Thanks for your answer and links.

Probably I could calculate some Isolation Index with FRAGSTAT at patch level.

The Euclidean Nearest-Neighbor Distance (ENN) in FRAGSTAT calculates the distance among patches of the same type, considering the shortest edge-to-edge distance. The problem of this index is that, in case of wetlands, aquatic organisms need to move through channels and rivers to go from a wetland to another. Thus, the shortest edge-to-edge distance could be not realitic. On other hands, some approximations may be acceptable.

Hi Mattias,

What species are you studying? Not all species may need connectivity via water, especially the small ones are sometimes transferred by birds flying from one wetland to another. I also wonder about the timel dimension: wetlands do not need to be connected all the time to have species exchange, occasional flooding or higher water levels  can connect two wetlands that are separated most of the time.

Kind regards,

Wietske

Hi Wietske,

The analysis concerns macro-zoobenthos community. All wetlands are permanently connected through channels.

Many thanks,

Mattias

There are many measures of connectivity that find application in landscape ecology.  I would recommend the following article for a review and introduction to the conceptual issues:

P. Kindlmann and F. Burel. 2008. Connectivity measures: a review.  Landscape Ecology 23: 879-890.

These authors make the point that there is not one single connectivity associated with a landscape -- rather, connectivity depends upon the interaction of a landscape and an organism.  Thus, much depends on the behavior and dispersal capability of the organism.  In your case, even if wetlands are physically connected by a water channel, they may still be functionally fragmented for an organism with limited dispersal capacity.  Alternatively, if your benthic community includes aquatic insects which have a non-aquatic adult phase, (e.g., dragonflies), there may be functional connectivity between wetlands which have no direct surface water connection.

I'm not an expert on software available for connectivity analysis, but the following website has links to various packages which look promising:

http://conservationcorridor.org/corridor-toolbox/programs-and-tools/

Hope this helps...

Alan

Thanks Alan,

So, considering that organisms dispersion occur through passive ways (e.g. human or bird transports) and adult aerial stage for insects, I could calculate euclidean distances among patches as a measure of isolation (i.e. lack of connectivity).

Zoobenthos data seems to show weak but significant negative relationship  between wetland areas and number of taxa, and probably also wetlands connectivity could play a role.

Hi Mattias,

try with this software/alghorithm

http://www.conefor.org/