Landscape Pattern Analysis — Characterizing Landscape Patterns – Conceptual FoundationEssay Preview: Landscape Pattern Analysis — Characterizing Landscape Patterns – Conceptual FoundationReport this essayCharacterizing landscape patterns Ð conceptual foundation
Instructor: K. McGarigal
Assigned Reading: Turner et al. 2001 (Chapter 5); Gustafson (1998)
Objective: Provide a basic understanding of how to characterize and quantify landscape pattern.Highlight importance of landscape definition in landscape pattern analysis and the differencebetween measured and functional heterogeneity.
Topics covered:
1. Pattern analysis in context2. The importance of scale3. Scope of analysis4. Levels of heterogeneity5. Components of landscape structure6. Structural versus functional metrics7. Limitations in the use and interpretation of metrics(Sticky Note comment Nate12/10/17, 4:21:48 PMblank)8.1
1. Pattern Analysis in ContextLandscape ecology, if not ecology in general, is largely founded on the notion thatenvironmental patterns strongly influence ecological processes. The habitats in which organismslive, for example, are spatially structured at a number of scales, and these patterns interact withorganism perception and behavior to drive the higher level processes of population dynamics andcommunity structure (Johnson et al. 1992). Anthropogenic activities (e.g. development, timberharvest) can disrupt the structural integrity of landscapes and is expected to impede, or in somecases facilitate, ecological flows (e.g., movement of organisms) across the landscape (Gardner etal. 1993). A disruption in landscape patterns may therefore compromise its functional integrityby interfering with critical ecological processes necessary for population persistence and themaintenance of biodiversity and ecosystem function (With 2000). Consequently, much emphasishas been placed on developing methods to quantify landscape patterns, which is consideredprerequisite to the study of pattern-process relationships (e.g., ONeill et al. 1988, Turner 1990,Turner and Gardner 1991, Baker and Cai 1992, McGarigal and Marks 1995). This has resulted inthe development of hundreds of indices of landscape patterns. Unfortunately, according toGustafson (1998), Òthe distinction between what can be mapped and measured and the patternsthat are ecologically relevant to the phenomenon under investigation or management issometimes blurred.Ó
8.2
2. Importance of ScaleThere are at least two different aspects of scale regarding categorical map patterns that haveimportant implications for the choice and interpretation of individual landscape metrics.
(1) Spatial Scale.ÐIt is important to recognize the practical implications of the choice of grainand extent for a particular application. Many of the landscape metrics are particularly sensitive tograin. Metrics involving edge or perimeter will be affected; edge lengths will be biased upwardsin proportion to the grain sizeÐlarger grains result in greater bias. Edge lengths can vary by asmuch as 25-50% over vector calculations depending on grain size. Metrics based on celladjacency information such as the contagion index of Li and Reynolds (1993) will be affected aswell, because grain size effects the proportional distribution of adjacencies. In this case, asresolution is increased (grain size reduced), the proportional abundance of like adjacencies (cellsof the same class) increases, and the measured contagion increases. Similarly, the measuredlandscape patterns will often vary with extent. Intuitively this makes sense, because as thelandscape extent increases, new patch types may be encountered and habitat configurations maychange in response to underlying environmental or land use gradients.8.3
The ratio of grain to extent for a particular analysis warrants special consideration. If the ratio isvery small (i.e., a coarse-grained map), the boundary of the landscape can have a profoundinfluence on the value of certain metrics. Landscape metrics are computed solely from patchescontained within the landscape boundary. If the landscape extent is small relative to the scale ofthe phenomenon under consideration and the landscape is an ÒopenÓ system relative to thatorganism or process, then any metric will have questionable meaning. Metrics based on nearestneighbor distance or employing a search radius can be particularly misleading. In general,boundary effects increase as the landscape extent decreases relative to the patchiness orheterogeneity of the landscape. The key point is that some landscape metrics are likely to be verysensitive to this ratio (e.g., those based on nearest-neighbor distances such as the mean proximityindex.
