A unit of Lasting Forests
Sustained forests; sustained profits
evolving since March 30, 1999

of an Alternative Wildlife Resource Management

On Edge

Practical Analyses of Wildland Edges (with Programs)

This is a unit about edges such as between a forest and a pasture, but it is also about analyses of such lines, costs to produce them, and where they fit into the picture of wildlife resource management. Edge has been used as an index to biodiversity -- a complex term for describing a vague idea now included in more than 20 federal laws.

"My teeth are on edge!" speaks of the frustrated condition of many resource managers who find old ideas presented as discoveries, words misused, and phrases like "edge effect for wildlife" used in nonsensical ways as political blocks rather than in conversations that may help solve local, usually unique problems.

I am hopeful that this unit will help

• the practicing wildlife manager make better decisions about creating and managing edge;
• the researcher formulate new hypotheses, create better designs and analyses of edge, and learn whether it significantly affects his or her target species;
• the sports-person and outdoor enthusiast do better studies, see the outdoors more clearly, and understand what and why the managers and researchers seem to be doing;
• the student, caught in a cross-fire of words with different meaning and with limited experience, to learn about wildlife resource management with edge as one unifying theme.

Edge or" edge effect" has been said to be an important part of habitat management. Edge is an element of wildlife habitat. With such statements the problems begin. "Habitat" is no longer a useful word in faunal system management (Moen 1973, Coulombe 1978, Hoover and Wills 1987, Harris and Kangas 1988). An animal is inseparable from its surrounding. A toxic molecule in a lung - is it inside the animal or in its environment? The moisture in a cell of an animal's nose - inside or out? The union of a frog or earthworm with its over-wintering conditions suggests that all is one. Habitat, niche, community -landscape, management area, forest - all are discussed and debated because of confusion in meaning, ignorance of or unwillingness to accept first-use definitions, unwillingness to move to common terminology and to develop alternative taxonomy as new knowledge suggests meaningful change.

There are multiple meanings and assumptions about "habitat" and whether it expresses (1) individual, population, or species assemblage needs; (2) differences in seasonal and migratory needs; (3) fundamental requirements or their associates; (4) needs in disjunct areas, or (5) whether the animals themselves are considered habitat. in the landscape ecology sense (Rodiek and Bolen 1990), all factors may be present for "good" habitat for a life group, but the size of the area or its nearness to other areas may be inadequate. Nearness to a large suitable area may make a "bad" area capable of supporting a population. "Habitat" is too elastic. Coulombe (1978) found four themes for the concept: (1) life requirements for single species or life groups, (2) resource allocation between or within species (as in "partitioning"), (3) spatial distributions and types (as in "nesting"), and (4) quality for a species or group. He noted that "habitat" rarely related to scale since that for an organism could be described as (a) being the bark of trees in general, or that (b) of a tree species, (c) of a mountain ravine, (d) of a topographic zone, (e) of meta-spaces within a region), or (f) of a biome.

A replacement of the word "habitat" is needed. To improve the health or well-being of an individual animal, you hire a sensitive veterinarian. To improve on the well-being of a population of animals you hire a sensitive modern faunal system manager. Rarely does the faunal manager deal with individual animals, only the population and usually by manipulating where it exists. The forest areas for wildlife management are very complex and complicated places and the manager must master their complexity

"Habitat" , like ecology has its roots in '"home"' and thus is misleading as the basis for a viable full-scale management system that deals with animals that migrate, and that are influenced by global changes. The habitat of many animals is faunal (microorganism of a wide range, parasites, and other animals, predators, nest mates and pack and herd associates) as well as flora. Siblings in a den are habitat. I find feathers or the outer 2 centimeters of hair very "habitat like" but yield reluctantly to convention; these are a part of animal studies, rarely of habitat. Rather than habitat, the "faunal surround" or "faunal space" are useful phrases, implying the outside-of-the-animal environment - everything up to the skin surface.

The faunal system manager creates and maintains space for animal populations and for users of these populations. A hunting "area" is managed for animals (usually to increase them) but also for users so that animals produced may be hunted or otherwise used in ways to maximize benefits. The resource system manager does not manage "animal homes", except at personal risk of mismanagement. Management is of spaces for desired populations and their legal users. Faunal space is available forage and protection from energy loss; legal protection; adjacent animals hiding places; special landscape patterns; and species-specific seasonal provisions. Habitat is a word which readers may lay aside for use only in general discussions.

