Reprint: Patterns of beetle species diversity in Castanopsis 
acuminatissima (Fagaceae) trees studied vjith canopy 
fogging in mid-montane New Guinea rainforest. 
Canopy Arthropods, 1997, pp 224-236 
Canopy Arthropods 
Edited by 
N.E. Stork 
Head of Biodiversity Division in the Department of Entomology, The 
Natural History Museum, London, UK; Noio Director of the Cooperative 
Research Centre for Tropical Rainforest Ecology and Management, James 
Cook University, Cairns, Australia 
J. Adis 
Senior Scientist, Tropical Ecology Working Group, Max-Planck-Institute for 
Limnology, Plan, Germany 
and 
R. K. Didham 
Commonwealth Scholar, The Natural History Museum, London and the 
NERC Centre for Population Biology, Ascot, UK 
CHAPMAN & HALL 
London ? Weinheim ? New Yori< ? Tol<yo ? Melbourne ? Madras 
11 
Patterns of beetle species 
diversity in Castanopsis 
acuminatissima (Fagaceae) trees 
studied with canopy fogging 
in mid-montane New Guinea 
rainforest 
A. Allison, GA. Samuelson and S.E. Miller 
ABSTRACT 
Canopy fogging with a pyrethnxm-based insecticide was used to study the structure 
and diversity of beetle communities in eight Castanopsis acuminatissima (Bl.) A.DC. 
(Fagaceae) trees at elevations of 1200-1400 m in the Wau Valley, Papua New Guinea. 
Arthropods were collected in 1-m- trays suspended beneath the trees, sorted to Order 
and all beetles were identified to morphospecies. The eight trees yielded 3977 indi- 
vidual beetles representing 418 morphospecies in 53 families. The mean number of 
beetle morphospecies per tree was 114 (range 82-155). Of these morphospecies, 199 
(47.6%) were represented by single individuals, and only 62 (14.8%) were represented 
by 10 or more individuals. Approximately 25% of the spedes on each tree were tree- 
specific, although those species represented only ca. 5-10% of the individuals. The 
similarity of the beetle fauna between trees, as measured by the Jaccard coefficient, 
ranged from 12 to 31%. This study is part of a much larger canopy fogging project in 
which 51 trees at study sites of 500 m, 1200-1400 m and 2100-2250 m have been fogged. 
Preliminary analysis of these samples indicates that diversity is similar at sites at 
500 m and 1200-1400 m, but higher at 2100-2250 m. 
Canopy Arthropods. Edited by N.E. Stork, J. Adis and R.K. Didham. Published in 1997 by 
Chapman & Hall, London. ISBN 0 412 74900 9 
study sites 225 
INTRODUCTION 
Canopy fogging techniques were originally developed to control pests. 
Gagn? was one of the first to recognize the applicabiHty of these tech- 
niques to study the diversity of insects in native forest hrees (Gagn?, 
1979). This led to further siniilar studies and in 1980 Erwin reported on 
his canopy studies in Panama (Erwin and Scott, 1980). The debate that 
this work sparked on the magnitude of global species richness (Erwin, 
1982; Stork, 1988,1993; May, 1990; Gaston, 1991, Hodkinson and Casson, 
1991; Hammond, 1992; 1994; Hodkinson, 1992) prompted us in 1987 to 
commence a canopy fogging study of raii\forest trees along an altitu- 
dinal transect from 500-2200 m in New Guinea. 
In order to put patterns of diversity into clear perspective it is neces- 
sary to meastire both within-habitat diversity (alpha-diversity) and 
between-habitat diversity, also known as turnover (beta-diversity) 
(Whittaker, 1975). To achieve this our experimental design emphasized 
a single species of tree, Castanopsis acuminatissima (Fagaceae). We 
fogged this species, together with related {Lithocarpus and Nothofagus) 
ancf unrelated genera (of Bxirseraceae, Clusiaceae, Dipterocarpaceae, 
Elaeocarpaceae, Grossiilariaceae, Juglandaceae, Lauraceae, Phyllo- 
cladaceae, Rosaceae, Sapindaceae) at study sites located at 500 m 
(Oomsis), 1200-1400 m (Wau VaUey) and 2100-2200 m (Biaru Road). We 
have now fogged a total of 51 trees and obtained more than 45 000 beetle 
specimens. The preliminary analysis of beetle samples from the first eight 
trees has been completed (two trees of C. aciiminatissima from each of 
the three study sites and two trees of Lithocarpus celebicus at 500 m) 
(Allison et al, 1993b). In this paper we report on the preliminary analysis 
of iiisect samples from eight C. acuminatissima trees (including two trees 
from the ear?er analysis) at the same elevation in the Wau Valley. 
