I'm working on marine fish ecological data, on an irregular time scale with data only for 1993,1994,2006,2007,2009,2014,2016 and 2017. For each year, I have computed several taxonomic and functionnal diversity indices over my study area. I would to determine if there are trends in my data series, but I'm not very confident with the method I should use. I tried with pastecs :: trend.test() but this Spearman's rank test considers that the input is a regular time series, and I imagine that the 12 years gap give a false output of the test. Would it work with a simple lm(indice~year) with using the positive or negative correlation as a trend ? Or should I find an other test?