applying principles of landscape design and...

Applying Principles of Landscape Design and Management to Integrate Old-Growth Forest Enhancement and Commodity Use DAVID J. MLADENOFF MARK A. WHITE

Natural Resources Research Institute University of Minnesota, Duluth Duluth, MN 55812, U.S.A.

THOMAS R. CROW

USDA Forest Service North Central Forest Experiment Station Forestry Sciences Laboratory Rhinelander, WI 54501, U.S-~.

JOHN PASTOR Natural Resources Research Institute University of Minnesota, Duluth Duluth, MN 55812, U.SMt.

Abstract: Using geographic information systems (GIS) and spatial analysis technique& we developed a landscape design to maintain old.growth forest remnants and integrate commodity production in the surrounding second-growth ma-

The 4500-ha forest landscape in northern Wisconsin contains scattered patches o f old-growth eastern hemlock (Tsuga canadensis) and northern hardwood~ predominately sugar maple (Acer saccharum). The design incorporates an o ld .growth res tora t ion z o n e s u r r o u n d i n g o ld-growth patches to buffer and enhance forest. interior habitat and l ink nearby old-growth remnant& This addition restores as- pecis o f landscape patch size and structure and ecosystem juxtaposi t ion that characterize a nearby, large and contig- uous natural old-growth landscape A larger secondary zone is delineated f o r uneven-aged forest managemen£ This zone provides a matr ix structurally s imilar to the old-growth patches but also accommodates harvesting. A larger outer zone is retained primari ly in even-aged forest o f aspen (Pop- ulns tremuloides) and paper birch (Betula papyrifera), but traditional clearcutting practices are modified to part ial cutting and mixed-species rotation~ This design meets limited goals o f biodiversity enhancement and integrated commodity production in a landscape that wil l remain largely harvested The landscape design is therefore improved not only by buffers and corridors provided to old-growth ecosys-

Paper submitted March 24, 1993; revised manuscript accepted Sep- teraber 30, 199~

752

Conservation Biology, Pages 752--762 Volume 8, No. 3, September 1994

Aplicando pincipios de disefio y manejo del paisaje para integrar el mejoramiento de bosques maduros y su utilizaci6n para consumo

Resumen: Utilizando sistemas de informaci6n geogrdflcos (SIG) y t$cnicas de andlisis espacla~ nosotros desarrollamos un diseho de paisaje para mantener los remanentes de bosques maduros e integrar la producci6n en la matr lz cir- cundante de cmctmiento secundario. Las 4500 ha de patsaje boscoso en el notre de Wisconsin, cont iene parches de crecimiento antiguo dispersos de abetos del este (Tsuga canadensis) y bosques de madem dura del nort¢ predominan- temente arce azucarero (Acer saccharum). El diseho incor- pora una Zona de Restauraci6n del crecimiento anttguo, rodeada por parches de remanentes maduros para amor- tiguar y mejorar el hdbitat del interior del bosque y unir remanentes cercanos de antiguo crectmiento. Esta adici6n restaura aspectos del tamaho de los parches del patsaj¢ estructura y yuxtaposici6n que caracteriza a un paisaje cer- cano, maduro grande y conttgua Una Zona Secundaria rods grandees delineada para el manejo de bosques de edad dis. par. Esta zona proveerd una matr iz estructuralmente similar a la de parches de antiguo crecimiento, pero tambidm incluye explotaci6n maderera Una Zona Exterior nuis am- p l la es retenida pr imariamente en bosques de igual edad de dlamo (Populus tremuloides) y de abedul (Betula papyrifera), pero los prdcticas de tala tradicionales son modifi- cadas hacia cortes parciales y rotaci6n de especies mixta~ Este diseho cumple con objetivos limitados de mejorar la

M/adeno~ eta/. /nteSrated Landsca/~ Des/gn 753

tem~ but by modifying the management o f the majority commodity lands matr ix as wel t

biodtversidad e integrar la producci6n para consumo en un paisaje que permanecerd ampl iamente explotado. De esta f o r w ~ el diset~o del paisaje es mejorado, no s61o p o t medio de zonas buffer y corredores que proveen protecci6n a los ecosistemas de crecimiento antiguo, sin6 tambidm por medio de la modificaci6n del manejo de la mayoria de la matr iz de las tierras explotadas para consumo.

Introduction

Background

Demands for land managers to produce commodities and conserve biological diversity concurrently are increasing (Probst & Crow 1991). As a result, greater at- tention is being given to the role that the vast areas under some form of active management for food and fiber can play in maintaining biological diversity (Frank- lin 1993). Clearly, there is a need for greater integration of biodiversity enhancement and commodity produc- t ion on managed landscapes (Wilcove 1989; Crow 1991; Hansen et al. 1991; Probst & Crow 1991).

