This is a really important question, and one I also ask myself! In terms of ecological scales of operation, I don't think there is a fixed spatial definition of a landscape. 

One way of looking at it that I find helpful is to consider what constitutes a landscape-scale study. For a typical non-spatial ecology study, your sampling unit is a single patch of habitat, i,e, in your model each row of data corresponds to a patch of habitat encompassing for example a single home range for your study organism. A landscape study considering the same organism would for each data point be considering multiple home ranges and their spatial characteristics, or at the very least one home range plus some information on the make-up of the surrounding landscape. So the landscape scale varies in this way from e.g. a matter of square meters for springtails to hundreds of square kilometres for a large, mobile mammalian or avian predator. What constitutes a landscape depends on what species you are!

Lenore Fahrig explains this really well this essay, I can't think of a published reference off the top of my head but if I come up with anything I'll post it here.  http://134.117.48.8/PEOPLE/PROFILES/lenoreFahrig/When%20is%20a%20Landscape%20Perspective%20Important.pdf. Almo Farina's various textbooks are also helpful. 

Of course, human or sociological perceptions of landscape tend to be more fixed and landscape conservation operates at the human-landscape scale. So whilst spatial processes might operate down to quite fine scales, it might make sense to address problems at the scale at which management actions will tend to take place, if that makes sense. 

This is a really important question, and one I also ask myself! In terms of ecological scales of operation, I don't think there is a fixed spatial definition of a landscape. 

One way of looking at it that I find helpful is to consider what constitutes a landscape-scale study. For a typical non-spatial ecology study, your sampling unit is a single patch of habitat, i,e, in your model each row of data corresponds to a patch of habitat encompassing for example a single home range for your study organism. A landscape study considering the same organism would for each data point be considering multiple home ranges and their spatial characteristics, or at the very least one home range plus some information on the make-up of the surrounding landscape. So the landscape scale varies in this way from e.g. a matter of square meters for springtails to hundreds of square kilometres for a large, mobile mammalian or avian predator. What constitutes a landscape depends on what species you are!

Lenore Fahrig explains this really well this essay, I can't think of a published reference off the top of my head but if I come up with anything I'll post it here.  http://134.117.48.8/PEOPLE/PROFILES/lenoreFahrig/When%20is%20a%20Landscape%20Perspective%20Important.pdf. Almo Farina's various textbooks are also helpful. 

Of course, human or sociological perceptions of landscape tend to be more fixed and landscape conservation operates at the human-landscape scale. So whilst spatial processes might operate down to quite fine scales, it might make sense to address problems at the scale at which management actions will tend to take place, if that makes sense. 

I think the best approach is to consider what constitutes a landscape for the species or processes under study.  The landscape impacting on an ant is generally going to be reduced compared to a deer for instance.  We need to get away from defining landscapes based on human perceptions, and consider how the species we study perceive them.

I can't improve on Chris's explanation of what constitutes 'landscape scale'. 

Very interesting discussion! I agree with you saying that the spatial scale of “landscapes” depend on the researcher’s issue or perspective. You described a very ecological, species-related view.

Let me add my landscape ecological point of view: In my work I consider “landscapes” as functional unit. Hence, landscape is a –to a certain degree- homogeneous functional unit, which is different from its adjacent units. That relates firstly to the primary landscape structure, which means geology, soil, vegetation, climate, etc. Secondly, landscapes are characterized by the secondary structure, which addresses the human influence on the primary structure due to management, settlement etc. E.g. Olaf Bastian (see his profile) has carried out a lot of work on landscape classification, e.g. http://link.springer.com/article/10.1007/s10980-005-5224-1. The human dimension often plays the crucial role, I would say. When it comes to practical research questions, dealing with landscape planning and management, I have experienced that finally we are working with administrative units, planning regions, management units etc.

In dependence from the research question, I would generally try to make my investigations at landscape level, referring to a holistic definition (http://carmelacanzonieri.com/library/6108-LandscapeEcoPlanning/Bastian-LandscapeEcologyUnifiedDiscipline.pdf). In order to derive practical implementation, the translation to administrative regions seems to be inevitable.

Coming back to your question: finally, apart from the socio-ecological perspective, the researcher's interest (or interest of funding institution) defines the spatial scale of your landscape.

...and here an example for German landscapes (definition and spatial scale):

https://belgeo.revues.org/13688?lang=de

Generic definition from one of the best - Jingle Wu (2013) (papers attached):

Landscape A geographic area in which variables of interest are spatially heterogeneous. The boundary of a landscape may be delineated based on

geographic, ecological, or administrative units (e.g., a watershed, an urban area, or a county) which are relevant to the research questions and objectives.