STUDY SITES 
Field work was based at the Wau Ecology Institute in Papua New 
Guinea. Part of the field work was conducted in remnant Castanopsis 
forest approximately 1 km east of Wau Ecology Institute on the road to 
the former New Guinea Goldfields mine at Nami (1250 m). The 
remaining work was conducted on the lower slopes of Mt Missim on a 
ridge between Poverty and Sandy Creeks (1250-1350 m). 
The Wau Valley generally receives ca. 1900 mm of rainfall annually. 
There is a slight seasonal pattern with ca. 60% of the armual total occur- 
ring from November to April vmder the influence of the north-west 
monsoons. The period from May to October, when the south-west 
tiradewinds form the dominant weather pattern, is sometimes very- 
dry (<25 mm of rainfall per montii). The mean armual temperature at 
226       Beetle species diversity in Castanopsis acuminatissima 
Table 11.1  Details of total number of species and individuals and mean body 
length (? S.D.) of beetles collected from a total of 126 l-m^ trays suspended 
beneath trees 
Tree Date Total Total Mean (S.D.) body Total no. of 
species individuals length (mm) trays used 
3* 15-10-87 82 335 2.66 (1.62) 13 
4 04-11-87 86 300 2.51 (1.74) 13 
9* 07-08-88 99 404 2.84 (1.74) 13 
10 18-08-88 155 800 2.89 (2.46) 17 
11 18-08-88 154 547 2.55 (1.63) 16 
12 19-08-88 129 696 2.30 (1.35) 15 
13 19-08-88 132 493 2.40 (1.65) 16 
14 20-08-88 106 402 2.39 (1.28) 23 
? Trees 3 and 9 are the same tree sampled 10 months apart. 
1230 m (Wau Ecology Institute) in the Wau Valley is 22?C with little 
annual variation (Gressitt and Nadkami, 1978). The vegetation in the 
area is generally classified as mid-montane rainforest (van Valkenburg 
and Ketiier, 1994). 
Study trees 
We fogged eight trees of Castanopsis acuminatissima over a 2-year period 
(Table 11.1). C. acuminatissima occurs throughout New Gviinea from 
nearly sea level to at least 2200 m and, like many tiropical fagaceous 
trees, is found mainly on ridge crests (Soepadmo, 1972; Shreimaim, 1983; 
van Valkenburg and Ketner, 1994). It is extiremely abimdant in the Wau 
Valley and occurs in both primary and secondary forest. In some areas 
it appears as a monospecific dominant. It is the only species of Castanopsis 
in New Guinea, although there are a large nimiber of related species to 
the west in Indonesia and south-east Asia, including a number of species 
that occupy subtropical and temperate environments (Soepadmo, 1972). 
Our original sampling design called for 10 tarays per tiree, but to guard 
against the possibility of samples being lost, we used at least 13 tirays 
for each tree and obtained a total of 126 tray samples (Table 11.1). 
METHODS 
Our methods are described in full in .Allison et al. (1993b) and are 
summarized here. We selected tir?es for fogging that seemed healthy, 
had no epiphytes or imderstorey, did not have flowers or fruit, and had 
Results 227 
a canopy that did not intermingle with adjacent trees. Each of the trees 
was 20-25 m tall and had a canopy voliune of ca. 3000-4000 m^. 
Trees were fogged at 06:00 h (when the air was almost still) for 15 
minutes using Pyranone? mixed with kerosene to form a solution of 5% 
active ingredient that was delivered by a Curtis Dynafog model 2610 
fogger with the fog density dial set on '5'. The fogger was operated 
manually at canopy level from an adjacent tree. No living insects were 
found on branches subsequently removed from fogged trees. 
Insect samples were collected in shallow plastic trays (126 in total), 
each 1 m^ in area, svispended approximately 1 m above the groxmd 
beneath each tree. Each tray was numbered individually and had an 
8-cm diameter aliiminium furmel inserted in the centre, flush with the 
tray surface. A 125-ml Nalgene? bottle approximately half full of 70% 
ethanol was attached to the base of each funnel. Paint brushes (ca. 5 cm 
wide) were used to sweep the ii\sect samples into the bottles for preser- 
vation. Trays were left in place for 2 hours after fogging although oui 
impression was that most of the insects dropped from the tree within 
20-30 minutes of fogging. Each sample bottle was labelled with tray and 
tree number. To ensure good preservation, the alcohol (70% ethanol) 
was changed at least four times in the month following collection. 
