Filtering a set of points in a "ppp" object by distance using marks

Dear R-Sig-Geo Members,

I'd like to find any way to filtering a set of points in a "ppp" object by minimum distance just only between different marks. In my example:

#Package
library(spatstat)

#Point process example - ants
data(ants)
ants.ppp<-ppp(x=ants$x,y=ants$y,marks=rep("antNests",length(ants$x)),window=Window(ants))


# Create a artificial point pattern - termites
termites <- rpoispp(0.0005, win=Window(ants))
termites.ppp<-ppp(x=termites$x,y=termites$y,marks=rep("termiNests",length(termites$x)),window=Window(ants))


#Join ants.ppp and termites.ppp
insects.ppp<-superimpose(ants.ppp,termites.ppp)


#If I try to use subset function:

subset(insects.ppp, pairdist(insects.ppp) > 20 & marks=="termiNests")

#Marked planar point pattern: 223 points #marks are of storage type
?character? #window: polygonal boundary #enclosing rectangle: [-25, 803]
x [-49, 717] units (one unit = 0.5 feet) #Warning message: #In ppp(X[,
1], X[, 2], window = win, marks = marx, check = check) : # 70751 out of
70974 points had NA or NaN coordinate values, and were discarded

Not the desirable result yet, because I'd like to calculate just only the > 20 "termiNests" to "antNests" marks and not the "termiNests" with "termiNests" too.

Please any ideas?

Thanks in advanced,