bcftools call: why is this a 1/1?

[samhokin]

I can't for the life of me figure out why bcftools call genotypes this position (and many, many others like it) as 1/1:

Here's the input VCF line:

10	227179	.	C	T,<*>	0	.	DP=191;I16=91,87,7,2,6502,238474,333,12321,235,3775,420,25200,3675,85899,97,1569;QS=0.733,0.267,0;VDB=0.00520152;SGB=-0.662043;RPB=0.000875247;MQB=0.0198712;MQSB=0.892843;BQB=0.980427;MQ0F=0.0209424	PL:DP:ADF:ADR:AD	64,255,0,255,27,95:187:91,7,0:87,2,0:178,9,0

and here's the output from bcftools call -m :

10	227179	.	C	T	34.7149	.	DP=191;VDB=0.00520152;SGB=-0.662043;RPB=0.000875247;MQB=0.0198712;MQSB=0.892843;BQB=0.980427;MQ0F=0.0209424;AC=2;AN=2;DP4=91,87,7,2;MQ=3	GT:PL:DP:ADF:ADR:AD	1/1:64,255,0:187:91,7:87,2:178,9

Note that DP4=91,87,7,2 -- there are 178 REF reads and 9 ALT reads used to make the call. How on earth is that ALT homozygous "1/1"??

I hugely appreciate any explanation of this call!