In termites soldiers are maintained for their intrinsic benefit to cost ratio - Article in the Journal of Evolutionary Biology by Estelle Roux and Judith Korb of the University of Regensburg, Germany.

PLoS Biology: Everything You Always Wanted to Know about Sexes: Open-access article, discussing this argumentation:
"No known organism needs more than one mother and one father. But even this assumption is now starting to break down at the level of biological systems. In a recently discovered hybrid system within the harvester ant genus Pogonomyrmex, queens must mate with two types of males to produce both reproductive individuals and workers. These ants are the first species known which truly has more than two sexes with colonies effectively having three parents argues Joel Parker of the University of Lausanne, Switzerland."

Very low genetic variability in the Indian queenless ant Diacamma indicum - Mol Ecol, Vol 13, Issue 7, pp. 2095-2100 (Abstract)

Limited phylogeographical structure across Eurasia in two red wood ant species Formica pratensis and F. lugubris (Hymenoptera, Formicidae) - Mol Ecol, Vol 13, Issue 7, pp. 1849-1858 (Abstract)

Proceedings of the Royal Society has a pair of ant articles up in first-cite:

Path integration in desert ants, Cataglyphis: how to make a homing ant run away from home, Andel and Wehner

Worker policing by egg eating in the ponerine ant Pachycondyla inversa, D'Ettorre et al

The latest issue of Insectes Sociaux is now online (subscription only). There are too many research articles to list here, so you might as well just visit the journal homepage and see for yourself. Papers include work on Colony distribution of Cataglyphis by Dillier and Wehner, a study showing that queenlessness in Argentine ants does not affect nestmate recognition by Caldera and Holway, a study by Gyllenstrand et al on genetic differentiation in two sympatric Formica, and a technical note by King and Porter confirming that 95% ethanol is the most useful preservative for ant specimens.

Pirk et al., 10.1073/pnas.0402506101 - Online before print: Worker reproduction in queenright honey bee cololnies (Apis mellifera) is low, in contrast to queenless colonies, most likely mediated via the removal of worker-laid eggs by other workers ("policing"). This behavior ist thought to evolve because of the high queen mating frequencies, as then workers are more closely related to the sons of their mother than those of their sisters (Hymenoptera). Alternatively it is hypothesized that worker-laid eggs are less viable than queen-laid eggs. Exactly this result is shown by this article for honey bees.