Biological diversity maintenance must be accomplished at a multitude of spatial scales in order to meet all needs (Noss & Harris 1986; Noss 1987; Crow 1991 ). Reserve areas, both large and small, remain essential for many ecological processes, and to provide protect ion and isolation from human influences for many species. For some processes and for wide-ranging species, effective reserves must generally be larger and more numer- ous than has typically been the case (Diamond & May 1976; Pickett & Thompson 1978; Baker 1989; Noss 1991). At the same time improved management and planning are needed on commodity lands to ensure future product ive potential as well as to contribute to diversity maintenance through more integrated and ecologically based land-use practices (Probst & Crow 1991; Mladenoff & Pastor 1993). Throughout eastern North America--and indeed much of the worM's temperate climatic zone---past and current land uses have created an extensively and intensively altered landscape with few remaining natural areas. Biodiversity maintenance in these landscapes poses significant challenges for enhancing the value and functioning of remaining natural fragments.

Objectives

We approached this situation on a forested Wisconsin landscape in the northern Great Lakes region (Fig. 1, inset ). Because of the complex mixed ownership of this landscape, simply delineating a large reserve area surrounding the old-growth patches is not a management option. This is increasingly the case elsewhere as com- petit ion be tween land uses increases even under public

ownership forests. The object ive of our study, old- growth enhancement in a commodity landscape, is limited to addressing a specific por t ion of biodiversity maintenance needs based on the opportunities provided by a real world example. Yet maintaining old growth as functional ecosystems and for the habitat values they provide is important in this region (Tyrrell & Crow 1989). Forests once characteristic of the region, largely old growth (Frelich & Lorimer 1991 ), are now reduced to less than 0.5% of their former ex ten t (Eckstein 1980). Late-successional forests, particularly with large conifers, are important as forest bird habitat because of their more diverse structure and composit ion (Jaako P6yry Consulting, Inc. 1992; Howe e ta l . 1993). In the Border Lakes landscape, much of the remaining old growth contains eastern hemlock (Tsuga canadens i s ) , which has been nearly eliminated regionally by past log- ging and land-use practices (Mladenoff & Stearns 1993).

Given this importance, our goal is to develop a landscape design that enhances the value of remaining natural old-growth forest by restoring characteristic landscape-scale features. To do so we used geographic information systems (GIS) and a hierarchical approach to develop our design. Our task here is to develop a design that protects and enhances the cur rent old- growth ecosystems both judiciously and effectively. A requirement of the design is to integrate harvesting with the protect ion and enhancement of old-growth forest. The landscape contains scattered, small remnants of old- growth forest wi thin a matr ix of managed second growth in mixed ownership (mostly private with some public land).

Integration of Previous Work

Results of previous work are key to our design and include a comparison of landscape pattern in the Border Lakes, characteristic of managed second-growth landscapes, with that of the Sylvania Wilderness, a nearby, reference old-growth landscape (Mladenoff et al. 1993). We also verified that the Sylvania landscape was com- parable to Border Lakes through a reconstruct ion of the pre-European landscape and historical vegetation. These data were used to analyze temporal landscape transi- tions to the present and to project future conditions (White & Mladenoff 1994). We previously found that

Conservation Biology Volume 8, No. 3, September .1994

754 Integrated Landscape Design Mladenoff et al.

the pre-European landscape, as reflected by Sylvania, had important characteristics that are now lacking on Border Lakes and in much of the region (Mladenoff & Pastor 1993). (1 ) The natural old-growth landscape has a greater range of forest patch sizes (ranging from less than 1 to more than 1000 ha) compared to Border Lakes (ranging from less than 1 to 200 ha), resulting in greater edge-inteilor ratios on Border Lakes. (2 ) Patch com- plexity, as measured by fractal dimension (Mandelbrodt 1977) is significantly greater on the natural landscape, providing for greater interspersion of ecosystems and habitats. ( 3 ) Characterist ic adjacency relationships, where particular ecosystems typically occur next to certain other ecosystems, no longer occur on the managed landscape. This adjacencY relationship may be important for complex, multiple habitat requirements, or for seed dispersal and stand regeneration. The strongest adjacency relationship found on the Sylvania landscape occurred between old-growth hemlock and lowland conifer ecosystems.

We examined the scales at which important landscape features are expressed in the natural landscape (Mladenoff et al. 1993) to decide what might realisti- cally be restored or managed at the scale of the Border Lakes landscape. We determined that several key objectives might be met, based on our previous analysis (Mladenoff et al. 1993). These objectives were (1) to restore significant ecosystem spatial adjacencY relationships that exist in the natural landscape, and (2 ) to enhance the integrity of the interior of remaining old- g rowth forest . We then d e v e l o p e d t echn iques to incorporate those features in a landscape design using modifications of landscape patch structure, and by pro- posing modified and spatially segregated forest harvesting. These techniques include adding and--potentially, over t imemrestor ing some old-growth forest. At the same time, we designed a plan where forest harvesting decreases in intensity from outer to core old-growth areas. Continued forest cutting is a given condition in this landscape.

Our purpose in this study has a limited scope. Using a real landscape, we seek to present an example of a con- ceptual approach and techniques for designing integrated forest landscapes using GIS. To illustrate, we compare our sample design with prior conditions and features found in a natural o ld -growth landscape through our previous work.

