70

. ·! • Reprinted from

TROPICAL RAIN. FOREST: ECOLOGY AND MANAGEMENT

Special Publication Number 2 of The British Ecological Society

Edited by S.L.SUTTON

T. c. W ·HITMORB

and A. C. CHADWICK

Blackwell Scientific Publications Oxford London Edinburgh

Boston Melbourne 1983

Beetles and other insects of tropical forest canopies at Manaus, Brazil, sampled by insecticidal fogging

TERRY L. ERWIN Department of Entomology, National Museum of Natural History,

Smithsonian Institution, Washington , D.C. 20560

SUMMARY

Data gathered from canopy insect samples taken by a fogging technique indicate that a high percentage of species are confined to one type of forest. Of the forest types in the Manaus, Brazil area, the Mixed-water inundation forest is richest in species, the White-water inundation forest is richest in numbers of individuals. The Terra firme, or non-flooded forest, contains the highest number of restricted species. The White-water forest species are on average larger than species of the other three forest· types studied; the species of the Terra firme forest are the smallest. In all forests, 97% of the beetle species are less than 8 mm total length. Herbivores are the largest species,' fungivores the smallest. Weevils and leaf-beetles (Curculionidae and Chrysomelidae) are by far the dominant forms of beetle life in the canopy. However, in all samples in all forests thus far fogged, Formicidae (ants) are dominant in number of individuals and biomass.

INTRODUCTION

Results and conclusions drawn from any field study must be considered in the light of the adequacy of methods utilized. Large scale and/or complicated studies often suffer from becoming more an experiment with tools than data gathering on the biota. Misleading data from sloppy (but at first glance sophisticated) field work results first in errors in the original paper, which then, by citation, ripple through subsequent works for years. Not only must we guard against accepting such data from others, we must strive to make our own experiments in the field conform to sound ecological principles, despite the difficulty of doing so under often adverse conditions. The purpose of the present paper is to document the history of the canopy fogging technique as used in the neotropics by certain workers and how that technique has evolved into a reliable tool for canopy arthropod studies. The second purpose is to draw attention to data and analyses previously published concerning canopy beetles of Panama and contrast them with new data from the Amazon Basin. Both geographic areas were studied with a

• primitive fogging technique, nonetheless certain aspects are scientifically interesting. Data concerning canopy arthropods of south-eastern Peru gathered with a much improved fogging technique are presently under analysis and will be reported in a separate paper.

This later study is designed to gather materials in such a way that current and future

59

60 Canopy insects at Manaus

studies are enhanced to a maximum degree satisfying what Raven, in a public address to the International Council of Museums (Mexico City, 1980), called a need to 'record coexistence and to be able to retrieve, for study, whole groups of organisms that occurred together for further analysis.'

MATERIAL AND METHODS

Processing of insect samples

Acquisition of material is described below. All Coleoptera and other insects discussed herein were prepared by the Smithsonian Institution Insect Sorting Center at San Jose State University in California, under the supervision of J . Gordon Edwards. Subsequent to prepara tion, Janice Scott labelled and alpha-sorted all material to morphospecies; final checking was done by me, my colleague Warren Steiner, and the USDA Coleopterists of the Systematic Entomology Laboratory in Washington (Anderson, Gordon, Kingsolver, Spilman, White and Whitehead). Scott, Gloria House, and Linda Sims measured a ll species by selecting the largest and smallest individuals in a morphospecies series to provide a range and then assigned each to a size class. Trophic levels were determined through consultation with many coleopter­ists. For further details, see Erwin & Scott (1980).

Methods of canopy fogging

The first attempt at true fogging of forest canopies seems to have been that of Roberts ( 1973). Others (Martin 1966; Gagne & Martin 1968) had earlier obtained canopy samples with hydraulic sprayers, but later Gagne (1979; based on work done in 1973). switched to the use of the resonant pulse fogger. Roberts' (1973) discussion of field methods describes almost exactly the techniques of Montgomery and Lubin in the first

_ fogging oflowland seasonal forest in Panama and subsequent studies in Venezuela and Brazil. The only difference was the adaptation by them of Gagne's (1979) use of a Dyna-fog machine and suspended cloth sheets for the collection of insect-rain. For later studies, Montgomery ran the suspended sheets along a 50-m transect with sheets paired and the fogger hoisted four or five times along and over the transect. For the last study of Manaus operated by Montgomery (August, 1979), I introduced moulded plastic funnel trays (Fig. I) with 4 oz Nalgene collecting bottles in place of the cloth sheets, in order to protect the more delicate species which were previously damaged using the suspended sheet technique. These plastic trays also served for better assessment of which insect species fell from which tree because they were more easily positioned. However, these trays were very heavy and difficult to transport far from roads or boat moorings; they also were too shallow and brushing was required to get specimens into the bottles, thus some damage and loss still occurred. Numerous problems with the modified Roberts' technique resulted in samples that could only be analysed in a general way (Erwin & Scott 1980; Erwin 198 1 ). These problems were: (i) wind drift of the insecticide fog away from the transect so that some trays or sheets had