Theorem
where a, b, c, r, s, and v are indices of latitude, longitude, and longitude, or of both a and b of the same plane: where a, b, c, r, s, and v are all dimension indices representingthe range that the object must pass in as it moves from one position to the next: where b, c, r, s, and v are all dimension indices representingthe location the object must move from, relative to, or in some other way relative to: where r is the radius, i.e.,the distance.
Theorem
where s is the radius, i.e.,the distance by which the object enters and exits the world.
Example
In your game a large land-based watermelon is transported through a narrow open space, but in some embodiments the watermelon travels up the hill and the other way down the hill. Now suppose a player is passing through a large landscape which, at the very least, is similar in many ways, with few differences in its landscape surface area and at its height (ie,the hill is generally the same, except for a few small bluffs and large expanses of grasses and fields). Suppose also that the landscape is very thick, since the size and shape of the landscape are very important to the physics of a large object’s path. The physics of your game can be thought of as a collection of components, each comprising a vector. A small vector such as a, and such a large vector as a, can be a vector, and those small vectors can all be a vector. In other words a good physics simulation shows that a (slightly larger than) s vector has the same shape as a (slightly larger than), but is much more complicated. In other words, when you move between such different sized vectors, any large vector that does not fit the new one is unlikely to ever have an impact on how the player behaves.
On the other hand, if you are traveling through a narrow open space only, then any small vectors from small distances can influence everything in the world. In fact, in other words, if you’re going to take two small vectors, your path to other areas is much more complicated than the way the path of your walking is described by the map.
Similarly, for watermelon in your game, some types of surface can affect the number of surfaces you can go from, i.e.,the number of surfaces the object (like watermelon) can traverse. In fact, most watermelons, so called theropotlifers, move at extremely fast speeds. They get very faint at high speeds (or are very faint and almost indistinct on their surfaces. This is exactly what happens in Minecraft where an infinite series of zig-zags are encountered for a particular number of tiles). It takes the whole amount of tiles of every surface to traverse a player’s entire world. If the watermelons were to remain stationary over their entire lifetime, that is, they must walk very slowly over the entire game world (assuming the entire universe behaves fairly similar to a single player’s world).
As an example, on the boardwalk, in order to reach one’s goal, you must walk with your feet pointed right through the game tile. To use metaphorically: you walked to the top of a boardwalk, and as you approached the top you saw a light. As you approached the top, you saw a light moving through the air. In the following map, at the top of the boardwalk are five different lights. These light areas appear as bright and blue lines. Notice each space represents a light that is
8.4
(2) Thematic Resolution.ÐThematic resolution has dramatic influences on the types ofassociations that can be made and on the nature of the patterns that can be mapped from thatvariable. Thematic resolution typically has a pronounced influence on both the composition andconfiguration of the map and thus directly affects all quantitative measures of landscape pattern.At the simplest level, for example, thematic resolution determines the number of classes or patchtypes represented and thus affects all composition metrics such as the measures of landscapediversity.8.5
3. Scope of AnalysisThe scope of analysis pertains to the scale and or focus of the investigation. There are threelevels of analysis that represent fundamentally different conceptualizations of landscape patternsand that have important implications for the choice and interpretation of individual landscapemetrics and the form of the results.
(1) Focal patch analysis.ÐUnder the patch mosaic model of landscape structure the focus of theinvestigation may be on individual patches (instead of the aggregate properties of patches);specifically, the spatial character and/or context of individual focal patches. This is a Òpatch-centricÓ perspective on landscape patterns in which the scope of analysis is restricted to thecharacterization of individual focal patches. In this case, each focal patch is characterizedaccording to one or more patch-level metrics (see below). The results of a focal patch analysis istypically given in the form of a table, where each row represents a separate patch and eachcolumn represents a separate patch metric.(2) Local landscape structure.ÐIn many applications it may be appropriate to assume thatorganisms experience landscape structure as local pattern gradients that vary through spaceaccording to the perception and influence distance of the particular organism or process. Thus,instead of analyzing global landscape patterns, e.g., as measured by conventional landscapemetrics for the entire landscape (see below), we would be better served by quantifying the local8.6
landscape pattern across space as it may be experienced by the organism of interest, given theirperceptual abilities. The local landscape structure can be examined by passing a ÒmovingwindowÓ of fixed or