In the laboratory all beetle specimens were removed, mounted and 
identified to morphospecies. Where possible our morphospecies concepts 
have been verified by specialists in each family. All information was 
entered into a computer database and reconciled with labelled specimens 
to ensxire accuracy. All beetles and tree voucher specimens are deposited 
in the Bishop Museum, Hawaii. Beeties were measured from the 
anterior-most part of the head to the apex of the elytra, except for species 
with hidden heads for which the anterior edge of the pronotiun was 
iised'and weevils for which the anterior edge of the eye was used. For 
species with more than a few individuals, several individuals, deter- 
mined by inspection to be of average size, were measured. Mass was 
calcidated from the equation derived by Schoener (1980) for Costa Rican 
rainforest beetles. 
RESULTS 
The eight tir?es yielded 3977 individual beeties representing 418 
morphospecies in 53 families (Table 11.1). The total number of species 
on each of the trees ranged from 82 to 155. Foiir families had more than 
25 species each: Chrysomelidae (39), Staphylinidae (39), Curculionidae 
(37) and Cocdnellidae (29), together comprising 34% of total species 
richness. They each had more than 200 individuals and together were 
represented by 1885 individuals (47% of the total). The only other family 
with  more   than  200   individuals  was   Attelabidae  which  had  431 
228      Beetle species diversity in Castanopsis acuininatissima 
individuals (but only three species), more than Curculionidae (206), 
Coccinellidae (229), but less than Chrysomelidae (601) and Staphylinidae 
(849). These five families comprised 58% of total individuals. Eleven 
families were represented by single species (number of individuals in 
parentheses): Lampyridae (1), Merophysiidae (1), Rhizophagidae (1), 
Sphindidae (1), Hydrophilidae (2), Trogositidae (2), Alleciilidae (3), 
Bostrichidae (4), Endomychidae (6), Eucnemidae (6) and Myceto- 
phagidae (114). 
There was no correlation between the number of collecting trays used 
beneath a tree and the niimber of individual beetles recorded from that 
tree (simple linear regression,. R^ <0.04, n = 8). In other words, the 
variance in numbers of individuals per tree masked any effect from 
differing numbers of trays. Therefore, to simplify analysis we simply 
report on the total insect sample recorded from each tree. 
Slightly fewer than half the species (199, 47.6%) were represented by 
single individuals ('singletons') and 19.9% (83) were represented by two 
individuals, with declining numbers represented by three or more indi- 
viduals. Only 62 species (14.8%) were represented by 10 or more 
individuals (Figure 11.1). More than half the beetle species (55.3%) were 
found on only one tree and fewer than one-third were found on three 
or more trees (Table 11.2). 
A separate analysis was performed on chrysomelids (herbivores) and 
staphylinids (predominantly predators; fovir species of omaliines, 
thought to be poUen-feeders, were excluded from this analysis). Patterns 
were similar to the overall trend described above, except that singletons 
comprised only 33.3% of chrysomelids (13 of 39 species) and 51.4% of 
staphylinids (18 of 35 species) (Figure 11.1). 
Table 11,2  Distribution of beetle spedes on eight Castanopsis 
acuminatissima trees from ca. 1200-1400 m elevation in the 
Wau Valley, Papua New Guinea 
Total no. of trees on No. (%) of 
which beetle species beetle species 
were recorded 
1 232 (55.3) 
2 70 (16.8) 
3 41 (9.8) 
4 30 (7.2) 
5 14 (3.4) 
6 7(1.7) 
7 13 (3.1) 
8 12 (2.9) 
Results 229 
200 
180 
160 
140 
S 120 
o 
s-100 
"o 
d 80 
l^yXA Slaphylinidae (n = 35) 
I        I Chrysomelidae (n = 39) 
4        5        6        7        8 
No. of individuals 
10      11 + 
Figtire 11.1   Comparison of numbers of individuals per species for all eight 
trees at 1200-1400 m. Categories are all beetles, Staphylinidae (except 
Omaliinae) and Chrysomelidae. 
