and 2006. Conservation
efforts cannot wholly negate the impacts of land use change on grassland, forest, and
wetland patches. However, this comparison reveals how targeting specific attributes of
the landscape with conservation measures (i.e. preservation of large patches, large core
areas, and wetlands) can reduce the impacts of increased residential use. This section
will first compare the composition, then the configuration, of grassland, forest, and
wetland patches inside the Lapeer State Game Area Dedicated Management Boundary.

The composition of forest patches changed greatly from 1978 to 2006. The
number of patches, total class area, and total core area exhibited substantial increases.
While there was a 17% increase in the amount of forest landcover, there was a 160%
increase in the amount of residential landcover. This trend of rapid expansion in
residential land has serious consequences for the non-built landscape in and around the
LSGA. The conservation alternative development scenario displayed noticeable
improvements in non-built landscape composition amidst residential development. In
the conservation alternative development scenario, the number of forest and grassland

patches remaining on the landscape increased. (Figure 15). The increases total forest

62

area and total forest core areas were not surprising, due to the implementation of
measures that targeted larger non-built patches for conservation (Figure 16 and Figure
18). The mean patch size for forest and grassland patches was less in the conservation
alternative development scenario than in the full capacity alternative development
scenario. This is because the conservation scenario constrained residential land use on
relatively small grassland and forest patches outside the parcel boundaries of the LSGA,
compared to the patches found and protected inside the LSGA parcels. In the full
capacity alternative development scenario, attrition takes place on the relatively small
grassland and forest patches found outside the LSGA parcels, which results in a higher
mean patch size (Figure 19).

While the composition attributes of grassland and forest patches were targeted
with conservation measures in the conservation alternative development scenario, the
conservation measures also impacted the configuration of grassland and forest patches.
The conservation alternative development scenario showed that by conserving large
habitat patches outside of the boundaries of the LSGA, the landscape-wide mean
proximity index was greatly decreased (Figure 19). The decrease in mean proximity
index resulted from the preservation of several large blocks of forest and grassland.

Through this study, the author has demonstrated that landscape ecology can
provide a useful set of metrics to evaluate alternative development scenarios for the
purpose of land use and management planning. Landscape metrics can also provide an
effective means of linking and incorporating biological and ecological information into

the MDNR and local municipalities’ planning processes.

63

MANAGEMENT CONSIDERATIONS AND FUTURE RESEARCH

This research demonstrates some of the incentives for guiding residential
development around areas of ecological significance in and around the Lapeer State
Game Area. This research also shows that if targeted conservation measures based on
ecologic criteria are implemented they can be an effective method to guide residential
grth patterns around ecologically important areas. The MDNR and surrounding
municipalities have tools in place that can be used to guide and control residential
growth patterns. There are opportunities for these two entities to collaborate and
engage in innovative conservation efforts. This section includes a discussion of the
tools and opportunities available for managing residential grth in the LSGA
Dedicated Management Boundary.

The MDNR has a set of tools specializing in interactions with individual land
owners. These tools can provide interesting opportunities for the MDNR inside the
LSGA Dedicated Management Boundary. As this study has shown, high quality forest
lands surround the LSGA. Almost 13% of the land that is adjacent to the LSGA is
vacant and 12% is used for agriculture (Nelson et. al., 2007). In areas adjacent to State
Game Areas across southern Michigan, over 23% of landowner respondents indicated
that they had an even chance or were very likely to sell their properties in the next five
years (Nelson et. al., 2007). The presence of vacant and agricultural lands adjacent to
State Game Areas, along with adjacent landowner willingness to sell their properties,
creates opportunities and a sense of urgency for innovative land use planning. With
increased residential development likely to continue on the lands surrounding the
LSGA, the price of land is likely to increase while the availability decreases.

One tool available to the MDNR is the fee-simple acquisition of land parcels

64

inside the LSGA Dedicated Management Boundary. Fee-simple acquisition of lands
can be accomplished using funds that are currently available, such as the Michigan
Natural Resource Trust Fund (MN RTF ). The MNRTF pay for over 20 million dollars
worth of land purchases each year for the purposes of conservation and recreation in the
State of Michigan. Another tool available to the MDNR is the Land Owner Incentive
Program. This program compensates land owners for the conservation of private
property. The fee-simple acquisition of lands and Land Owner Incentive Program are
often limited to opportunistic application. This thesis demonstrates one possible
method for the MDNR to prioritize available lands by targeting landcover patches of
high ecologic importance and patches at high risk of being developed. Such land
prioritization could allow the MDNR to use these tools in a strategic conservation
design.

