Reproductive biology of two Temognatha species 
behaviour are available are, co-incidentally, closely 
allied and members of the Temognatha ( Temognatha) 
variabilis species-group, as defined herein. This 
species-group contains the following species: T. 
variabilis (Donovan, 1805) (type species of 
Temognatha), T. fortmimi (Hope, 1843), T. mitchellii 
(Hope, 1846), T. bruckii (Thomson, 1878), T. 
chalcodera (Thomson, 1878), T. regia (Blackburn, 
1892). The T. variabilis species-group is diagnosed 
by the following character combination: size 
moderately small to large (TL 19—16 mm); eyes 
strongly converging dorsally (MIDAV 10-33% of 
HW, sexually non-dimorphic intraspecifically); 
pronotum with explanate lateral margins; pronotal 
base same width as (not wider than) elytral base; 
scutellum shape scutiform (subpentagonal) to 
subcircular; elytra punctate-striate; epipleuron 
entire (non-serrate), without caudally directed 
ventral spur at level of hind-coxae, and only 
weakly expanded anterior to this level; elytron 
apex bispinose with narrow arcuate or broad 
truncate excision between spines (spine length 
reduced in some taxa); sternite 7 sexually 
dimorphic: females convexly rounded/more 
elongate, males concavely/arcuately excised 
(notched); presence on female sternite 7 of an 
arcuate subapical carina (parallel to and near apical 
margin) with dense subapical setae; female 
proctiger comprised of completely fused epiproct 
and paraprocts, without discernible suture between 
both; male proctiger comprised of incompletely 
fused epiproct and paraprocts, with visible suture 
between both; basal hind-tarsomere length 1-1,5x 
that of following hind-tarsomere; tarsal claws 
moderately to strongly curved and with weak to 
strong basal lobes (intraspecifically constant); 19- 
56 tubules per testis ( bruckii and fortnumi not 
examined for this character); "shape" of male 
genitalia; ovipositor broad (wider than long), 
dorsal valve longitudinally striate dorsally, ventral 
valve not strongly sclerotized ventrally. 
On the limited data known (this paper; pers. obs.; 
n = 3 species), members of the T. variabilis species- 
group breed in Allocasuarina spp. (contra 
Macqueen: 1948; 1964). In the discussion following 
Macqueen's (1948) observations there is mention 
(p.3), probably attributable to A.P. Dodd, that T. 
regia "breeds near the base of Casuarinas". In the 
absence of observations of reproduction in other 
members of this species-group and other 
Temognatha taxa, the phylogenetic generality of the 
behavioural characteristics discussed herein is 
indeterminate. Variation in Temognatha ovipositor 
form/structure (see Peterson, 1991: 121,123, figs 2- 
4) may be indicative of alternative mechanisms of 
particle collection/oviposition in other Temognatha 
species. However, the slight structural 
modifications of female sternite 7 within the T. 
variabilis species-group are likely to relate to sand/ 
207 
charcoal-scraping, and thus this latter habit (in 
conjunction with the unusual egg-laying behaviour) 
may be characteristic of this species-group. 
Additional observations are required to determine 
the extent and constancy of species-level 
differences in the type of particles collected. 
The functional significance of covering the egg 
with sand (aside from the confirmed function of 
aiding glueing of egg to trunks of plants), and 
possibly charcoal, is unknown. In the interests of 
stimulating further research, I propose three 
potential functions for the sand-covering, acting 
singly or in combination: protection against 
parasitism/predation of egg; protection against 
dessication of egg; thermoregulation to maximise 
rate of development of egg to first instar larva. 
Additionally, the nature/duration of this type of 
oviposition behaviour is likely to place adult 
females at less risk from predation, which would 
be increased if they had to spend extra time to 
penetrate the tough non-fissured living bark to lay 
eggs. This type of oviposition also puts the onus on 
the larvae to penetrate into the hostplant and thus 
suggests it is likely to be more energy efficient 
because of the following factors: larval mandibles 
are presumably better designed to penetrate living 
unfissured bark/timber than the "reduced" 
stigmoderine ovipositor; the hole created by first 
instar larva is smaller (only required to be width of 
larval pronotum, the widest part of larval body) 
than that required to oviposit an egg into. 
Three major types of oviposition behaviour have 
been previously described in the family 
Buprestidae: exophytic (egg layed directly into soil 
so larva, upon hatching, can feed externally on 
roots while initially protected by soil: Julodini 
(Holm, 1979: 98; Holm and Gussmann, 1992: 3), 
Julodimorpha bakeivelli (White) (Hawkeswood and 
Peterson, 1982: 242)); simple endophytic (egg layed 
directly into crevices/fire-scars in timber/bark, or 
into leaf/stem tissue so larva, upon hatching, can 
feed internally on cambium, bark or leaves/stems 
while initially protected by timber/leaf/stem 
surface: majority of buprestid genera (e.g. 
Hadlington and Gardner, 1959: 325, 326)); complex 
endophytic (egg layed on surface of leaf/stem or 
timber/bark and covered by a presumably 
protective coating of material until larva hatches 
and bores into hostplant: Agrilus obtusus Horn 
(Manley, 1977: 80, fig.2.)). 
The behaviour described in this paper generally 
conforms to the latter oviposition method, which I 
consider ecologically intermediate between the two 
former oviposition types. However, Manley's 
observations differ from mine in being apparently 
based on captive beetles, and A. obtusus adults 
utilize their own faeces to provide a mat to lay the 
egg on, and to then cover and presumably 
camouflage exposed parts of the egg. This suggests 