Although there was a large niitnber of singleton species, they 
comprised only 5.0% of total individuals. The percentage of singletons 
on individual trees ranged from 10.1 to 17.2% (Table 11.3). These higher 
figures include species occurring as singletons on individual trees. 
However, about half these species occur on more than one tree and 
therefore are not treated as singletons in the overall sample (Table 11.2). 
The species diversity of the tree samples as measured by the Q-statistic 
ranged from 32.31 to 68.72 (Table 11.3). (Q is used instead of a because 
the underlying species-abimdance distribution did not follow the log- 
series; the Q-statistic is a non-parametric distribution based on the 
inter-quart?e slope of the cumulative species abundance curve; 
Magurran, 1988.) Higher meastires of diversity reflect lower total 
numbers of individuals in the sample. 
The siir?larity in the bee?e faimas of the various trees, as measxired 
by the Jaccard coefficient, Cj (Magurran, 1988), ranged from 0.12 to 0.37 
(Table 11.4). Trees 3 and 9, the same individual tree fogged about 10 
months apart, had a Jaccard coefficient of 0.30, similar to the 0.31 for 
trees 3 and 4 (fogged 20 days apart and separated by a distance of ca. 
50 m). The greatest similarity between samples, and the or?y samples 
230       Beetle species diversity in Castanopsis acummatissima 
7- 
6 
CO 
?g 
14 
_? 
2 3 
o 
O) 
g 2 
A 
A 
A 
*?A 
A              A 
i   A^ A 
A 
A 
A 
A A A\i A\^ 
A 
A 
i 
A A*   ^    if*^    A   AA    i*     *"        1      AA AAAJ      AAA        lAAAAAA           A 
AA AA     A   AUAAA A                 f.  i-U.   A                A 
1   >      1       ill      4   l?     1   1     1          AAA    i?            A                A 
A 
A 
A 
-A  
AA AAAA A    MA AAiUA A AA       A   A AA AA AAA 
' "'   111? l?? ?    fl '* ?A?AA 
-3 -1 0 
Log mass (mg) 
Figure 11.2   Double logarithmic plot of body mass versus abtmdance for 
beetles from eight trees at 1200-1400 m. Body mass was calculated from 
length using the formula for Costa Rican rainforest beetles derived by 
Schoener (1980). 
with a Jaccard coefficient above 0.31, were from trees 10 and 11. These 
trees were ca. 15 m apart and were fogged at the same time. 
Mean length of beetle species in the overall sample was 2.86 mm and 
ranged from 0.50 to 22.0 mm (SD = 2.23). Mean length of beetles in each 
of the tree samples ranged from 2.30 to 2.89 mm (Table 11.1). In a double 
logarithmic plot of abundance-body mass, the data points fall within an 
approximately triangular-shaped zone, as observed for many other 
assemblages (Gaston et ai, 1993a,b) (Figure 11.2). Blackburn et al. (1990) 
examined similar samples from Bnonei (Morse et al., 1988) and concluded 
that this pattern was caused by a preponderance of middle-sized species. 
However, it is not clear whether the observed patterns of size and abxm- 
dance result from biological processes, or from statistical or sampling 
properties common to distributiorts of any collection of objects subject 
to optimal design constraints (see also Gaston et al, 1993b). 