Thanks everyone for the very good contributions.

When I or other ecologist think of a landscape, we very much think of it from a species perspective, thus limiting the maximal scale a landscape can have through a focal species dispersal / mobility capabilities as Chris pointed out. However for a social scientist or someone who interacts with social-ecological properties of a landscape, the very same landscape is defined by how people live and interact along-side nature and how processes within nature affect the people living in it.

I kinda think those two perspectives often differentiate quite a lot from each other (not only spatially).

I acknowledge that both approaches are dependent on the question / hypothesis to be answered, but how is that working out for instance in conservation management perspective? What if for instance the people in a certain "landscape" depend upon the ecosystem services delivered by them through the nature they are living in. This very nature is acting on a scale, where people benefit the most from it. They might still appreciate a forest far away from their village, but they are not dependent on it and nether are there (so I assume) no processes linked to this social community. On the other hand local conservationist intend to put the very same landscape in a conservation focus to protect a species (for instance a brown bear) that has a much larger home-range and disperses into areas far away from the people view/dependencies of their landscape.

Is it possible to address both in landscape management framework? Are landscape boundaries set by humans purely artificial ? Can they be justified from a ecological point of view (via for instance a process such as seed dispersal)? Too often we seem to just correlate patterns we observe on a local level with an indiscriminate scale of other patterns we observe on greater spatial scale (the landscape).

Martin

There is a journal called Landscape Ecology.  Papers therein will answer the two questions in your last paragraph.   On the second question, we in the US are exploring something we call Ecosystem Based Management; ecological versus political boundaries (I don't know of any species other than humans that recognize the latter!).  Here's a proceedings we did on the subject.  Regards,

Mike

Well, in Ecology, we define "landscape" as an area that humans can perceive as relatively homogeneous functional unit, in terms of environment (geology, topography, vegetation, etc.) plus the occurrence of human related structures (urban tissue, settlements, etc.) that can deeply change the natural aspects of the environment (forest with city, farmlands in mountains, etc.). I agree with Susanne's explanation before. The combination of these factors can create the units of landscape. For this reason, the landscapes can be defined at a multispatial scale.

Other authors stress mainly the need for a morphographic approach, in order to define topographic land-use untis, on landscape studies. One good explanation: http://www.kolleg.loel.hs-anhalt.de/studiengaenge/mla/mla_fl/conf/pdf/conf2003/41blasch.pdf

However, also if landscapes remain as the human's "reading" of the territory, several other animal or vegetal species can read the territory in similar ways. Some bird species, for example, can be related to some landscape previously defined by humans, because correlated to environmental, climate or opportunistic variables offered in those landscapes. 

Any possible landscape boundaries, if defined under ecological reasons, often are present more like a "gradient zone" or "transitional area" between two different landscape units previously defiend, than a pure boundary.

Other interesting aspects are constituted by the typical structures capable to "penetrate" and cross different landscape units: linear structures as roads. Roads can cross different units, adding some blending effects, very hard to quantify in ecology.

Hello,

Interesting question. I don't think we can have a unique answer to it. Rather, the "landscape scale" is dependent upon the scale of the studied phenomena or species landscape use.

Hello Martin, 

I have learned a lot from the contribution of others. However, i read that a landcaspe scale is between 10^2m - 10^3m, a Regional scale is between10^3m - 10^5m, while a Biogeographic scale is between 10^5m  - 10^7m. I saw this in Nathan et al. 2008: Mechanisms of long-distance seed dispersal. and here is the url to the paper : http://www.sciencedirect.com/science/article/pii/S0169534708002723

I hope this spatial delineation of a landscape scale is good enough.

Good Luck

i think from a landscape ecology view, we categorize it as the boundary made by the change in the soil type and formation from one landscape to the other. this will give us  a surrogate measure of the spatial opportunity provided by a particular vegetation within the environment.

Here what is written in the e-book Ecology: Concepts and Applications of Manuel Molles:

"A landscape is a heterogeneous area composed of several ecosystems. The ecosystems making up a landscape generally form a mosaic of visually distinctive patches. These patches are called landscape elements. Landscape ecology is the study of landscape structure and processes. Landscape structure includes the size, shape, composition, number, and position of patches, or landscape elements, in a landscape. Landscape structure influences processes such as the flow of energy, materials, and species across a landscape. Landscapes are structured and change in response to geological processes, climate, activities of organisms, and fire.  "

A 2002 book chapter by John Weins and others offers a useful discussion on the issue of the "landscape scale":

Wiens, J. A., B. Van Horne, and B. Noon. 2002. Integrating landscape structure and scale into natural resource management. Pages 23–67 in J. Liu and W. W. Taylor, editors. Integrating landscape ecology into natural resource management. Cambridge University Press.

