Star formation in high-redshift quasars: excess [O II] emission in the radio-loud population

We investigate the [O?ii] emission-line properties of 18?508 quasars at z < 1.6 drawn from the Sloan Digital Sky Survey (SDSS) quasar sample. The quasar sample has been separated into 1692 radio-loud and 16?816 radio-quiet quasars (hereafter RLQs and RQQs) matched in both redshift and i'-band absolute magnitude. We use the [O?ii]?3726 + 3729 line as an indicator of star formation. Based on these measurements we find evidence that star formation activity is higher in the RLQ population. The mean equivalent widths (EW) for [O?ii] are EW([O?ii])RL = 7.80 +/- 0.30 angstrom and EW([O?ii])RQ = 4.77 +/- 0.06 angstrom for the RLQ and RQQ samples, respectively. The mean [O?ii] luminosities are log10[L([O?ii])RL/W] = 34.31 +/- 0.01 and log10[L([O?ii])RQ/W] = 34.192 +/- 0.004 for the samples of RLQs and RQQs, respectively. Finally, to overcome possible biases in the EW measurements due to the continuum emission below the [O?ii] line being contaminated by young stars in the host galaxy, we use the ratio of the [O?ii] luminosity to rest-frame i'-band luminosity; in this case, we find for the RLQs log10[L([O?ii])RL/Lopt] = -3.89 +/- 0.01 and log10[L([O?ii])RQ/Lopt] = -4.011 +/- 0.004 for RQQs. However the results depend upon the optical luminosity of the quasar. RLQs and RQQs with the same high optical luminosity log10(Lopt/W) > 38.6 tend to have the same level of [O?ii] emission. On the other hand, at lower optical luminosities log10(Lopt/W) < 38.6, there is a clear [O?ii] emission excess for the RLQs. As an additional check of our results we use the [O?iii] emission line as a tracer of the bolometric accretion luminosity, instead of the i'-band absolute magnitude, and we obtain similar results. Radio jets appear to be the main reason for the [O?ii] emission excess in the case of RLQs. In contrast, we suggest active galactic nuclei feedback ensures that the two populations acquire the same [O?ii] emission at higher optical luminosities.