DISCUSSION 
Our results are similar to those obtained from other canopy fogging 
studies (Erwin and Scott, 1980; Erwin, 1983; Stork and Brendell, 1990; 
Stork, 1991; Allison et al, 1993b). A prominent feature common to these 
studies is that there is a strong relationship between the number of indi- 
viduals and the number of species, primarily because a high percentage 
(47.6% in our study) of the species are singletons. Oxir figure is lower than 
Discussion 231 
Table 11.3   Details of number and percentage of beetle species represented by- 
single individuals ('singletons') 
Tree No. Total singletons Total individuals Singletons (%) Q'Statistic 
3 51 336 15.2 32.31 
4 48 300 16.0 38.23 
9 44 404 10.9 38.23 
10 85 800 10.6 68.72 
11 58 547 10.6 46.89 
12 70 696 10.1 45.81 
13 78 493 15.8 58.26 
14 69 402 17.2 42.78 
Table 11.4  Similarity of the beetle favmas of eight Castanopsis acumina- 
tissima trees in the Wau Valley, Papua New Guinea using the Jaccard 
coefficient 
Tree 4 Tree 9 Tree 10 Tree 11 Tree 12 Tree 13 Tree 14 
Tree 3 0.31 0.30 0.19 0.22 0.19 0.19 0.16 
Tree 4 0.23 0.19 0.17 0.20 0.14 0.12 
Tree 9 0.22 0.23 0.19 0.18 0.18 
Tree 10 0.37 0.29 0.30 0.26 
Tree 11 0.29 0.32 0.32 
Tree 12 0.32 0.32 
Tree 13 0.29 
that reported by Morse et al. (1988) from Brunei (57.8%), but higher than 
that reported by Basset and Kitching (1991) from Australia (35.7%). The 
reasons for this variation are ^mdoubtedly complex and probably reflect 
differences in the insect communities as well as differences is sampling 
methodology and effort. Oxir results and those of Basset and Kitching 
(1991) are based on single tree species. Morse et al. (1988) sampled single 
individuals of 10 species of lowland trees and their samples may include 
greater beta-diversity. In any case, it is the generally high percentage of 
singleton species in canopy fogging samples (assiimed to include the total 
insect fauna of the tree canopy) that led to the extremely high estimates 
of the diversity of the rainforest canopy insect fauna advanced by Erwin 
and others (Erwin, 1982,1988; Stork, 1988). However, in our study, of the 
species recorded from individual trees as singletons, approximately half 
occurred on other trees, sometimes in reasonably large numbers. 
232       Beetle species diversity in Castanopsis acuminatissima 
The ecology of singleton species can be better understood by docu- 
menting their presence and absence on other related and unrelated 
species of trees. Our experimental design allows for this comparison. At 
1200-1400 m we fogged four Lithocarpus trees and one tree each of 
Engelhardia mersingensis, Callophyllum sp.. Primus sp., and Carpodetus 
arboreus. Beetle species that have no particular affinity to C. acuminatis- 
sima or fagaceous trees in general should show up on some of these 
unrelated species of trees. Other species may occur only on fagaceous 
trees. Analysis of these data is necessary to better assess the true niunber 
of species that are restricted to C. acuminatissima. Our preliminary resxilts 
indicate that herbivores (the leaf-chewing family Chrysomelidae) are far 
less likely to occtir as singletons than are predators (non-omaiiine 
Staphylinidae) (Figure 11.1). General insect predators might be expected 
to occtir widely throughout the forest on any available leaf substrate 
while leaf-chewers are more likely to be restricted to a given species of 
tree because of specialized adaptation to leaf chemistry and other factors. 
Farrell and Erwin (1988), in a study of Peruvian canopy beetles, found 
that Chrysomelidae were more closely associated with individual 'forest 
types' than were Staphylinidae (singletons were excluded from their 
analysis). On balance, we would therefore predict that predators would 
be more likely to occur as singletons. This is consistent with our resxilts 
but is something that requires further analysis. 
Montane forests in Papua New Guinea are not stable ecosystems but 
are constantly subjected to natural disturbances, including landslides, 
drought, fire and storms Qohns, 1986; van Valkenburg and Ketner, 1994). 
Castanopsis acuminatissima appears to be well adapted to this successional 
regime (van Valkenburg and Ketner, 1994). Insects associated with 
successional forests that include Castanopsis would therefore be expected 
to have a high dispersal capacity. A high percentage of the singleton 
species recorded in this study may represent dispersing individuals that 
are not otherwise associated with this tree species. 
Another important component of diversity is turnover (beta-diversity). 
Our experimental design allows us to compare the canopy insect fauna 
of C. acuminatissima at different elevations. Preliminary results suggest 
that ttimover is very high between the three elevational study sites 
(Allison et al., 1993b). Pair-wise Jaccard coefficients between C. acumi- 
natissima trees at different elevations (500 m, 1200-1400 m and 2100 m) 
averaged ca. 0.06 compared with 0.24 for the eight trees sampled at 
1200-1400 m, a four-fold difference. 
Comparison of beetle diversity (Q-statistic) in our study trees at three 
elevations confirms the conclusion of our previous altitudinal transect 
(Allison et ai, 1993b). Table 11.5 shows that diversity is similar at the 
500 m and 1200-1400 m sites and higher at the 2100 m site. This rela- 
tionship between diversity and altitude needs more examination (see 
Discussion 233 
Table 11.5  Comparison of the beetle faunas of Castanopsis 
aciiminatissima at different elevations using the Q-statistic. 
(See Allison et ai, 1993a for details of trees 1, 2, 7 and 8.) 