•

•

T. L. ERWIN 61

little or no insect-rain; in analysis one could not tell if it was drift or lack of insects on a certain tree; (ii) a rather haphazard rotation of the fogging machine in the canopy together with fog dispersion from the ground to about 30 m in the canopy resulted in collections of undercanopy material as well as that desired from the tree-tops; (iii) variance in fog application left little confidence in the attainment of consistent samples from one site to another or even within a single transect; (iv) the amount of insecticide

• and the time of dispersion were inconsistent so that equitability among sites was not

FtG. 1. Transect fogging technique used in Manaus, Brazil, showing moulded plastic sampling trays suspended over water of Black-water inundation forest.

attained; (v) misadjustment of the fogging device allowed droplets of fog to fall on trays, thus damaging small specimens during brushing resulting in inconsistent samples; (vi) no test was used to assess wind drift of insect-rain. These and other problems caused a reassessment of the entire sampling regime. The new design is described below in some detail and is currently in use in Peru by the Tropical Forest Canopy Laboratory of the Smithsonian Institution's National Museum of Natural History.

Current fogging regime in Peru

After forest types are delimited and selected, 144-m2 plots are established by locating a " tree with large upper branches which will sustain pulley systems; this tree then becomes

the north point of the plot. A plot (Fig. 2) is established by selection of three more trees with suitable branches which delimit a vertical column with four sides, 12 m to a side and as high as the canopy top in that particular forest type. This plot is suitably

62 Canopy insects at Manaus

surveyed and marked; perpendicularly, at the middle of each side, a 15-m line is marked off. With the aid of a line-throwing gun (Fig. 3}, pulleys are placed as high as possible in the four trees and to the outside of the plot (not directly over it). Over these pulleys, 'pull-up' ropes are set and tied off for hoisting the fogger during fogging

N

\ F10. 2. Configuration of plot for canopy sampling at Tambopata Reserve, Peru. showing assigned tree numbers/letters. tray numbers, tree dbh and canopy height/depth, and fogger pull-up sites (letters A-D).

Plot size equals 12 x 12m. tray size equals I m2.

operations. Within the plot a rea of 144 m2 lines are tied horizontally between trees in a non-arbitrary way at about 2 m height. Arrangement of trays (Fig. 2) is scattered depending upon availability and position of trees on which to tie the support lines. On these 'clothes-lines' are suspended twenty-five trays (Fig. 4), each I m2 in surface a rea relative to the ground, as sampling surfaces to catch the ' rain' of dead insects after fogging the canopy. The trays are collapsible plastic with square aluminium frames and .

•

•

•

T. L. ERWIN 63

are funnel-form; at the bottom of the funnel is screwed a 4 oz Nalgene bottle with 70% alcohol preservative. Outside the plot area, perpendicular to each side, are strung three additional trays at 5-m intervals (Fig. 2) to catch individuals which leave the fogged area before dying and dropping. In addition, these outlier trays collect wind drift

• should there be any, thus providing a measure of the success rate of capture within the

Fto. J. Line-chro wing gun used to establish rope and pulley system in canopy. Adapted fishing reel aids in line retrieval.

plot area, and also telling something of the kind of species with higher escape capabilities. Other tests of internal consistency are run at selected sites; e.g. sites are

• refogged a few days later to determine recolonization and recruitment, drop times are duplicated sequentially, and branches are shaken after initial samples are bottled to see how much of the biota is left lying on leaf surfaces or on branch tops. Special collecting trays for rainy season samples are now under design; during a fogging operation a sudden downpour will instantly fi ll the plastic tray with water and wash away specimens, thu~ ruining several days work.