The tools available to local municipalities are focused on guiding broader
patterns of development. Local municipalities have regulatory tools, including the
master planning and zoning processes, which dictate land management within their
jurisdictions. Local municipalities also have the Park, Recreation, and Greenspace
planning processes, which can be incorporated into their land management system to
guide residential development. The cost of rapid residential development can be
destabilizing to the fiscal resources of local municipalities; herein lies the incentive for
managing land use change. The long term costs associated with providing services to
residential areas can quickly outpace the tax revenues created by those residential areas.
Municipalities often attempt to control the amount of residential development through
zoning measures, such as agricultural zoning with the oft stated goal of preserving rural

character. These measures tend to be ineffective at controlling growth. Part of the

65

difficulty in controlling growth is the lack of technical resources for incorporating
biological criteria into rural community zoning ordinances.

There are potential benefits for local municipalities and the MDNR to
collaborate in innovative ways to control growth inside the LSGA Dedicated
Management Boundary. By utilizing the resources that both entities have at their
disposal, strong growth control measures can be developed. With the expertise of the
MDNR, local municipalities could develop conservation based zoning ordinances that
target specific species and ecologic communities of importance. Also with the expertise
of the MDNR, local park, recreation, and greenspace plans could reflect the ecosystem
management goals for the LSGA. This would provide the opportunity for increased
leverage among both entities for finding sources such as the MNRTF. The
development of easy to use tools such as the Potential Conservation Areas Assessment
by the Michigan Natural Features Inventory can serve as the vehicle to include
biological information into the planning process. By engaging municipalities with
conservation based zoning ordinances and local non-profit groups (e. g. land trusts), the
MDNR could help in identifying ecologically important pieces of the landscape and
ensure that they remain intact.

Third party non-profits such as land trusts, resource advocacy groups, and
science-based conservation organizations can serve as helpful intermediaries to work
between MDNR biologists and land managers and local municipalities. In Lapeer
County, there are several groups that could act as intermediaries. For instance, the
conservation alternative development scenario used the GLS Greenlinks green
infrastructure network boundaries to identify the areas considered for conservation. The

GLS Greenlinks green infrastructure network could contribute to the facilitation of

66

conservation measures within the LSGA Dedicated Management Boundary because it
considers social preferences for conservation as well as biological criteria.

Considerations for future research should include alternative development
scenarios that account for landscape change within the LSGA boundary based on forest
and old field succession, natural disturbance, and future management action. Research
that seeks to understand threshold responses to changes in landcover for game and
nongame species important to the LSGA would also be useful. Such research would
allow managers to further their understanding of the potential impacts of future
development. Future research could better address the types of development that are
actually occurring. This study assumed that all areas would be built-out at 100% and
that all lot sizes were one acre. Both of these assumptions vary greatly in their accuracy
based on the zoning class. Similar studies at coarser scales would also be important in
understanding broad development patterns and the impacts of these patterns on the
structure of forest, grassland, and wetland landcover.

It would be useful to develop a landcover dataset for 1978 using the same
methods used to develop the 2006 landcover dataset. This would allow for a more
reliable change comparison. The author notes that the change comparison is not
essential to the planning process, as currently implemented, however, it is the author’s
opinion that using landscape metrics to evaluate the alternative scenarios is a good way

to incorporate biological information into the planning process.

67

LITERATURE CITED

Alberti, M., & Marzluff, J. M. (2004). Ecological resilience in urban ecosystems;
Linking urban patterns to human and ecological functions. Urban Ecosystems , 241 -
265.

Benedict, M. A., & McMhon, E. T. (2001). Green Infrasturcture: Smart Conservation
for the 21st Century. Washington DC: Sprawl Watch Clearinghouse.

Berke, Phillip R., (2007) Ecology and New Directions for Land Use Planning: Barriers
and Opportunities to Change, In E. L. Institute, Lasting Landscapes: Reflections on the
Role of Conservation Science in Land Use Planning

Bierwagen, B. B. (2007). Connectivity in urbanizing landscapes: The importance of
habitat configuration, urban area size, and dispersal. Urban Ecosystems , 29-42.

Botequilha, A., & Ahem, J. (2002). Applying landscape ecological concepts and
meterics in sustainable landscape planning. Landscape and Urban Planning, 59, 65-93.

Cohen, J. (2004). Natural comminity abstracthr oak openings. Lansing: Natural
Features Inventory.

Cohen, J. (2001). Natural community abstractfor oak barrens. Lansing : Natural
Feature Inventory.

