We analyse the relationships between galaxy morphology, colour, environment and stellar mass using data for over 105 objects from Galaxy Zoo, the largest sample of visually classified morphologies yet compiled. We conclusively show that colour and morphology fractions are very different functions of environment. Both colour and morphology are sensitive to stellar mass. However, at fixed stellar mass, while colour is also highly sensitive to environment, morphology displays much weaker environmental trends. Only a small part of both the morphology–density and colour–density relations can be attributed to the variation in the stellar-mass function with environment.
Galaxies with high stellar masses are mostly red in all environments and irrespective of their morphology. Low stellar-mass galaxies are mostly blue in low-density environments, but mostly red in high-density environments, again irrespective of their morphology. While galaxies with early-type morphology do always have higher red fractions, this is subdominant compared to the dependence of red fraction on stellar mass and environment. The colour–density relation is primarily driven by variations in colour fractions at fixed morphology, in particular the fraction of spiral galaxies that have red colours, and especially at low stellar masses. We demonstrate that our red spirals primarily include galaxies with true spiral morphology, and that they constitute an additional population to the S0 galaxies considered by previous studies. We clearly show there is an environmental dependence for colour beyond that for morphology. The environmental transformation of galaxies from blue to red must occur on significantly shorter time-scales than the transformation from spiral to early-type.
We also present many of our results as functions of the distance to the nearest galaxy group. This confirms that the environmental trends we present are not specific to the manner in which environment is quantified, but nevertheless provides plain evidence for an environmental process at work in groups. However, the properties of group members show little dependence on the total mass of the group they inhabit, at least for group masses .
Before using the Galaxy Zoo morphologies to produce the above results, we first quantify a luminosity-, size- and redshift-dependent classification bias that affects this data set, and probably most other studies of galaxy population morphology. A correction for this bias is derived and applied to produce a sample of galaxies with reliable morphological-type likelihoods, on which we base our analysis.