Batley & Houston: Australian bee genus Trichocolletes 
43 
There was a small discrepancy in the labelling of 
the holotype of T. nigroclypeatus. A male and a female 
specimen were found carrying red Allotype and Type labels, 
respectively, together with identification labels in Rayment’s 
hand. Collection information on the labels corresponded 
with details given in Rayment’s original descriptions 
(Rayment, 1929, 1931) except that the date attached to the 
holotype (Oct. 1929) did not match that given in the original 
description (12/10/27). It is possible that the holotype was 
(re)labelled by Rayment when he designated the allotype. 
Both specimens are T. venustus. 
Rayment (1931) refers to specimens of T. venustus from 
Western Australia when discussing bees in the collection 
of the Department of Agriculture of Western Australia. The 
only specimen of T. venustus found in that collection was 
from Black Rock, Victoria and no specimens bearing the 
collection data referred to by Rayment were found in that 
or any other collection. It is unlikely that the species occurs 
in Western Australia. 
Distribution. Southeastern Australia including Tasmania 
(BBS, FLB, KAN, NAN, NET, NNC, NSS, SB, SCP, SEH, 
SEQ, TSE, VM) (Fig. 119). 
Discussion 
Subgenera. Most of the interspecific differences between 
females are small and matters of degree, while male sexual 
characteristics such as modified legs may be an unreliable 
indicator of relatedness. The female characters used to define 
the subgenus Callocolletes (Michener, 1965), and most of the 
extraordinary modifications of the male, occur individually 
in some of the newly described species. For reasons given 
in the remarks for T. pulcherrimus, we have chosen not to 
recognize this subgenus. 
A group of thirteen species have hairy eyes, females 
with simple tarsal claws (minute inner rami in 2 cases) and 
males with similarly shaped S7. The remaining species show 
considerably greater variation in the shape of the male S7 and 
it may be that the thirteen species represent a relatively recent 
radiation. Their designation as a subgenus would, however, 
leave the remaining 27 species as a paraphyletic group and 
one that included three hairy-eyed species. 
None of the species known to Michener (1965) had 
indistinct metasomal bands and hairy eyes, so he proposed 
a species group based on the absence of metasomal bands. 
The material now available contains one species with hairy 
eyes and indistinct metasomal bands and others that display 
intraspecific variation in the visibility of the metasomal 
bands. 
The species pairs T. chrysostomus / T. sericeus, T. 
leucogenys n.sp. / T. albigenae n.sp., T. brunilabrum n.sp. 
/ T. micans n.sp. and possibly T. lacaris n.sp. / T. aeratus 
n.sp. appear to be examples of species that have diverged on 
either side of the continent. It is less obvious why the rather 
similar Western Australian species T. rufibasis and T. soror 
n.sp. should have separated. 
As the similarities that unite the genus greatly outweigh 
differences between any possible species groups it was 
decided that no subgeneric division is appropriate. 
Flower visitation. Information about flower visitation 
(including that in Houston [2000]) was present on the labels 
of 53% of the specimens examined in this study. All but six 
species were recorded as visiting pea flowers (Fabaceae: 
Faboideae) and the number of recorded visits to other flower 
groups by these 34 species was 5% of all visits. This pattern is 
consistent with the idea that the bees are broadly oligolectic, 
but occasionally visit other flowers to obtain nectar, but 
no examination of pollen loads was performed. The other 
flower families visited were Aizoaceae, Amaranthaceae, 
Apocynaceae, Caesalpiniodeae, Dilleniaceae, Ericaceae, 
Goodeniaceae, Lamiaceae, Myoporaceae, Myrtaceae, 
Polygalaceae and Rutaceae. 
Only for two of the species, T. centralis n.sp. and T. soror 
n.sp., did visits to flowers other than peas exceed 15% of all 
records. Trichocolletes centralis n.sp., with 25% non-pea 
visits, has a wide distribution across arid areas, which might 
account for an increased dependence on alternative nectar 
sources. The number of records for T. soror n.sp. (13 versus 
85 for T. centralis n.sp.) is fairly small and the 23% non-pea 
visits may be statistical variation. 
Host flowers were not recorded for the two known 
specimens of T. avialis n.sp., but the remaining five species 
may specialize in visiting other plant types. This is certainly 
true for T. eremophilae and T. multipectinatus, which 
have been found only on Eremophila species and have an 
elongated proboscis that would help them obtain nectar from 
these flowers (Houston, 1990). Whether they visit a limited 
number of flowers in order to obtain nectar (oligotropic) or 
pollen (oligolectic) or both, remains to be determined. The 
species T. hackeri has to date been collected in rainforest 
areas on the Eastern coast of Australia. Flower visits were 
recorded on only three occasions, each involving a different 
plant family, Apocyanaceae, Dilleniaceae and Myrtaceae. 
On one of those occasions, pollen collection from Hibbertia 
scandens was observed (MB pers. obs.). Trichocolletes 
macrognathus n.sp. has been observed visiting five different 
plant genera in three plant fa mili es or subfamilies, Fabaceae- 
Caesalpinioideae, Lamiaceae and Myrtaceae, but not pea 
flowers. There is no obvious relationship between the flowers 
visited and the unusual elongation of the malar area in this 
species. The final species, T. luteorufus n.sp., has been 
collected on only two occasions. 
Rayment (1935) reported that T. serotinus n.sp. and T. 
venustus nest in the ground and gave brief descriptions of 
the nests. Three other species have been seen entering nests 
in the ground: T. orientalis n.sp. at Black Mountain, ACT 
(MB pers. obs.); T. dowerinensis at Boorabbin Rock, WA 
and T. macrognathus n.sp. at Wannoo, WA (TFH pers. obs.). 
Acknowledgments. The authors are greatly indebted to Mr 
David Notton (BMNH), Dr John Asher (AMNH) and Mr James 
Hogan (OUM) for their expert assistance with type material 
and wish to thank the following people who provided facilities 
for the examination of specimens or who loaned specimens for 
examination: Mr C. Brumley (WAA), Ms N. Fisher (ANIC), Mr 
B. Hanich (WAM), Dr P. Hudson (SAM), Mr D. Notton (BMNH), 
Dr K. L. Walker, Dr S. Winterton (UQIC) and Ms S. Wright (QM) 
and Dr K. Hogendoom who provided helpful comments on the 
manuscript. One of us (MB) would like to thank Dr D. Britton and 
Dr D. K. McAlpine for helpful discussion and guidance. 
