The Problem of the Urban Landscape

I was recently accused of bringing a wilderness or rural approach to urban landscapes.  This surprised me since I have long had what I think of as an inclusive concept of landscape.  Oddly, an inclusive, urban-friendly view of landscape can be considered to emerge from each of the two different ways that ecologists have used the term landscape.

Landscape as a Rung on the Scala Naturae

Fig 1: Levels of Organization
Traditionally, ecologists and philosophers of biology have found it useful to think of the world as arranged as a nested hierarchy of units.  This nested hierarchy runs something like this: atom – molecule – cell – tissue – organ – organism – population – community – ecosystem (Odum, 1971; Mayr, 1997).  From this sequence, it is clear that larger kinds of object are made up of smaller kinds, and any given kind of object can be a component of still larger objects.  Each kind of object is called a “level of organization” (Figure 1).  The ladder or stairway of units is supposed to represent increasing complexity as one climbs upward.
To give a specific example from one of the nodes above, populations of a given species are composed of all the individual organisms of that species in an area of interest.  Populations are nested within communities.  Continuing up the ladder from ecosystem, contemporary ecologists group communities into ecosystems, and in turn assemble ecosystems into landscapes.  This may be continued on up through biomes and then to the entire planetary biosphere.  In this level-of-organization approach to scientific objects, landscape is considered to be a level between ecosystem and biosphere.  Forman and Godron (1986) in their pioneering North American textbook on landscape ecology, define landscape to be a kilometers-wide area encompassing ecological heterogeneity or patchiness.  The patchiness is expressed as different ecosystem types, such as fields, forests, and lakes.  This is consonant with the way landscape has been used in art and perhaps also in design.  Forman and Godron, however, are clear that landscapes must include humans and human artifacts and settlements.  Patches, corridors, and matrix are the  fundamental kinds of elements of the Formanian landscape (Figure 2).
Fig. 2: Patch (P), Corridor (C) and
Matrix (M) in a landscape.
The characteristics of each level of organization are distinct and specific to that level.  Populations have characteristics like age structures and sex ratios.  Communities have species richness and three dimensional structure.  A mass of molecules of a particular compound will have a boiling point, but it would be silly to try to apply that kind of attribute to cells.  Cells, on the other hand, are characterized by internal structures like vacuoles, mitochondria, and so on.  Specialists studying each level of organization employ appropriate observational tools and protocols, different kinds of measurements, ask research questions, or seek applications that are especially relevant to objects on their focal level.  Unfortunately, they often argue about which level is the best to focus on, with the unjustified bias that smaller, lower levels are better than higher, more inclusive levels.  A true understanding of the differences in subject matter and methodology for different levels would do away with such arguments.

Landscape as an Approach to Analysis

The single hierarchy or stairway of nature, presents some problems, however.  Jim MacMahon and colleagues (1978) noted that if one starts with a focus on organisms, several different hierarchies can be conceived.  Leading up to organisms, they maintain the atom-through-organ nesting.  But once one has focused on organisms, several perspectives are possible.  One perspective is through the lens of evolution.  Populations are a unit within which genetic variation and subsequent natural selection are made manifest.  Along this nested hierarchy, species, genera, families, and the rest of the taxa produced by evolution are arrayed above the organism level.  Another perspective is the focus on the processes of metabolism, that is energy and matter exchange, above the level of the organism.  This arm of hierarchy might include ecosystems, biomes, and the biosphere as units within which the flow of energy and the cycling of limiting nutrients and contaminants result from the physiology and decomposition organisms and their aggregation into larger units.  A final hierarchy above organisms runs through demes, populations, and communities.  This is called a coevolutionary hierarchy, and its intention is to emphasize resource partitioning and adaptive interaction among organisms of different species.
Where does landscape fit in these sequences?  MacMahon et al. (1978) are silent about this, becsuse landscape ecology as a discipline did not become widely understood until the mid 1980s.  Often, ecologists interpose landscapes in the matter-energy flow hierarchy, to suggest that different kinds of ecosystems exist and are spatially arranged over large areas.  The ambiguity of where to place landscapes arises from the recognition that there are different hierarchies in which organisms may participate, and which of the hierarchies is chosen depends upon the research interest or observational window. 
This ambiguity was resolved by Allen (1998), who reframed common ecological units not as some preexisting ranking of natural reality, but as reflections of the interaction of the observer and the material world.  In other words, criteria of observation may be a more effective way to classify ecological observables than a fixed stairway of nature. 
Criteria of observation are illustrated by the ways that MacMahon and colleagues (1978) placed organisms in several hierarchies.  One criterion is autecological – the combination of organism physiology and anatomy.  This criterion emphasizes the role of organisms in assimilation of energy and nutrients, and how they transform them into other forms.  Continuing this concern into the larger world in which those materials are in turn used by still other organisms, whether they be consumers, disease agents, or decomposers, identifies an ecosystem or biogeochemical criterion of observation.  In contrast, how organisms relate to one another through the use of the resources already captured by or sought by other organisms, is portrayed as a niche or community hierarchy.  Here, organisms are viewed as competitors, mutualists, or commensals, for example.  Finally, the hierarchy of taxonomy is an evolutionary one, reflecting the increasingly inclusive units of inheritance and deep history.
Allen and Hoekstra (1992) generalize and extend the insights pioneered by MacMahon and colleagues (1978).  The general criteria of observation include 1) physiology, 2) spatial heterogeneity, 3) evolutionary potential and mechanisms, 4) transformation of energy and matter, 5) species coexistence, and so on.  These concerns are not nested hierarchically, nor represented uniquely by specific spatial scales.  The concepts, models, and empirical understanding required by each of these criteria of observation may be applied or generated on any number of spatial scales.  Criteria of observation are scale independent, and can thus be applied on any spatial scale.

