If you want to know which setting to use with trinity:

1) map some reads with STAR (you will need some 32-34Gb memory). The index has to be built with an annotation built-in, or you can provide an annotation at run time. You don't need to output a bam file, just the gene counts:

  STAR --outFileNamePrefix temp. --genomeDir /path/to/STAR_index \
--outSAMtype None --readFilesIn file.fastq.gz --readFilesCommand zcat \
--quantMode GeneCounts

2) the gene counts output will be called temp.ReadsPerGene.out.tab, and will have four columns, one with gene names, and three with counts:

column 2:  counts for unstranded RNA-seq
column 3:  counts for the 1st read strand aligned with RNA (htseq-count option -s yes)
column 4:  counts for the 2nd read strand aligned with RNA (htseq-count option -s reverse)

You know already your data is stranded, so inspect columns 3 and 4 to decide which one is correct. The correct one will have lots of counts, whilst the other one will have very few counts.

3) if the counts are accumulated at the 3rd column, it means you have to use --SS_lib_type F with Trinity, whereas if the counts are accumulated on the 4rth column, you have to use --SS_lib_type R.