Eleoation (m) Tree no. (s) Q-statistic 
2100 1 103.31 
2100 2 67.36 
1200-1400 3, 4, 9, 10-14 32-68 (see Table 11.2) 
500 7 43.24 
500 8 53.25 
Allison et al, 1993b). It is possible that it represents phenological differ- 
ences between trees at the varioiis sites, but coxild also be because of 
o-reater topographical and habitat heterogeneity at higher altitudes 
(N. Stork, personal communication). 
As mentioned earlier, one tree was fogged twice and is included in 
this analysis as trees 3 and 9. The total number of individuals and species 
increased slightly in the second fogging, presumably indicating that 
beetles had rapidly recolonized the tree. The fauna of the recolonized 
tree was as similar to the original faixna (Jaccard coefficient = 0.30) as 
the original tree (no. 3) was to tree 4 located close by (Jaccard coefficient 
= 0.31) and fogged 20 days later. These findings are consistent with 
predictions of island biogeography theory (MacArthur and Wilson, 1967) 
that species niimber remains fairly constant but that species composi- 
tion changes diiring turnover events. 
With the exception of Erwin and Scoti: (1980), Gagn? (1979) and several 
temperate studies (Barnard et al, 1986; Blackburn et at, 1993; Gaston et 
al, 1993a), published canopy fogging studies are not based on multiple 
foggings of the same tree species at the same site, making statistical eval- 
uation difficult. Also, early studies did not segregate the overall samples 
from individual trees (the insects were collected on large sheets rather 
than multiple replicate small tiays). Thus, our study and otiiers in this 
volume (Roren and Linsenmair, 1997, Chapter 16; Mawdsley and Stork, 
1997, Chapter 6; Stork and Hammond, 1997, Chapter 1; Wagner, 1997, 
Chapter 9, this volume) are among the first to report data from multiple 
individual conspecific tiees from the same site with subsampling by 
replicate trays. 
To our knowledge the only other species of Castanopsis that has been 
fogged was that reported by Stork (1991) firom lowland Borneo. One tiee 
yielded 396 individuals representing 103 species. This compares with an 
average of 497 individuals (range 300-800) and 117 species (range 
82-155) from our eight tiees. This is sligh?y higher than the mean of 
234       Beetle species diversity in Castanopsis acuminatissima 
112 species that we earlier reported from our 500-m study site (two trees) 
and lower than the mean of 192 species for the 2100-2200 m study site 
(two trees) (Allison et al, 1993b). 
The distributional patterns of individual species are extremely 
complex and will require detailed analysis to understand and summa- 
rize. That is beyond the scope of this paper and will be forthcoming as 
we complete analysis of additional samples. Elsewhere, we have 
discussed the implications of this issue to overall estimates of species 
numbers (Basset et al, 1996a). 
To further refine our understanding of the role of singletons it will be 
necessary to study the feeding ecology of individual species. Such work 
has been undertaken by our colleague, Yves Basset, as part of a compre- 
hensive study of the feeding ecology of the insect fauna on 10 species 
of trees in the Wau Valley, iachiding C. acuminatissima (Basset, 1994, 
1997, Chapter 12, this volume; Basset et al, 1996b and unpublished data). 
Acknowledgements 
This study is an expanded version of a paper presented to the New 
Guinea Biological Society but published only as an abstract (Allison et 
al, 1993a). It was partially funded by the New England Biolabs 
Foimdation and tmdertaken in collaboration with Wau Ecology Institute. 
Harry Sakulas facilitated logistics at Wau Ecology Institute. We are 
greatly indebted to many systematists who have verified morphospecies 
assigrmaents; they w?l be acknowledged fully in a subsequent paper. 
Candida Cardenas, Karin Kami, Jime Ibara, David Preston and Valerie 
Hedlund prepared the beetles for study. Gordon Nishida prepared the 
illustrations. Our colleagues Y. Basset, J. Coddington, T.L. Erwin, and 
N.E. Stork provided useful discussion of some of the ideas presented 
here. 
REFERENCES 
Allison, A., Miller, S.E. and Samuelson, G.A. (1993a) Patterns of beetle species 
diversity in Castanopsis acuminatissima (Fagaceae) trees studied with canopy 
fogging techniques in mid-montane New Guinea rainforest, in Proceedings of 
the Biological Society of New Guinea, (ed. R. H?ft), Wau Ecology Institute Wau, 
Papua New Guinea, p. 89. 