Community Viz. (2001). Water Resource Protection Black Earth Creek Wisconsin.
Bolder: Community Viz.

Environmental Law Institute. (2003). Conservation T hresholdsfor Land Use Planners.
Washington DC: Environmental Law Institute.

Folke, C., Carpenter, S., Elmqvist, T., Gunderson, L., Holling, C. S., Walker, 3., et al.
(2002). Resilience and Sustainable Development. The Environmental Edvisory Council
to the Swedish Government.

Forman, R. T. (1995a). Land Masaics. New York: Cambridge University Press.

Fulton, W., Pendall, R., Nguyen, M., & Harrison, A. (2001). Mo Sprawls Most? How
Growth Patterns Differ Across the US. Washington DC: Brookings Institute.

Hyde, D., Paskus, J ., & Enander, H. (2006). Genesee Lapeer Shiawassee Counties
Potential Conservation Areas Assesment. Flint: GLS GreenLinks.

Jenerette, G. D., & Wu, J. (2001). Analysis and simulation of land-use change in the
central Arizona — Phoenix region, USA. Landscape Ecology , 611-626.

68

Kaufman, M. M., & Marsh, W. M. (1997). Hydro-ecological implications of edge cities.
Landscape and Urban Planning , 277-290.

Leonardi, J. (2001). The Flint River Assesment. Lansing: Michigan Department of
Natural Resources.

McDonald, L., Allen, W., Benedict, M., & O'Connor, K. (2005). Green Infrastructure
Plan Evaluation Frameworks. Journal of Conservation Planning , 12-43.

McGari gal, K. (2007). What is a Landscape? Lecture notes from Landscape Ecology,
NRC 621, University of Massachusetts .

Merenlender, A. (2007). Protecting Wildland Beyond the Urban Fringe. Washington
DC: Environmental Law Institute.

Michigan Department of Natural Resources. (2003). Resource Management: Wildlife
Division 2003 Program Descriptions. Michigan Department of Natural Resouces.

Michigan Department of Natural Resources (1998), Changes to MIRIS Land Cover/Use
Classifacation System.

Nelson, C. (2001). Economic Implications of Land Use Patterns for Natural Resource
Recreation and Tourism. Public Sector Consultants.

Nelson, C., Steffey, E., Clark, E., Steger, K., & Danforth, K. (2007). Michigan State
Game Area Use and User Assesment.: March I5-December I5, 2006. East Lansing:
Department of Community Agriculture Recreation and Resource Studies.

Perlman, D.L., (2007) Views of a Conservation Biologist, In E. L. Institute, Lasting
Landscapes: Reflections on the Role of Conservation Science in Land Use Planning

Pickett, S., Cadenasso, M., Grove, J ., Nilon, C., Pouyat, R., Zipperer, W., et a1. (2001).
Urban Ecological Systems: Linking Terrestrial Ecological, Physical, and
Socioeconomic components of Metropolitan Areas. Annual Review of Ecological
Systems , 127-157.

Stern, M. A., & Marsh, W. M. (1997). Editor Introduction, the decentered city: edge
cities and the expanding metropolis. Landscape and Urban Planning , 243-246.

The American Heritage. (2005). Dictionary of Cultural Literacy. Houghton Mifflin.

Theobald, D. M. (2007a). Challenges in bridging conservation science and land use
planning. In E. L. Institute, Lasting Landscapes: Reflections on the Role of
Conservation Science in Land Use Planning (pp. 13-24).

Theobald, D. M. (2007b). Evaluating the cumulative effects of development by analzing
alternative build-out scenarios: A case study of Ouray County, Colorado. Journal of
Conservation Planning , 57-78.

69

Theobald, D. M., Hobbs, N., Bearly, T., Zack, J. A., Shenk, T., & Riebsame, W. E.
(2000). Incorperating biological information in local land-use decision making:
desinging a system for conservaiton planning. Landscape Ecology, 15 (1), 35-42.

Verburga, P. H., Koninga, G. H., Koka, K., Veldkampb, A., & Bouma, J. (1999). A
spatial explicit allocation procedure for modelling the pattern of land use change based
upon actual land use. Ecological Modeling , 45-61.

Vogt, C. A., Egeler, C., Jordan, E., & Clark, E. (2006). Genesee, Lapeer, Shiawassee
Counties Conservation Needs Assesment. Flint: GLS GreenLinks.

White, R., Engelen, G., & Uljee, I. (1997). The use of constrained cellular automata for
hi gh-resolution modelling of urban land-use dynamics. Environment and Planning B:
Planning and Design , 323 — 343

70

   

3|33313333331333333133333l