Patterns reflect arrangement in space (area). There may be various method to appraise the pattern. You may use the NDVI (through remote sensing data) for regional pattern. You may use NNI (Natural Neighborhood Index) for local distribution patterns.

One of the preliminary methods I use is a simple DCA of my relevés including different variables. It gives you a preliminary idea about your data...including species grouping, but also which variables are determinant over them.

I suggest you to read : De Cáceres, M., Legendre, P., Moretti, M. 2010b. Improving indicator species analysis by combining groups of sites. Oikos, 119: 1674–1684.

I found these articles useful:

http://www2.bren.ucsb.edu/~fd/misc/595/borgogno-vegpatternformation-revgeophys09.pdf

http://kellylab.berkeley.edu/storage/papers/2010-Kelly-etal-EI.pdf

I would first ask: Why do I need to know these vegetation patterns? Perhaps different hypotheses explaining patterns require different methods to describe the patterns and their details. How much detail is required? For instance, some use photos from airplanes to describe vegetation patterns at landscape scales, but then you don't always see what is under the tree canopy perhaps hiding vegetation patterns at some scales of analysis......

Some people would first like to understand the causes of vegetation patterns, and then use the patterns as proxies/measures of other environmental characteristics, like soil quality or water availability. In very dry regions, presence of vegetation obviously indicate presence and spatiotemporal dynamics in water sources...

I agree with Marcel; the numeric treatment of your data will depend on what you want to test.

I think that this book is a classic and can be extremely useful when analyzing big ecological data sets. http://store.elsevier.com/Numerical-Ecology/Pierre-Legendre/isbn-9780444538680/

If you going to held a pilot study then you need a biodiversity man and plant taxonomist. Then you can use Transect and quadrate method while, if you are wont to have monitoring for long time then you need to protect fix areas to get your measurements.

Long-term study requires long-term monitoring to capture the dynamics of the vegetation patterns involved. Plants grow or disappear. Trees become bigger. Forests get older, which may be related to spatiotemporal changes in biotic and abiotic factors.

A biodiversity man could also be represented by several people, like an animal taxonomist, a wildlife biologist, a behavioural ecologist, a physicists, etc.. in other words a multi-disciplinary team....

Again, you need first to divided your location into range site (information available online) then launched your action plan for implementation.

What i understood and I think GIS software has the option of this.Layered mapping

Not sure if you use ArcGIS, but there's a tool called Tabulate Area (http://resources.arcgis.com/en/help/main/10.1/index.html#/Tabulate_Area/009z000000w2000000/) that allows you to specify zones (which in you case, might be slope or aspect or some other environmental, climatic or topographic variable) and determine the vegetation pattern or area or variation within each zone. I've used this tool extensively with excellent results.

Hope this helps,

Justin