There are few topics more difficult to discuss than the spaces for forest wildlife. There are so many species and life groups, so many biomes, so many different requirements of each life group, so many techniques (Thomas 1979, Cooperrider et at. 1986, Bonham 1989, DeGraaf et al. 1989, Patton, 1992) and, if those were not enough reasons for the difficulty, then there can be added the potential conflicts between almost any user of the land and people with a wildlife interest. Resolving or easing these difficulties is possible, but it is very hard work. Knowing that there are long-standing, deep-seated disagreements, and that resolution is hard work may help reduce the frustrations and inspire concerted effort for people of the world ahead.

"Edge effect" is a well established phrase that is not consistently used and one for which there have been few definitive studies. It was used by Leopold in Game Management and later by Odum (1971) to mean the increases in species richness and variety often found at the edge of plant communities. The difference is compared to the numbers within the interiors of the contiguous communities. Kroodsma (1982) listed some of the difficulties in studying the perceived phenomenon:

• difficulties in selecting suitable study areas
• changes in vegetation (one or both communities) along a single linear feature over time
• debatable sampling strategy (parallel-to or across the edge)
• biases in estimating populations
• calculating densities within a thin strip
• uncertainty in the width (if any) in "edge"
• separating use of edge (the measured condition in the field) from edge effect, the animal response
• inconsistent responses from species, some populations seemingly harmed by, others helped by the presence of edge, and many (probably most) imperceptively affected.

Giles (1972) showed the means by which populations in contiguous areas may overlap, the sum of species in the overlap being greater than in either community. Sports-people hunt the edges because they know from experience that game occurs there, if not in variety, at least in abundance. Proven or not, edge is where richness and abundance are likely to be found. Since these two are primary objectives in many wildlife resource programs, edge needs to be managed to produce it, maintain or stabilize the amounts of it, or increase it for the species likely to be benefited by it. In the future the concept will probably lose favor and be replaced by summations of the effects of edge volume (described below) on single species for a season of the year.

Island Edge

Animal species numbers seem closely related to the size of islands ( real or hypothetical, such as a county, a timber stand, or a forest surrounded by agricultural land).The larger the island, the more species... to a limit, so the relationship can be expressed as:

S = C A ^z

where A is the area of the island, C a proportionality constant relating A to S, for example in relation to the units of measure used for the area. Z is the "magic number" derived within analyses of past data, probably some productivity factor p and some square-root relationship of area so

Z = f(p/2)

and it turns out that it is approximately 0.33 in many studies. (e.g., Connor and 19 ; Harris ).

The parsimonious relationship for S has been expanded by many authors to account for its failure in specific cases. Intuitively, the older the island (e.g., volcanoes) the more species will have reached it. Similarly, the closer to other islands or the mainland, the more invaders.

All islands are not circular, so a modifier for circularity ,kof the islands will probably remove some of the variance and improve the fit of such a model for S as used above. The more amoeboid the shape, the larger becomes k. It too is limited (rarely more than 5 in natural settings), so it may later be found best expressed within the exponential function. Revised, the model is:

S = kC (A ^z)

I have not seen the height of the vegetation used as an explanatory factor (along with area) for the number of species present. I perceive that animals live in volumes, not merely in areas. The more layers of vegetation (the thicker the floral mass covering the island), the more species and more numbers there will likely be. Clearly there are species and density differences between grassland covered islands and those covered by tropical rain forests. The elevation of the ground surface influences species. The vegetation layer is on something and that varies in productivity from very little in high-elevation mountainous islands to highly productive areas. Full scale topographic analyses of slope, aspects, and elevation seem to be needed to describe the land on which layers of solar-fixing plants consumed by animals live. The approximation, without age or distance-away phenomena, or between-layer effects is thus:

S = khC (A ^z)

Ocean islands are easily seen as closed systems. It is feasible to imagine a woodlot in an agricultural field or a grassy clearing in a forest as an island -- any distinctively different sub area. Cities are usually islands within counties; counties or regions are islands within a state. This perspective can be useful, at least to raise the question: is this county an island (e.g., of deer harvest) or is it like contiguous lands. The longer the edge of any island in an area:

• the greater the contiguities of "touches"
• the greater the corners
• the greater the variance in contiguities

Contiguity Matrix

A contiguity matrix for the area shown nearby is in Table 1. Such matrices can help understand and analyze the edges of any area. Edges exist between two distinctly different land-use or cover types. The matrix depicts the joins of the lines of contiguity. After the contiguity matrix, the length of each line is placed in a similar table. A table with all "1's" is likely to provide maximum opportunities for edge-related animals. The joins are important, but the length is also so some optimization or tradeoff made between the two. Hexagons for each type will tend to yield 6 joins per type even though they may be short and perhaps not functional. A minimum length that is relevant needs to be specified as a constraint on any solution system.