Also, T.F. H. Allen made a forceful argument against conceptualizing the landscape in a hierarchical framework of scales in this chapter:

Allen, T. F. H. 1998. The landscape “level” is dead: persuading the family to take it off the respirator. Pages 35–54 in D. L. Peterson and V. T. Parker, editors. Ecological Scale: Theory and Applications. Columbia University Press.

Excellent question!

Folks

There is an unbelievably large and diverse literature on this subject.  We have only begun to scratch the surface of the topic with our comments, and the few citations we listed.  Go to Google Scholar and type in the term to see what I mean.  You need to read fairly widely to get an appropriate perspective, and not get overly biased by our (to date) relatively small sample!

Mike

I think your landscape needs to be defined by the movement or distribution of the key variables in question. So it can be movement of soil or water downslope, or the foraging range of the animals you are interested in, for example. Both of those examples can be scaled up from a single hillslope to an entire watershed or coastline or biome, depending on your question.

Landscapes are dynamic systems in space and time, so think of them in terms of a constantly changing system and define the boundaries of the relevant changes for your study. It might be patterns of grazing across seasons, which is in turn influenced by a combination of movement of nutrients, spatial distributions of soils and topography, and management (fences, fire). For an ornitholigist, it might be far bigger, and influenced by the scale and distribution of foraging and nesting sites of groups of birds, which could expand the greater landscape over thousands of square kilometres. Or it could be patterns of plant seed dispersal, which is influenced by prevailing winds or movements of animals, as well as the distribution of suitable habitats for germination. All of these examples are dependent upon movement and change of both living and non-living things, within a matrix of static (or more slowly changing) abiotic factors. Define the movements, and you will be a long way towards defining your landscape.

This is a tricky question which ecologists must be careful in advancing a straight answer. I quite agree with the views of Michael P. Weinstein above but rather advice that landscape rarely respect boundaries. In other words, the subject matter of the research may well define the limit of landscape and that landscape is a geographical expression of a spatial entity.

By  Beruchashvili, N.L. (Beroutchachvili, Nicolas ),1979,1995.... , Landscapes imply natural complex formed under the conditions of a homogenous geological basis, in a single type of dominating relief (e.g., alpine ancient glacial, flat country accumulative, etc.) and homogenous climate (humid subtropical, moderate dry, etc.). Under these conditions, the combination of vegetation and soil, typical of the landscape, is formed. The fact that the landscape has been formed as a result of the successive joining of smaller natural complexes - a facies, is of great significance. They are closely interconnected and dependent on one another while creating the framework of the landscape. The area of landscapes in mountainous countries ranges from several to dozens of square kilometers.

       Landscapes to some extent are transformed by a man. The scope of man's activity depends on the amount of the area of the landscape that is occupied by cultivated land, cities and villages, as well as industrial and hydro-technical constructions, roads and other units. 

The landscape has horizontal and vertical structure as well as temporary structure. The vertical structure implies dividing of the landscape into the certain circles in its vertical profile (in geography these circles are referred to as geohorizons). The horizontal structure is connected with division of the landscape into small natural-territorial complexes (facies). The temporary structure is connected with dynamics of state of landscapes. The horizontal structure is well decoded on the basis of space teledetection. In different seasons of the year different results can be received while studying temporary structure. For example, in summer and in winter. The vertical structure is decoded only through indirect signs.

author Beroutchachvili, Nicolas  http://www.worldcat.org/identities/lccn-n85193986/

Alternative Names: Beruchashvili, N.L.

1.      The Caucasus: Landscapes, Models, Experiments - UNEP, GRID Arendal, WORLD BANK,  Tbilisi, Ed. Tbil.St.Univ., 1995, 314 p. (in Russian, Englich)

2.      Geosystems et  paysages.  Bilan  et  Methodes",  Paris,  Armand Colin, 1991, 302 p., (co-author G.Rougerie).

etc.,etc.

often in science especially in the landscape ecology, there is spatial heterogeneity, so if one were to study a certain process in the landscape ecology, they would have to be very specific about the aims of their study and their boundaries therein, because spatial heterogeneity does exist in nature and also differs with relation to time (after events of floods and erosion). 

i do however agree with your view martin 

Here is a paper that clearly describes that defines landscape ecology

• Pickett, S. T., & Cadenasso, M. L. (1995). Landscape ecology: spatial heterogeneity in ecological systems. Science, 269(5222), 331.

Considering landscape as a complex system, the scape depends on the type of landscape you are assessing. For example, urbanscapes or riverscapes should have a different scaled approach...