Description
The Input Features contain groups of disconnected clusters of features. There are more of these clusters than the Number of Clusters you specified, and the Spatial Constraints parameters do not allow these clusters to be combined. Consequently, there are more clusters going into the analysis than the number specified to come out of the analysis.
Solution
While you can increase the Number of Clusters specified, you should realize that the Spatially Constrained Multivariate Clustering tool will begin with the disconnected groups of features already representing clusters. For those clusters, the spatial relationships alone determine cluster membership (there is no analysis of the feature attributes for those clusters). The Spatial Constraints parameter that you specify determines which features can be clustered together and which features can't. If you see this message about groups of disconnected clusters, you may want to relax your Spatial Constraints parameter. Creating a spatial weights matrix using K nearest neighbors and increasing the Number of Neighbors parameter until there are no more disconnected groups of features will ensure that both spatial and data relationships are incorporated into the clustering algorithm. More information on how the Spatially Constrained Multivariate Clustering tool works with disconnected groups of features is described in the tool help.