Why does gscatter create extra groups that are already represented?

I am using gscatter to plot data and then color the data by site. I have 6 sites (PK, PO, EK, EO, LK, LO) to plot but I get 9 groups from gscatter. I am working around it for now by coloring redundant groups the same but I would like to understand what is happening. Does anyone have any insight about this?

This is my code for plotting the data (I have also attached a .mat file with the data in question).

h=gscatter(DIC,TA,SiteID);
EKgroup = h(1); EKgroup.Color = [0 0 1]
EOgroup = h(2); EOgroup.Color = [0 0.8 1]
PKgroup = h(3); PKgroup.Color = [0 .4 .2] 
POgroup = h(4); POgroup.Color = [0 .8 .2]
LKgroup = h(6); LKgroup.Color = [.8 0 .6] 
LOgroup = h(7); LOgroup.Color = [1 .4 .8] 
EOextra = h(5); EOextra.Color = [0 0.8 1]
LOextra = h(8); LOextra.Color = [1 .4 .8] 
EKextra = h(9); EKextra.Color = [0 0 1]
set(gca,'fontsize',16); 
xlabel('DIC');
ylabel('TA'); 

Answer 1

[Walter Roberson]

The reason will be obvious if you unique(SiteID)

h = gscatter(DIC, TA, strtrim(SiteID));

[Heidi Hirsh]

Thank you Walter! It turns out that some of the IDs had extra spaces!