Study Landscape

Our study landscape, known as the Border Lakes area, is located in north-central Wisconsin, U.S.A. (46°12'N, 89°34'W), with Michigan's upper peninsula adjacent to the north (Fig. 1). Climate is continental, with mild sum- mers and long, cold winters with snow cover from No-

vember to April. Mean temperatures are - 1 0 ° C for January and 18 ° C for July, with approximate annual precipitation of 85 cm.

Our design encompasses the eastern two-thirds of the Border Lakes landscape (4415 ha), which contains all current old-growth forest remnants. Like much of the region, this morainal landscape contains many lakes and wetlands within an upland forest matrix that was formerly dominated by conifers such as eastern hemlock and northern hardwood species, predominantly sugar maple (Acer saccharum) and yellow birch (Betula al- leghaniensis). White pine (Pinus strobus) was present as an admixture and locally abundant on sandier softs (Curtis 1959; White & Mladenoff 1994).

Today, only scattered remnants of the original old- growth hemlock and northern hardwood forest remain. Second-growth forests of successional hardwoods such as aspen-birch (Populus tremulotdes-Betula papyrifera) and conifers such as spruce-fir (Picea glauca- Abies balsamea) predominate in pure stands and in a variety of mixtures, along with second-growth northern hardwoods dominated by sugar maple (Mladenoff et al. 1993). These second-growth forests are common to this boreal -nor thern hardwood transition region (Wolter et al. 1995). Past human disturbance and the vagaries of site, climate, and stochastic events have produced a more complex and varied assemblage of types on the landscape than originally occurred (Pastor & Mladenoff 1992).

Methods

The original landscape map was derived from 1:24000 color infrared stereo photography. The photos were in- terpreted and forest and land cover mapped onto a rec- tiffed base map. The map was digitized into the ARC/ INFO geographic information system software (ESRI, Inc. 1987). Data acquisition and GIS database creation methods are detailed in Mladenoff et al. (1993). All second-growth types delineated in Mladenoff et al. (1993) are mapped here as general upland and lowland catego- ries. Old-growth forest contrasts markedly from the dominant younger forests on the site. Old growth was identified from air photos and historical information. In our discussion we refer to these mapped patches as vegetation types or ecosystems, as seems most appro- pilate in a given context. Ecosystems may be described at any scale, and different forest types in this region have different characteristic ecosystem processes (Pastor & Mladenoff 1992). Where we are discussing potential habitat value of these different patch types or ecosystems, we refer to them as habitats.

Our analysis methods here are techniques used to meet design objectives. We used GIS and several spatial analysis methods to develop and assess the landscape

Conservation Biology Volume 8, No. 3, September 1994

M/~enoff eta/. / n ~ La~dsca~ Des/~ 755

Figure 1. Study landscape with edge~interior analysis of old-growth patche~ The shading delineates the penetra- tion of a l O0-meter edge effect into the old-growth polygons, and remaining old-growth interior.

structural changes in our proposed design. These structural changes were created by designing an old-growth r e s to r a t i on zone a round the exis t ing o ld -g rowth patches. Complementary management changes are in- corporated into the design by aggregating harvesting methods into zones and assuming qualitative changes in their implementation. Thus a moderately harvested secondary zone was designated beyond the restoration zone, as well as a more intensively managed outer zone (Fig. 2).

Edge/Interior Analysis

We used the BUFFER procedure in ARC/INFO to evaluate edge/interior conditions of a habitat patch by gen- erating a band of specified width around the interior of the polygon boundaries, to yield a map with habitat edge and interior differentiated (Fig. 1). We then calcu- lated amounts and proport ions of old-growth edge and interior, and edge/interior ratios for the original landscape and the designed landscape. In our example we used a 100-meter edge effect extending into old-growth patches adjacent to immature forest or wetlands (Frank- lin & Forman 1987). We also chose the 100-meter edge value as a guide in restoring old-growth buffers, assuming that many adjacent areas wili eventually be cut, cans- ing greater edge effects than occur now. Other specific

edge-effect distances can be justified for specific objectives. For example, Brittingham and Temple (1983 ) found higher nest parasitism by cowbirds up to 300 meters into forest patches from agricultural edges. In our landscape, we are assuming microclimate effects that may influence tree reproduct ion or ecosystem processes as well as effects on habitat value (Franklin & Forman 1987; Chen et al. 1992).

Our main design objective is to insure the integrity of all existing old-growth forest. Based on our 100-meter edge effect assumption, we delineated a 100-meter-wide buffer around all existing old-growth polygons within upland areas, designating these areas as the old-growth restoration zone. We then compared the spatial effects of this change on the landscape with the prior condition. The width of the secondary zone (300 meters )was also chosen based on our 100-meter edge assumption. Any areas of the secondary zone will have some mini- mum habitat interior even if isolated by disturbance or serving as a corridor be tween larger patches.

Adjacency Analysis

We conducted an adjacency analysis using the electivity index of Jacobs (1974) and Jenkins (1979), and as applied previously (Pastor & Broschart 1990; Mladenoff et al. 1993) to evaluate how the juxtaposition of different


756 /ntegrated Landscape Des~ Mladenoff et M.