Table 1. The triangular contiguity matrix showing what cells are "touching" or contiguous (designated as 1) at the sides (not the corners) of each different land-use type. By convention and as will be seen later, areas are designated as contiguous to themselves.
	A	B	C	D	E	F	G	H	I	J
A	1	1	1	0	1	1	0	0	0	1
B		1	1	1	0	0	0	0	0	1
C			1	1	0	1	1	1	0	1
D				1	0	1	1	1	0	1
E					1	1	0	1	0	1
F						1	1	1	0	0
G							1	1	0	0
H								1	1	1
I									1	0
J										1

After length, the manager can enter in a similar matrix the width of each edge. Edges are lines on maps, but in the field, they have width. Some are thin, merely noticeable because of a wire fence. Others are wide, with forest influence extending some 12-15 feet out into a field; field influences extending 5-10 feet into the forest. An average width along each edge segment (the value 1 in the matrix) in each cover type should be entered.

Next the vegetation height for each segment is entered. Since the height changes with plant growth and age, the height of the edge changes, for example as a clear-cut regrows and becomes almost indistinguishable from the residual surrounding forest.

The result is that edge is a volume, an imaginary "tunnel" around openings or forest stands. It has height, width, and length, the product being a volume. My observations are that certain animal populations are related to the vegetation and conditions (opening, insects, etc.) in this volume, more than to the single factor of "length."

There remains one more factor to allow the manager to evaluate the edge volume. That is it precisely. What is its likely value to a species or species group of managerial interest. Each "1" in the matrix expresses the presence of an edge between 2 types. This is the essence of one measure of "interspersion", a term tossed around by wildlife managers but rarely quantified. Another term used is "juxtaposition" Herein, I believe that juxtaposition means the value of the junction of two types for a species. Where a stream and a food supply come together is "good". Where food plants touch plants providing wind protection , that also is "good". Red oak trees adjacent to white oak trees form an edge, but depending on the age, probably not significant to any (or many) species. Edges have estimable value to species being managed. That value is called the juxtaposition value and it too can be added to an identical table, multiplied through to get the relative, estimated value of the land use pattern... and the value of the pattern after the managers has completed proposed or planned project work. Some day, the optimum pattern will be computed. Until then, experienced workers can "play with" alternative patterns, simulate, and select patterns that will provide significantly improved scores for reasonable costs of land manipulation and re-planting. Stated above, the manager needs to stay aware that the edges like the tracts themselves change, but at different rates and the volume changes in almost inconceivable ways requiring species by species analyses. The analyses are time consuming, readily assisted by computers, and thus costs of the analyses suggest the need for a large scale of operation (e.g., analyses for areas within a region.)

An Exercise

The manager, for example, working on increasing edge length for superior quail space, may be given a habitat like that in the next figure. The task is to make a matrix (like the above) of the present fields and tracts, then, within the same ownership (the outer boundary) modify the cover pattern, retain the same four types of vegetation, and create a map for which the new triangular matrix will be filled with "1's." The task: make all tracts contiguous.There are several ways to achieve this, perhaps one example of equifinality.

You need to think through along with rough sketches, that edge unlike in mathematics is a line but it must have width. Imagine a "field" through a forest only 1 centimeter wide.!? There is a practical minimum width, perhaps a meter? How long must an edge be to occupy an acre? A hectare? If birds nest within 50 feet of edge, it is reasonable to think of the nesting acres created in the process if creating edge. What is the ratio of nesting area (say in acres) to length of edge (say in yards or 100 foot segments)?

The minimum edge for an area is that when it is a circle. What is the edge of a circular acre? A square acre? An acre with length 4 times its width? A circularity index is merely the actual length of a known area compared to the length of the edge of that same area if it were a circle. Amoeboid and jagged shapes rarely have an index above 5.0. Which has the most edge, an acre on a map as a triangle or an acre as a circle?

Imagine cutting 100 1/4th acre clearcuts as circles in a forest. How much edge length will you produce? If grouse nest 50 feet from the opening edge, how much nesting area potential will you create? If edge disappears gradually in 8 years from cutting, how can edge area be stabilized over the next 150 years?

Edge is always a line. Perhaps it is the vertical "wall" between forests and field. If so, then it must be reasonable to imagine animal populations that live in layers of the forest (like bats) to be influenced by horizontal layers, the imagined boundaries between the layers, the layer edges.

References

See the agricultural search site

Odum,E.P. 1971. Fundamentals of ecology, W.B. Saunders, Philadelphia, Pennsylvania

Kroodsma, R.L. 1982 Edge effect on breeding forest birds along a power-line corridor. J. Appl. Ecol. 19:361-370

Go to the top.

Last revision January 18, 2001.