The systematic study of the environment of BL Lacs is relatively recent. The work of Prestage and Peacock (1988), discussed in Sec. 5.3, is usually quoted (wrongly) as evidence that BL Lacs avoid rich environments; it actually shows that (within large uncertainties) the spatial correlation amplitude of (4) BL Lacs is consistent with that of FR I radio galaxies.
Fields of low-redshift BL Lac objects that have been studied intensively tend to show statistically significant excesses of galaxies, compatible with clusters of galaxies of Abell richness class 0 and 1 (Falomo et al. 1993; Pesce et al. 1994; Fried et al. 1993; Wurtz et al. 1993; Smith et al. 1995), in good agreement with the environments of FR I type radio galaxies (Hill and Lilly 1991). The environment of PKS 0548-322 (Fig. 12) is clearly rich in galaxies, including two companion elliptical galaxies within ~ 40 kpc projected distance (if at the redshift of the BL Lac, z = 0.069) and numerous other galaxies with brightnesses appropriate for a cluster at z = 0.069. Some nearby BL Lacs lie in somewhat poorer clusters, as do some FR Is (Wurtz et al. 1993; Laurent-Muehleisen et al. 1993). In short, BL Lacs and FR Is do appear to have similar environments.
Reversing the question leads to an unexplained discrepancy. A spectroscopic survey of 193 radio sources found in Abell clusters revealed 186 FR I galaxies but no BL Lac objects, when according to beaming models roughly 8 would have been expected in the observed luminosity range, a discrepancy significant at the > 99% level (Owen et al. 1995). That is, this unbiased cluster-selected survey for BL Lac objects suggests they occur preferentially outside of normal FR I environments. We note that the FR I luminosities observed by Owen et al. (1995), L1.4 < 1026 W Hz-1, and the observed BL Lac radio luminosity function, for which L5 > 1025 W Hz-1 (Fig. 17; Urry et al. 1991a), overlap by less than one decade. In this interval there are only 18 FR Is and the absence of BL Lacs is only marginally significant (95% confidence; Owen et al. 1995). Furthermore, the conflict in number density is actually between the observed BL Lac and FR I luminosity functions rather than with the predictions of the beaming model. It is possible that since the redshift limit of the cluster sample is low, z < 0.09, recognition effects (Browne and Marchã 1993) could be important.
At higher redshifts (z ~ 1) there is no indication of a density enhancement near the BL Lacs, as expected because cluster galaxies at these redshifts would be well below the typical detection limits (Fried et al. 1993). It is important to study blazar environments at high redshift because it is there that quasar environments appear to undergo significant evolution. At z ~ 0.6, radio quasars are often found in clusters of galaxies while locally they are in regions of lower galaxy density (Yee and Ellingson 1993). Furthermore, if there were a connection between the more luminous BL Lac objects and radio quasars (see Sec. 7.2), it would only be apparent at higher redshift.