Figure. 2. Border Lakes study landscape (4415 ha) with remnant old-growth hemlock and hardwood forest type~ The ditmme matrix o f surrounding second-growth forest types (Mladenoff et aL 1993) are mapped as gen. eral upland or lowland forest type~ Map includes landscape design with all zone~. (1) current old-growth, (2) lO0-meter restoration zon~ (3) 300-meter secondary zone, (4) general commodity management outer zone

ecosystem types change due to the addition of the restoration zone to the landscape:

Eij = ln[(r¢X 1 - P¢)/(P#X 1 - r¢)],

where Etj is the electivity index for the nearest neighbor probabilities of patch types i and j, and re is the proportion of shared boundary between type i and type~ pc is the proportion of shared boundary of type i with all other patch types except typej. These electivity index values are compared with the chi-square distribution for significance of positive or negative association between ecosystem types.

We mapped a band of specified width (here 20 m) around the outside of all polygons of a type to deter- mine the type and relative perimeter proportion of the adjacent ecosystem patch types using the BUFFER procedure in ARC/INFO (Fig. 3). We applied this analysis to the existing landscape and compared it to an identical analysis of the proposed landscape design with hypo- thetically induced changes due to management.

We also used ARC/INFO to summarize patch types by number and area, to generate management-zone boundaries of specified width around the old-growth forest types, and to assess future changes in old-growth patch sizes and forest interior based on our design.

Results and Discussion Edge/Interior Analysis

The total edge area in the current landscape, based on the lO0-meter buffer, is 282.6 ha, or nearly 84% of the total old-growth area (Fig. 1, Table 1, upper values). The old-growth hemlock type is most affected by edge, because the complex, linear shapes leave little forest interior (Table 1, upper values, Fig. 1). Hemlock forests create a highly modified microclimate within stands compared to hardwoods (Godman & Iancaster 1990). They also differ significantly in ecosystem processes such as nitrogen cycling (Mladenoff 1987) and structural features such as coarse woody debris (Tyrreli & Crow 1989; Tyrrell 1991 ). Logically, edge contrast may affect forest processes most in these stands.

The remaining old-growth hardwoods are less affected by edge than hemlock, though still significantly reduced in interior (Fig. 1, Table 1, upper values). Hard- wood patches retain proportionately nearly three times as much area in forest interior as hemlock (20.9% versus 7.1% ), due both to larger patch size and simpler shape. Similarly, edge habitat surrounding interior hardwoods is proportionately less than that surrounding hemlock (7.0% versus 13.8% ) (Fig. 1, Table 1, upper values).

Conaervation Biology Volume 8, No. 3, September 1994

M/adenoff~a/. l n ~ / a n d s c a p e D e s ~ 757

Figure 3. Example from adjacency analysis showing buffering of old-growth hemlock patches following addition of lO0-meter restoration zone Buffer indicates type, extent and location of adjacent ecosystem type~ Map illustrates the upper left sub-area of the study landscape

Adding the lO0-meter buffer as an old-growth restoration zone means that all existing old growth will be- come forest interior, thereby maximizing its integrity. By adding this relatively small but key area to the old- growth forest, reserved land area is increased by only 174.8 ha, but wi th substantial effective changes in the

landscape. Besides maximizing o ld-growth interior , many formerly isolated old-growth patches now co- alesce, thus increasing their potential habitat value for these ecosystems and bet ter assuring their functional integrity (Fig. 2). The number of separate old-growth patches has been reduced f rom 19 to eight, and the

Table 1. values).*

Edge/interior analysis of current old growth (upper values) and old growth plus lO0-meter exterior restoration zone (lower

Overall Interior Edge

Area % n ~ Area % n ~ Area % n

HEMLOCK (original) 222.7 66.0 15 14.8 15.9 7.1 6 2.7 206.8 92.9 15 13.8 Restored 328.2 64.1 14 23.5 44.8 13.7 12 3.7 283.4 86.3 14 20.2

HARDWOOD (original) 114.6 34.0 13 8.8 24.0 20.9 11 2.2 90.6 79.1 13 7.0 Restored 183.9 35.9 10 18.4 43.5 23.6 13 3.3 140.4 76.4 10 14.0

COMBINED (original) 337.3 100.0 19 17.8 54.8 16.2 10 5.5 282.6 83.8 19 14.9 Restored 512.1 100.0 8 64.0 105.9 20.7 15 7.1 406.3 79.3 8 50.8

* Edge and interior classes are based on an assumption o f 100 meters edge effect extending into old-growth adjacent to second.growth forest (see Methods). Data are for total area (ha), percentage o f tota~ number of patches in each clas~ and mean patch area (ha).


758 lntegra/ed Landscape Des/go M/adenoff et al.

mean patch size, possibly the most important factor, increased more than three-fold from 17.8 to 64.0 ha (Table 1).

Most of the added 100-meter restoration zone ends up under the edge category, not forest interior, and much of it serves as relatively narrow links between formerly isolated patches when a lO0-meter edge is as- sumed (Table 1). Currently, less than 3.0% of the study landscape is in clearcuts and other openings (Table 2). However, nearly all of the second growth is at or ap- proaching harvest age. Therefore, under conventional management, additional harvest patches would be ex- pected soon to increase significantly. We therefore feel that our conservative approach to this design criterion is justified when planning for future management. Adja- cent to new clearcuts, edge effects much greater than 100 meters may be created(Frankl in & Forman 1987; Chen et al. 1992).

