Analysis of NVC using Mavis

[laurangutan]

Hello All

I am trying to better understand the interpretation of the coefficients that are generated by Mavis when data is analysed in terms of NVC communities. Below are the first 3 communities suggested by Mavis for some data from quadrats of homogenous vegetation analysed together. The results are a bit inconclusive but what I am trying to understand is how well my data matches the suggested communities. Does 54.95 (I assume this is a percentage?) imply a strong match? What does a coefficient of say 50 mean compared to 20 or 70?

NVC: MG4 54.95
NVC: MG9 53.75
NVC: MG9a 53.45

Regards

Laurangutan

[Johnny Redgate]

I've only used MATCH myself so my help may be limited. Higher score in MATCH is a better fit to the NVC community. I found that field data scores above 60 weren't common. More of an index than a percentage. Similar scores (within 5-10 or so) for several on the face of it fairly different looking communities is quite frequent because many species are common to several communities. Treat it as a guide and refer to the NVC texts for detailed community descriptions for the constants and the indicator species + habitats, distributions etc. Don't rely only on the computer programme.

Other than that try and get hold of the programme manual or search online. Good luck and keep doing those quadrats!

[amydenness]

Originally Posted by laurangutan

Hello All

I am trying to better understand the interpretation of the coefficients that are generated by Mavis when data is analysed in terms of NVC communities. Below are the first 3 communities suggested by Mavis for some data from quadrats of homogenous vegetation analysed together. The results are a bit inconclusive but what I am trying to understand is how well my data matches the suggested communities. Does 54.95 (I assume this is a percentage?) imply a strong match? What does a coefficient of say 50 mean compared to 20 or 70?

NVC: MG4 54.95
NVC: MG9 53.75
NVC: MG9a 53.45

Regards

Laurangutan

The higher the coefficient the more confident MAVIS is of the community. However, the results should only be used as a guide and you need to consult the floristic tables and descriptions in the NVC volumes in order to decide which community best fits your survey results. A coefficient of 50+ is pretty good but due to the amount of species overlap there is between similar communities you cannot simply infer that the community with the highest MAVIS score will be the correct one. I find it is best to use MAVIS and then look at the communities it suggests are most likely and make a judgement using both sets of information.

[Johnny Redgate]

Looks like we agree, Amy