Dense clumps in cool molecular clouds are thought to be the "cradles" of star formation, but finding molecules at low metallicities has been exceedingly difficult. In dwarf galaxies with nebular oxygen abundances < 20% solar, despite vigorous on-going star formation, there seems to be very little H2, at least as inferred from 12CO(1-0) emission.

Although we expect some dependence of CO content on metallicity, this gross deficit is rather surprising. Generally, the gas component of the interstellar medium (ISM) and star formation are intimately linked; the star-formation rate (SFR) in galaxies is found to depend on gas surface density, Σ(gas), through a power-law relation known as the Schmidt law (more recently Kennicutt-Schmidt, KS law). This correlation holds over a wide range of environments from quiescent spiral disks with low SFR surface densities (Σ(SSFR)) to circumnuclear starbursts and ultraluminous infrared galaxies (ULIRGs) where the SFRs are much higher Interestingly, however, metal-poor dwarf galaxies which are forming stars at high Σ(SSFR) do not follow the "standard'' KS relation. While most dwarfs form stars relatively slowly, some of them form stars almost as violently as ULIRGs. However, unlike ULIRGs, they do this in an ISM which is apparently almost entirely atomic ; how such galaxies form stars is the subject of this ISSI consortium.

We propose to explore alternatives to CO as a tracer of molecules and star formation at low metallicity; this will be done both from theoretical and observational points of view, and will include an assessment of how molecular content depends on metal abundance, dust grain properties, interstellar radiation field, gas density, and other properties of the ISM.