Adjacency Analysis

The adjacency analysis map with the "restored" old- growth hemlock polygons (original old-growth hemlock plus lO0-meter res tora t ion zone ) shows old- growth neighbor associations (Fig. 3). In the current landscape, hemlock is not spatially associated with lowland conifers. In the restored landscape, old-growth hemlock is again positively associated with lowland conifers (Table 3). In addition, the negative relationship between old-growth hardwoods and lowland conifers, and the positive associations between old-growth hardwoods and hemlock, still remain (Fig. 2, Table 3). These associations were all important in the natural Sylvania landscape and the pre-European Border Lakes (Mlade- noff et al. 1993; White & Mladenoff 1994).

Discussion Landscape Design

Much is unknown about the relation between ecological processes and landscape patterns. Yet we have found

that there are characteristic patterns that define both natural old growth and disturbed, managed landscapes (Mladenoff et al. 1993). In this context it is prudent to attempt to mimic natural landscape structure, restoring aspects that can be managed at different scales (Crow 1991 ). Habitat interior, old-growth patch size, and landscape linkage can be significantly increased at the scale of our landscape design ( lO00s of hectares) , even where constrained by commodity management. A design using the fine-grained manipulations possible with GIS allows important ecosystem juxtapositions to be restored on the landscape while maintaining efficient land use. These improvements may benefit area-sensitive species requiring mature forest ecosystems, as well as restore a degree of connectivi ty for ecosystem processes and species dependent on greater landscape linkage.

The natural old-growth landscape had attributes of landscape structure that we could not restore at this scale (Mladenoff et al. 1993). The full range and distribution of old-growth patch sizes on Sylvania (less than 1 to more than 1000 ha) could not be replicated on the Border Lake landscape. The natural old-growth landscape also contained forest patches of more complex shape than those of the disturbed Border Lakes. More- complex patch shapes increase the relative amount of forest edge. In the natural landscape, however , the greater number and size of old-growth patches allows both forest interior (habitat) and patch interspersion (connectivi ty) to be maximized (Mladenoff et al. 1993). Also, in the natural landscape fewer patch edges are high-contrast, induced edges characteristic of heavily disturbed forest landscapes. These features cannot generally be restored at the limited spatial extent of our study landscape.

Complementary Management Oranges

Our design to maintain the integrity of the old-growth forest remaining on the Border Lakes landscape reduces

Table 2. C o ~ of current and restored isadmq~, baaed oa the ~ design.*

Area (ha) % Area Number o f Patches

Restored Original Restored Original Restored Original

Mean Patch (ha)

Restored Original

Old Growth Hemlock 329.9 223.0 7.5 5.1 14 15 23.6 14.9 Hardwood 183.9 114.6 4.2 2.6 10 13 18.4 8.8 Second Growth N. Hardwood 1301.1 1388.3 29.5 31.5 95 74 13.7 18.8 Mixed Hardwood 540.6 549.5 12.2 12.5 42 39 12.9 14.1 Hardwood and Conifer 375.9 387.0 8.5 8.8 33 32 11.4 12.1 Hardwood over Conifer 1022.7 1052.0 23.2 23.8 131 107 7.8 9.8 Mixed Conifer 549.3 585.4 12.4 13.3 155 126 3.5 4.7 Upland Open 109.4 115.0 2.5 2.6 22 22 5.0 5.2 Total 4414.7 4414.7 100.0 100.0 502 428 12.0 11.0

* Restored version has 100 meters added to perimeter o f all old.growth patche~


Mladello~ et xl. l n ~ t e d landscape Design 759

Table 3. Adjm:ency analysis o f patch types in current h m d s m p e (upper symbols) and r e s t o r ~ I m d s m q ~ ( lower symbols) ."

Old Growth Second Growth

LC HE HI) NHD MHD H/C HOC MC

Lowland Conifer 0 0 0 0

Old-Growth Hemlock

Old-Growth Hardwood

0 + + +

0 + 0 +

- 0 0 + - 0 0 +

0 0 0 - - 0 0 -

0 0 0 0 - 0 0 -

0 +

m

Second-Growth Northern Hardwood - 0 0 0 0 0 +

- - 0 0 + 0 +

Mixed Hardwood 0 0 0 0 0 + 0 - 0 0 0 +

Hardwood/Conifer 0 0 0 + 0 0 0 0 0 0 0 0

Hardwood over Conifer + - 0 0 0 0 + - 0 0 + 0

Mixed Conifer + - 0 + 0 0 - 0 - 0 + 0 0 -

m

* Restored landscape has 100 meters buffer addition to old-growth patcbez Associations shown for X 2 > 7. 78, 1 dr, p < 0.005.

the isolation and increases the size of the original old- growth patches, but the distribution of these ecosystems leaves a complex map with a large propor t ion of existing and potential old growth (restorat ion zone) in the edge category ( Fig. 2). The secondary zone buffer of 300 meters surrounding the restoration zone and existing old-growth partly ameliorate this condition (Fig, 2). The p roposed forest management within this zone is uneven-aged, with eventual creation of a mature forest that is structurally and compositionally similar to the adjacent old-growth. This will p rov ide a secondary buffer to the entire design and particularly to the narrow old-growth linkages be t ween larger old-growth patches (Fig. 2).

