Species accumulation curve = accumulation rates of new species over sampled areas, so it is hard for you to use this definition because you are using non-plot samples. But you could plot number of species on the y axis versus cumulative number of samples on the x axis, or the natural logarithms, etc. 

See: 

The Species-Accumulation Curve and Estimation of Species Richness

Karl I. Ugland; John S. Gray; Kari E. Ellingsen

The Journal of Animal Ecology, Vol. 72, No. 5. (Sep., 2003), pp. 888–897. 

I hope this helps :)

Numerosity of a single genus in different herbaria may indicate a trend but it is usually met by plot scores made at purpose in places wich were not familiar to the genus Met inSchmidt et al. (Diversity Distrib.)(2005)11, 509–516

if you have enough samples and their localities (site names, gps points or so) you could plot the herbarium individuals on a map, then you can "sample" in different (increasing) areas the plotted species (samples). Nevertheless, this method assumes that samples from all available species in the herbarium are included (at least all the species reported for a particular region). Maybe it is as demandant as field work...

I was using the concept that species accumulation curve uses sampling effort. For bats, each night sampling is also used instead of area or net, so I would like to try an alternatiive with herbaria specimens.

Thank you all for your ideas, I'll consider them all.