Landscape as a General Concept and Criterion

Following the view of criteria of observation, a general definition of landscape emerges.  Most simply and generally, a landscape is a collection of spatial areas that differ in some attribute(s) of interest.  These areas are often represented as patches, but landscapes may also represented by fields or gradients arrayed in three dimensional space.  Landscapes can exist at any spatial scale, and apply to any collection of areas that differ from one another in a defined way.  Note that patches do not have to be internally homogeneous.  Rather, they can represent characteristic mixtures of structural elements or processes.  All that is required is that the mixture in one kind of patch be discernibly different from the mixtures in other kinds.  The same is true for the mechanisms or outcomes of processes, if that is the way that patches are to be differentiated.

Landscape: City or Country?

Fig. 4. Riverine patchiness in Kruger
(S. Pickett)
Fig. 3. Patchiness in Baltimore
(S. Pickett)
It is clear that this general definition of landscape applies just as well to a city (Figure 3) as to Kruger National Park (Figure 4).  Spatial differentiation is notorious in cities (Clay, 1973), and is key to understanding them, managing them, and assessing environmental inequities, for example (Pickett et al. 2011).  Urban patches can be distinguished based on architecture, road patterns, the kind and amount of vegetation, the density of human inhabitation or use, and the characteristics of the human populations inhabiting different areas (Cadenasso et al. 2007, 2013).  In places like Kruger, spatial differentiation may be the result of vegetation responses to slope, elevation, and drainage.  In addition, Kruger patchiness is generated by the grazing of large ungulates, by fire, and as a legacy of local resource management.  
The message is that entire urban areas are as much landscapes in the conception of criteria of observation as are rural or wild areas.  Urban landscapes involve flows and transformations of energy, matter, organisms, and information, just as do wild and rural ones.  Of course, they also have massive flows of social capital, power, financial capital, and commodities, among others.
This may be different than some uses of landscape in urban design or landscape architecture.  Notably, landscape in the urban context as described here does not just mean the conspicuously green spaces, and certainly does not require that fluxes of energy, matter, information, and organisms take the same form as they do in wilder or more rural areas.  Landscape is an important and gereralizable concept in an inclusive urban ecological science.


  • Landscapes are heterogeneous mosaics or three-dimensional spatial fields containing contrasts in structures or processes.
  •  Landscapes can exist on any spatial scale.
  • Wild areas and urban areas can be represented in their entirety as landscapes.
  •  Landscapes can change through time, and their component patches or gradients can shift as a result.
  • The areas making up landscapes can be differentially connected to each other, and to areas outside of an immediate landscape by teleconnections.


Cadenasso, M. L., S. T. A. Pickett, and K. Schwarz. 2007. Spatial heterogeneity in urban ecosystems: reconceptualizing land cover and a framework for classification. Frontiers in Ecology and Environment 5:80-88.
Cadenasso, M. L., S. T. A. Pickett, B. McGrath, and V. Marshall. 2013. Ecological heterogeneity in urban ecosystems: reconceptualized land cover models as a bridge to urban design. Pages 107-129 in S. T. A. Pickett, M. L. Cadenasso, and B. McGrath, editors. Resilience in ecology and urban design: linking theory and practice for sustainable cities. Springer, New York.
Pickett, S. T. A., G. L. Buckley, S. S. Kaushal, and Y. Williams. 2011. Social-ecological science in the humane metropolis. Urban Ecosystems 14:319-339.