A combinat ion of secondary buffering with a narrow restoration zone maximizes old-growth enhancement while still maintaining commodi ty production. This integration of harvesting and biodiversity protect ion can be further enhanced by matching spatial buffering with what is, in effect, temporal buffering through selection cutting. In selection harvesting, cutting of trees in stand is spread over many years. Selection cutting includes a variety of harvesting intensities, and thus it is not ttecessarily a conservat ion panacea. Removing a large propor t ion of basal area ( > 3 0 % ) can result in i nadequa te in tac t c a n o p y for s o m e habi ta t needs (Lorimer 1989), and frequent entries can also require maintenance of large road networks. Existing guides provide the management flexibility to grow more trees into larger d iameter classes whi le maintaining stand

structure (Arbogast 1957; Crow et al. 1981). Retaining a greater number of trees in larger size classes, maintaining higher stand basal area, increasing the t ime between cuttings, and creating canopy gaps for regeneration all increase stand structural diversity (Crow et al. 1981; Mladenoff & Pastor 1993). Such an approach will provide higher-value products, thus providing some incentive for these changes (Tubbs 1977; Marquis 1978). The existing forest types within the secondary zone are largely suitable to this modified selection management . Over 75% of the uplands are dominated by nor thern hardwoods (largely sugar maple) or are succeeding to northern hardwoods and conifers (Table 4).

The outer zone of the landscape design, beyond the secondary buffer, can accommodate more in tens ive- - but modif ied-- -even-aged m a n a g e m e n t des igned to maintain successional species such as aspen and birch, as well as uneven-aged management of nor thern hardwoods (Pastor & Mladenoff 1993). Successional types can be managed by partial cutting in mixed stands with conifers such as spruce/f i r - - for example, alternating harvests of aspen with rotations of fir (Graham et al. 1963). Selected species can be retained in clumps and large individuals for structural diversity that more re- sembles naturally regenerat ing stands (Hansen et al. 1991; Franklin 1993; Mladenoff & Pastor 1993). Over 50% of the uplands in this zone are mixed stands of aspetL birch, spruce, and fir that can be maintained using these modified silvicultural systems.

This t iered landscape design is in some respects a


760 Intuited ~ Design Mladeno~ et al.

Table 4. Smmary of p m ~ types withlu otd-Smr.h ~ zoue ( lOO meu~), ~O-mmer ~ m t r zme, - , a omcr zone.

Area (ha) % Area

Old Growth Old Growth and (100 m) 300-m and (100 m) 300-m Restoration Buffer Outer Restoration Buffer

Zone Zone Zone Zone Zone Outer Zone

Lowland Conifer 136.2 240.2 711.6 20.9 34.8 17.1 Old Growth Hemlock 329.9 - - - - 50.8 - - m Hardwood 183.9 - - - - 28.3 - - Second Growth N. Hardwood - - 172.4 1130.7 - - 25.0 27.2 Mixed Hardwood - - 39.2 501.4 - - 5.7 12.1 Hardwood and Conifer - - 51.9 324.0 - - 7.5 7.8 Hardwood over Conifer - - 95.5 926.0 - - 13.8 22.3 Mixed Conifer - - 86.8 460.0 - - 12.6 11.1 Upland Open m 5.2 104.2 ~ 0.8 2.5

Total 650.0 691.2 4157.8 100.0 100.0 100.0

modification of the multiple-use module (MUM) pre- sented by Harris (1984) and Noss and Harris (1986). Their approach recognizes that reserve cores and corridors do not exist in isolation f rom the larger managed landscape context (Noss 1987; Frartklirl 1993). In our design, however, we place equal emphasis on modifying forest management pract ices within the zones beyond the old-growth core as well as on the spatial gradient of management intensity of the different zones per se. Har- vests can be spread temporally through the use of modified selection harvesting, which is well suited to many forest types in this region. This provides a more fine- grained and lower contrast integration of harvesting and biodiversity protec t ion than in the simplest application of the MUM concept. By necessity of ownership, the harvested matr ix zones ( secondary zone and outer zone) will continue to consti tute 11.9% and 75.9% of the landscape design, making their dual purpose very important and providing incentive for finer integration of the dual management objectives (Table 4). In an integrated design, however , the landscape plan must be evaluated on the basis of changes in the management of the commodi ty forest matrix as well as the amount of buffers and corridors provided.

Summary

As appfied here, integrated forest management landscape designs can increase the value of small, isolated ecosystem remnants for protect ion of certain biodiversity values, even in situations where forest harvesting will continue. Selected landscape structural attributes can be restored at characteristic scales while maintaining greater within-stand structure by combining spatial changes wi th qualitative management modifications.

The hierarchical management zones efficiently buffer the core area while allowing harvesting intensity to in-

crease with distance from the core. Buffer zones are managed temporal ly as wel l as spatially. Inne rmos t zones are managed as uneven-aged forest favoring re- tention into larger diameter sizes. Management of the outer zone includes even-age harvesting accomplished with partial cutt ing techniques (Mladenoff & Pastor 1993).

