Star formation in Perseus

Aims. Molecular outflows provide an alternative method of identifying protostellar cores, complementary to recent mid-infrared studies. Continuing our studies of Perseus, we investigate whether all Spitzer-identified protostars, and particularly those with low luminosities, drive outflows and if any new protostellar cores (perhaps harbouring low-mass sources) can be identified via their outflows alone.Methods. We have used the heterodyne array receiver HARP on JCMT to make deep 12CO 3–2 maps of submm cores in Perseus, extending and deepening our earlier study with RB and bringing the total number of SCUBA cores studied up to 83. Our survey includes 23/25 of the Spitzer low-luminosity objects believed to be embedded protostars, including three VeLLOs.Results. All but one of the cores identified as harbouring embedded YSOs have outflows, confirming outflow detections as a good method for identifying protostars. We detect outflows from 20 Spitzer low-luminosity objects. We do not conclusively detect any outflows from IR-quiet cores, though confusion in clustered regions such as NGC1333 makes it impossible to identify all the individual driving sources. This similarity in detection rates despite the difference in search methods and detection limits suggests either that the sample of protostars in Perseus is now complete or that the existence of an outflow contributes to the Spitzer detectability, perhaps through the contribution of shocked H2 emission in the IRAC bands. For five of the low-luminosity sources (including two previously believed to be embedded), there is no protostellar envelope detected at 350 μm and the Spitzer emission is entirely due to shocks. Additionally, we detect the outflow from IRAS 03282+3035 at 850 μm with SCUBA with 20–30% of the submm flux due to CO line contamination in the continuum passband.