4 
D.P. GORDON AND P.D. TAYLOR 
Percentage occurrences of different bryozoan colony morphologies in different settings based on numbers of species, not numbers of colonies or 
Table 1 
colony fragments (* from Gordon (1987); + from Lee er al. (1997)). 
Red Bluff Recent, shell 
Tuff gravel* 
MORPHOLOGY 
Shell-boring 2 0 
Total encrusting 50 72 
runner 7 6 
patch 38 61 
mound 5 5) 
Total fixed-erect 29 12 
tree 22 8 
rod 7 0 
lamellar 0 4 
Total flexible-erect 17 15 
tree U7/ 15 
rod 0 0 
lamellar 0 0 
Total dwarf-rooted 2, 0 
conical 0 0 
lobate 2 0 
pedunculate 0 0 
Free-living 0 <1 
Red Bluff Tuff at Pukekio (ca. 50% of all species), this proportion 
does not conform to the generalised pattern of modern bryozoan 
colonial morphologies on hard substrata around New Zealand. 
Gordon (1987) has shown that from a range of modern habitats, 
excluding the deep-sea and regardless of bryozoan diversity, the 
proportion of bryozoan species as planar or mounded encrusters of 
hard substrata is fairly consistent, ranging from 67—76% of the total 
bryofauna. This is a taxonomic measure, independent of actual 
numbers of colonies in each category, which are very difficult to 
obtain. The lower proportion of encrusting species in the Red Bluff 
Tuff samples could be correlated with the relative paucity of hard 
substrata (e.g., basaltic pebbles and cobbles, shell and other bio- 
genic carbonate like stylasterid colonies). Alternatively, it may be a 
function of poor sampling. For Cenozoic cheilostome faunas in 
tropical America, Cheetham & Jackson (1998) have shown that 
encrusting species are far more numerous than erect or free-living 
species but have fewer occurrences per species and much less 
abundance per species. They are consequently the least well-sam- 
pled, and this may also be the case for the Red Bluff Tuff. The 
relative numbers of encrusting species at the Pukekio site is even less 
than that (59%) at Alma, Oamaru, where there is a notable mobile 
rockground dominated by bryozoans (Lee et al., 1997). At the Alma 
site (and also at nearby Fortification Road, Kakanui) it is possible to 
count actual numbers of colonies because of the large numbers of 
lithoclasts, available for attachment in life, and the preservation of 
the bases of fixed-erect colonies. (Numbers of colonies of species 
that are articulated in life are impossible to determine owing to 
fragmentation.) The numbers of fixed-erect colony bases (mostly 
Recent, shell/ Eocene, mobile _Terrigenous, 
volcanic gravel* rockground+ carbonate mud 
(deep sea)* 
<1 0 0 
a 59 10 
15 7 0 
59 50 8 
3 2) 2} 
8 26 10 
6 26 8 
<1 0 0 
2 0 2 
13 14 39 
13 13 31 
0 1 2) 
0 0 6 
2 0 41 
2 0 33 
0 0 4 
0 0 4 
0 1 0 
indeterminable to species but representing only 8 out of 301 bryozoan 
colonies counted) was significantly less than the numbers of fixed- 
erect species (20) represented in the Alma fauna and Fortification 
Road bryofauna from broken fragments found in the sieved sedi- 
ment matrix between the clasts. Smith (1995) has summarised the 
range of taphonomic and other problems that confound 
palaeoecological reconstruction of bryozoan faunas, and these apply 
to the Red Bluff Tuff and Oamaru sites. One feature of the Red Bluff 
Tuff fauna may be noted, however — the unusually high number of 
cellariid species, including four species of Cellaria. Even account- 
ing for taphonomic filters, this preponderance of Cellaria species at 
a single locality is truly noteworthy compared with later Cenozoic 
and Recent bryofaunas in New Zealand. Generally, the basally 
rooted, often articulated, erect colonies of cellariids are adapted to 
live in significant current speeds, as well as being able to tolerate 
moderate amounts of fine sedimentation (Lagaaiy & Gautier 1965). 
SYSTEMATIC PALAEONTOLOGY 
Specimen repositories and abbreviations: NHM, The Natural His- 
tory Museum, London; IGNS, Institute of Geological & Nuclear 
Sciences (formerly New Zealand Geological Survey), Hutt City, 
New Zealand. 
The species described were studied by scanning electron 
microscopy (SEM), using type and other specimens. Sorted material 
was soaked in hypochlorite solution overnight then washed in water 