The constraints of our dual object ives mean t that some characteristics of the old-growth landscape, such as the range of patch sizes, the shape complexi ty of large patches, and dominance by an old-growth matrix, could not be restored on the Border Lakes. Clearly, not all conservation objectives can be met at only one or several scales; context as well as content of reserves must be considered in their design, and a need remains for very large reserves if the full array of natural processes are to be maintained on the landscape. Our landscape design is an example of a concept and technique. Other landscapes and ownerships may provide opportunit ies for modifying the balance of commodi ty product ion and biodiversity protect ion in different ways.

Such designs, coupled wi th modif ied silvicultural techniques, can contr ibute to more ecologically sustainable forest management. Not only do such techniques directly benefit the ecological health of the system, but they increase biological diversity, provide a further ecological buffer and sources of colonization, and hedge against future change. In this way such landscape designs both extend biodiversity protec t ion into the managed forest matrix and help to insure the future produc- tive potential of managed forests.

Acknowledgments

This study was funded jointly through a grant f rom the U.S. Forest Service, North Central Forest Experiment Station, Landscape Ecology Program, and The Nature


M/adenoff et 92. laa~zag-d Laadsc~ Oea~ 761

Conservancy , W i s c o n s i n Chapter , to D.J. Mladenof f and J. Pastor. Bren t Hag lund p l a y e d a key ro l e in in i t ia t ing this p ro j ec t , and the W i s c o n s i n DNR p r o v i d e d logis t ical s u p p o r t . C o m m e n t s f r o m A n d r e w Hansen , W i l l i a m Baker, and an a n o n y m o u s r e v i e w e r subs tan t ia l ly i m - p r o v e d the manusc r ip t . This is c o n t r i b u t i o n n u m b e r 135 o f t he C e n t e r for W a t e r and the Env i ronment , Nat- ural Resources Resea rch Ins t i tu te , and n u m b e r 31 of t he

Natura l Resou rce s GIS Laboratory .

Literature Cited

Arborgast, C., Jr. 1957. Marking guides for northern hardwoods under the selection system. Paper LS-56. Lake States Forest Experiment Station, U.S. Forest Service, St. Paul, MN.

Baker. W.L. 1989. landscape ecology and nature reserve design in the Boundary Waters Canoe Area, Minnesota. Ecology 70:23-35.

Brittingham, M.C., and S.A. Temple. 1983. Have cowbirds caused forest songbirds to decline? Bioscience 33:31-35.

Chen, J., J.F. Franklin, and T.A. Spies. 1992. Vegetation re- sponse to edge environments in old-growth Douglas-fir forests. Ecological Applications 2:387-396.

Crow, T. IL 1991. Landscape ecology: The big picture approach to resource management. Pages 55-65 in D.J. Decker, M. E. Krasny, G. IL Goff, and others, editors. Challenges in the conservation of biological resources: A practitioner's guide. Westview Press, Boulder, Colorado.

Crow, T. IL, IL D. Jacobs, IL 1L Oberg, and C. H. Tubbs. 1981. Stocking and structure for maximum growth in sugar maple selection stands. Research Paper NC-199. North Central Forest Experiment Station, U.S. Forest Service, St. Paul, MN.

Eckstein, R. G. 1980. Eastern hemlock (Tsuga canadensis) in north central Wisconsin. Research Report 104. Wisconsin De- partment of Natural Resources, Madison Wisconsin.

Curtis, J.T. 1959. The vegetation of Wisconsin. University of Wisconsin Press, Madison, Wisconsin.

Diamond, J. M., and R. M. May. 1976. Island biogeography and the design of natural reserves. Pages 163-186 in IL M. May, editor. Theoretical ecology. Saunders, Philadelphia, Pennsyl- vania.

ESRI, Inc. 1987. ARC/INFO user's manual. Environmental Sys- tems Research Institute, Inc., Redlands, California.

Franklin, J. F. 1993. Pages 127-144 in G. H. Aplet, J. T. Olson, N. Johnson, and V.A. Sample, editors. Defining sustainable forestry. Island Press, Washington, D.C.

Franklin, J. F., and R. T. T. Forman. 1987. Creating landscape patterns by forest cutting: Ecological consequences and principles. Landscape Ecology 1:5-18.

Frelich, L. E., and C. G. Lorimer. 1991. Natural disturbance re- gimes in hemlock-hardwood forests of the Upper Great Lakes Region. Ecological Monographs 61:145-164.

Godman, IL M., and K. Lancaster. 1990. Tsuga canademts (L) Carr. eastern hemlock. Pages 604--612 in IL M. Burns and B. H. Honkala, coordinators. Silvics of North America. Agriculture Handbook 654. U.S. Forest Service, Washington, D.C.

Graham, S.A., IL P. Harrison, and C. E. Westeil, Jr. 1963. As- pens: Phoenix trees of the Great Lakes region. University of Michigan Press, Ann Arbor, MichigatL

Hansen, A.J., T. A. Spies, F.J. Swanson, andJ. L. Ohmann. 1991. Conserving biodivers i ty in managed forests. Bioscience 41:382-392.