AUison, A., Samuelson, G.A. and Miller, S.E. (1993b) Patterns of beetle species 
diversity in New Guinea rainforest as revealed by canopy fogging: preliminary 
findings. Selbyana, 14, 16-20. 
Barnard, P.C., Brooks, S.J. and Stork, N.E. (1986) The seasonality and distribution 
of Neuroptera, Raphidioptera and Mecoptera on oaks in Richmond Park, 
Surrey, as revealed by insecticide knock-down sampling. Journal of Natural 
History, 20, 1321-31. 
Basset, Y. (1994) Palatability of tree foliage to chewing insects: a comparison 
References 235 
between a temperate and a tropical site. Acta Oecologia, 15, 181-91. 
Basset, Y. (1997) Species abundance and body size relationships in insect herbi- 
vores associated with New Guinea forest trees, with particular reference to 
insect host-specificity, in Canopy Arthropods (eds N.E. Stork, J. Adis and R.K. 
Didham), Chapman & Hall, London, pp. 237-64 (this volume). 
Basset, Y. and Kitching, R.L. (1991) Species number, species abundance and body 
length of arboreal arthropods associated with an Australian rainforest tree. 
Ecological Entomology, 16, 391-^02. 
Basset, Y., Samuelson, G.A., AlUson, A. and MiUer, S.E. (1996a) How many 
species of host-specific insects on a species of tropical tree? Biological Journal 
of the Linnean Society (in press). 
Basset, Y., Samuelson, G.A. and Miller, S.E. (1996b) Similarities and contrasts in 
the local insect fatmas associated with ten forest tree species of New Guinea. 
Pacific Science, 50, 157-83. 
Blackburn, T.M., Harvey, P.H. and Pagel, M.D. (1990) Species number, popula- 
tion density and body size in natural communities. Journal of Animal Ecology, 
59 335-46. 
Blackburn, T. M., Gaston, K.J. and Stork, N.E. (1993) Temporal dynamics of body 
size of beetles on oaks: a cautionary tale. Ecological Entomology, 18, 399-401. 
Erwin, T.L. (1982) Tropical forests: Their richness in Cole?ptera and otiier 
arthropod species. The Coleopterist's Bulletin, 36, 74-5. 
Erwin, T.L. and Scott, J.C. (1980) Seasonal and size patterns, tirophic structure, 
and richness of Cole?ptera in the tropical arboreal ecosystem: the faima of the 
tree Luehea seemannii Triana and Planch in the Canal Zone of Panama. The 
Coleopterist's Bulletin, 34, 305-22. 
Erwin, T.L. (1988) The tropical forest canopy: The heart of biotic diversity, in 
Biodiversity (eds E.O. Wilson and P.M. Peter), National Academy Press, 
Washington, DC, pp. 123-9. 
Erwin, T.L. (1983) Beetles and other insects of tropical forest canopies at Manaus, 
Brazil, sampled by insecticidal fogging, in Tropical Rain Forest: Ecology and 
Management (eds S.L. Sutton, T.C. Whitinore and A.C. Chadwick), BlackweU 
Scientific Publications, Oxford, pp. 59-75. 
Farrell, B.D. and Erwin, T.L. (1988) Leaf-beetie community structure in an 
Amazonian rain forest canopy, in Biology of the Chrysomelidae (eds P. Jolivet, 
E. Petitpierre and T. Hsiao), Kluwer Academic Publishers, Dordrecht, pp. 73-90.. 
Floren, A. and Lii\senmair, K.E. 1997. Diversity and recolonization dynamics of 
selected arthropod groups on different tree species in a lowland rainforest in 
Sabah, Malaysia, in Canopy Arthropod^ (eds N.E. Stork, J. Adis and R.K. 
Didham), Chapman & HaU, London, pp. 344-81 (this volume). 
Gagn?, W.C. (1979) Canopy-assodated arthropods in Acacia koa and Metrosideros 
tar?e communities along an altitudinal transect on Hawaii Island. Pacific Insects, 
21, 56-82. 
Gaston, K.J. (1991) The magnitude of global insect species richness. Conservation 
Biology, 5, 283-96. 
Gaston,   K.J.,   Blackburn,   T.M.,   Hammond,   P.M.   and   Stork,   N.E.   (1993a) 
Relationships between abundance and body size: where do tourists fit? 
Ecological Entomology, 18, 310-14. 
