I'm performing a DCA analysis of a data-matrix of 150 releves and the result is 3.3 SD units for the first axis and 3.8 SD units - for the second. The configuration of releves is reasonable and they are well differentiated but these values are confusing. Most analyses I've seen have resulted with smaller length of second axis than the first but now the case is different. Does this mean that DCA is not appropriate or there is something else? Total gradient length is 3.8 SD units long and for that reason I chose DCA.