Harris, L D. 1984. The fragmented forest. University of Chi- cago Press, Chicago, Illinois.

Howe, IL, S. Temple, and M. Mossman. 1993. Forest management and birds in northern Wisconsin. Passenger Pigeon 54:297-305.

Jacobs, J. 1974. Quantitative measurement of food selection: A modification of the forage ratio and Ivlev's electivity index. Oecologia 14:413-417.

Jenkins, S.H. 1979. Seasonal and year-to-year differences in food selection by beavers. Oecologia 44:112-116.

Jaako P6yry Consulting, Inc. 1992. Generic environmental im- pact statement on timber harvesting and forest management in Minnesota: Wildlife technical paper. Prepared for Minnesota Environmental Quality Board, St. Paul, Minnesota.

Lorimer, C. G. 1989. Relative effects of small and large distur- bances on temperate hardwood forest structure. Ecology 70:565-575.

Mandelbrodt, B. B. 1977. The fractal geometry of nature. W. H. Freeman, New York.

Marquis, D.A. 1978. Application of uneven-aged silviculture and management on public and private lands. Pages 25-61 in Uneven-aged silviculture and management in the United States. United States Forest Service, Washington, D.C.

Mladenoff, D.J. 1987. Dynamics of nitrogen mineralization and nitrification in hemlock and hardwood treefall gaps. Ecology 68:1171-1180.

Mladenoff, D.J., and J. Pastor. 1993. Sustainable forest ecosystems in the northern hardwood and conifer region: Concepts and management. Pages 145-180 in G. H. Aplet, J. T. Olson, N. Johnson, and V.A. Sample, editors. Defining sustainable forestry. Island Press, Washington, D.C.

Mladenoff, D.J., and F. Stearns. 1993. Eastern hemlock regeneration and deer browsing in the northern Great Lakes Region: A re-examination and model simulation. Conservation Biology 7:889-900.

Mladenoff, D.J., M.A. White, J. Pastor, and T. IL Crow. 1993. Comparing spatial pattern in unaltered old-growth and disturbed forest landscapes for biodiversity design and manage- merit. Ecological Applications 3:293-305.

Noss, R.F. 1987. Protecting natural areas in fragmented landscapes. Natural Areas Journal 7:2--13.


762 In~9"a~ed landscape Oesi~ MlMeno~ et al.

Noss, R.F. 1991. Sustainability and wilderness. Conservation Biology 5:120-122.

Noss, 1Z F., and L. D. Harris. 1986. Nodes, networks, and MUMs: Preserving diversity at all scales. Environment Management 10:299-309.

Pastor, J., and M. Broschart. 1990. The spatial pattern of a northern conifer-hardwood landscape. Landscape Ecology 4:55-68.

Pastor, J., and D.J. Mladenoff. 1992. The southern boreal - northern hardwood forest border. Pages 216-240 in H.H. Shugart, R. Leemans, and G. B. Bonan, editors. A systems analysis of the global boreal forest. Cambridge University Press, Cambridge, England.

Pastor, J., and D.J. Mladenoff. 1993. Modeling forest harvesting effects on diversity, productivity, and habitat structure. Pages 16-29 in D.C. Le Master and IZ A. Sedjo, editors. Modeling sustainable forest ecosystems. American Forests, Washington, D.C.

Pickett, S. T. A., and J. N. Thompson. 1978. Patch dynamics and the design of nature reserves. Biological Conservation 13:27- 37.

Probst, J., and T. IZ Crow. 1991. Integrating biological diversity and resource management. Journal of Forestry 89:12-17.

Spies, T.A., and B.V. Barnes. 1985. A multifactor ecological classification of the northern hardwood and conifer ecosys- terns of Sylvania Recreation Area, Upper Michigan. Canadian Journal of Forest Research 15:949-960.

Tubbs, C.H. 1977. Manager's handbook for northern hardwoods in the North Central states. General Technical Report NC-39. North Central Forest Experiment Station, U.S. Forest Service, St. Paul, MN.

Tyrrell, L.E. 1991. Patterns of coarse woody debris in old- growth hemlock-hardwood forests of northern Wisconsin and western upper Michigan. Ph.D. thesis. University of Wisconsin, Madison, Wisconsin.

Tyrrell, L, and T. R. Crow. 1989. Old-growth in the Lake States: What is it? Why is it important? The Bulletin of the Botanical Club of Wisconsin 21:37-40.

White, M.A., and D.J. Mladenoff. 1994. Old-growth forest landscape t ransi t ions from pre-European se t t l emen t to present. Landscape Ecology, in press.

Wilcove, D.S. 1989. Protecting biodiversity in multiple-use lands: lessons from the US Forest Service. Trends in Ecology and Evolution 4:385-388.

Wolter, P. T., D.J. Mladenoff, G. E. Host, and T. R. Crow. 1995. Improved forest classification in the northern Lake States using multi-temporal Landsat imagery. Photogrammetric Engineer- ing and Remote Sensing, in press.


applying principles of landscape design and...

Documents