Gaston, K.J., Blackburn, T.M. and Lawton, J.H. (1993b) Comparing animals and 
automobiles: a vehicle for imderstanding body size and abxmdance relation- 
ships in species assemblages? Oikos, 66, 172-9. 
Gressitt, J.L. and Nadkami, N. (1978) Guide to Mt Kaindi: Background to New 
Guinea Montane Ecology, Wau Ecology Institute Handbook, Wau, Papua New 
Guinea. 
236       Beetle species diversity in Castanopsis acuminatissima 
Hammond, P.M. (1992) Species inventory, in Global Biodiversity: Status of the 
Earth's Living Resources (ed. B. Groombridge), Chapman & Hall, London, 
pp. 17-39. 
Hammond, P.M. (1994) Practical approaches to the estimation of the extent of 
biodiversity in speciose groups. Philosophical Transactions of the Royal Society of 
London, Series B, 345, 119-36. 
Hodkinson, I.D. (1992) Global insect diversity revisited. Journal of Tropical Ecology, 
8, 505-8. 
Hodkinson, I.D. and Casson, D. (1991) A lesser predilection for bugs: Hemiptera 
(Insecta) diversity in tropical rain forests. Biological Journal of the Linnean Society, 
43, 101-9. 
Johns, R.J. (1986) The instability of the tropical ecosystem in New Guinea. Blumea, 
31, 341-71. 
MacArthur, R.H. and Wilson, E.O. (1967) The Theory of Island Biogeography, 
Monographs in Popxilation Biology, Princeton University Press, Princeton, 
Nev^ Jersey. 
Magurran,   A.E.   (1988)   Ecological  Diversity  and   its  Measurement,   Princeton 
University Press, Princeton. 
Mawdsley, N.A. and Stork, N.E. (1997) Host-specificity and the effective special- 
ization of tropical canopy beetles, in Canopy Arthropods (eds N.E. Stork, J. Adis 
and R.K. Didham), Chapman & Hall, London, pp. 104r-30 (this volume). 
May, R.M. (1990) How many species? Philosophical Transactions of the Royal Society 
of London, Series B, 330, 293-304. 
Morse, D.R., Stork, N.E. and Lawton, J.H. (1988) Species number, species abun- 
dance and body length relationships of arboreal beetles in Bornean lowland 
rain forest trees. Ecological Entomology, 13, 25-37. 
Schoener, T.W. (1980) Length-weight regressions in tiopical and temperate forest- 
tmderstorey insects. Annals of the Entomological Society of America, 73, 106-9. 
Soepadmo, E. (1972) Fagaceae. Flora Malesiana, 7(2), 265-403. 
Stork, N.E. (1988) Insect diversity: facts, fiction and speculation. Biological Journal 
of the Linnean Society, 35, 321-37. 
Stork, N.E. (1991) The composition of the arthropod fauna of Bornean lowland 
rain forest trees. Journal of Tropical Ecology, 7, 161-80. 
Stork, N.E. (1993) How many species are there? Biodiversity and Conservation, 2, 
215-32. 
Stork, N.E. and Brendell, M.J.D. (1990) Variation in the insect faima of Sulawesi 
- trees with season, altitude and forest type, in Insects and the Rain Forests of 
South  East Asia  (Wallacea)  (eds  W.J.  Knight  and J.D.  Hoiloway),  Royal 
Entomological Society of London, London, pp. 173-90. 
Stork, N.E. and Hammond, P.M. (1997) Sampling arthropods from tree-crowns by 
fogging with knockdown insecticides: lessons from studies of oak tree beetle 
assemblages in Richmond Park, UK, in Canopy Arthropods (eds N.E. Stork, J. Adis 
and R.K. Didham), Chapman & Hall, London, pp. 3-26 (this volume). 
Streimaim, H. (1983) The Plants of the Upper Wattit Watershed of Papua New Guinea, 
National Botanic Garden, Canberra, 
van Valkenburg, J.L.CJi. and Ketner, P. (1994) Vegetation changes following 
human disturbance of mid-montane forest in the Wau area, Papua New 
Guinea. Journal of Tropical Ecology, 10, 41-54. 
Wagner, T. (1997) The beetle faima of different tree spedes in forests of Rwanda 
and East Zaire, in Canopy Arthropods (eds N.E. Stork, J. Adis and R.K. Didham), 
Chapman & Hall, London, pp. 169-83 (this volume). 
Whittaker, R.H. (1975) Communities and Ecosystems, 2nd edn, MacMillan, New 
York.