64 Canopy insects at M anaus

OGCC the set-up is in order for fogging, preparations are made to return to the site early the next day when there is little or no wind blowing and warm air is rising from the forest floor. Although the commercial Dyna-fog machine is normally hand-operated, I attach a small radio-controlled servo unit to the on/off lever so that the machine can be automatically turned on while up in the canopy. The radio-controlled fogging machine is started and hoisted into the canopy; at 15m above the forest floor the fog is turned on by the radio and it is run for l min at a dial setting of' 5' ( d~nsity of fog). The process is

FtG. 4. Plot fogging technique used in Tambopata. Peru, showing foldable plastic sampling trays suspended in Terra Iinne forest.

repeated at each of .the four corners. Formulation of the insecticide is set to provide 40 000 cubic metres of fog during this 4-min period, thus completely over-fogging the amount of canopy available in the plot (about 3-4000 m3). The fogger is moved vertically above 15 m and rotated 90° at each of the four pull-ups (Fig. 5). In this way, the entire 4000 m3 of canopy receives an even amount of insecticide fog and is overfogged enough to make up for any special circumstances, e.g. deeper canopy, denser foliage, wind, or pull-up lines being slightly away from corners, etc. It is the 'over-fog' capability which allows comparisons between samples: replicate series are a back-up to confirm that samples a re comparable. Drop-time is 3 hours from the end of fogging at each plot. Material is then collected from the trays and subsequently transferred from 4 oz bottles to 4 dram vials for return to the processing laboratory in Washington. Measures are made electronically (Erwin 1978) and all data (including species numbers, counts, and measures. and all information about tree sizes and positions, species, and ages) are stored as computer readable and accessible. The

•

..

•

T. L. ERWIN 65

number of plots needed per forest type to assure adequate coverage of tree species present is determined using a planned transect method (Scott Mori, pers. comm.) and a survey system like that of Hubbell & Foster ( 1983). Concurrent recolonization studies to assess faunal recovery in previously fogged canopy are under way and will be presented separately.

'. ~. : ~ _.~ .DLw:::Jif\;,.111~'!'>·~ !~

~

FIG. 5. RadiO<Ontrollcd Dyna-fog machine in canopy being rotated through 90° angle.

Protocol for sampling regime

Forests fo r sampling are selected at reserves or near research stations where logistics are not overwhelmingly complicated. The amount of field gear and personnel needed for this kind of project are considerable and certain transportation requirements can be found only in areas accustomed to large scientific expeditions. Sites at each forest are selected by visual observation of canopy and understorey conditions with regard again to logistics of operating machinery and also maturity or type of forests, remoteness from homes or settlements and remoteness to other long-term study sites where living insects are being studied.

FAUNA OF EARLY DRY SEASON (WITH INUNDATION) OF FOUR TYPES OF

FOREST AT MANAUS, ~RAZIL

Erwin & Scott ( 1980) provided an analysis of the Coleoptera fauna of a single species of canopy tree in lowland seasonal forest of Panama. Although the fogging technique was

66 Canopy insects at Manaus

in a primitive state of development, certain aspects of the fauna could be ascertained. Similarly, samples from the second large study at Manaus (August-September, 1979) could be analysed in a general way. Results and data from this second study are here presented for the first time.

All histograms and tables herein are based on analyses of 49% of the trays in ten 50-m transects, three in each of the Black-water inundation forest, Mixed-water inundation forest, Terra firme forest, and one in the White-water inundation forest. Replicate transects within each forest type were proximate to each other, not more than 1000 m distance. All sites are within 70 km ofManaus (Fig. 6) and were fogged in the early dry season when the rivers were beginning to subside, but the forests were still completely inundated. The inundation forest types are fully described by Adis (1981); Beck (1969, 1971); Irmler (1975, 1977); and Prance (1979). Ranzani (1980) described the soils of the Terra firme fogging site. Beck ( 1971) described the Terra firme forest at Reserva Ducke near Mana us, similar visually to that fogged some 40 km north, but no description of the fogged forest as such exists. Figure 7 shows the total number of adult and immature insects (usually prepared by pinning, the alcohol specimens are not yet counted) in ascending order of quantity. Hymenoptera, excluding ants, rank only third behind Coleoptera and Homoptera; however, as in all samples since those in Panama taken in 1974, ants are dominant in the canopy, both in numbers and in biomass. Ants are certainly the force to be reckoned with in the tropics. Field observations and preliminary counts of Peruvian samples recently obtained uphold these figures, with the exception that Diptera (ranked fourth in Manaus) seems to be most dominant, at least in the rainy season samples. Ants constitute one-sixth of the arthropods in rainy season samples of the young Terra firme forest at Tambopata, Peru.

Figure 8 shows the number and percentage of adult and immature insects per forest type. Based on extrapolation to I 00% of the samples, ~he single sample from the White-water system is lower than any other forest, but only slightly lower than the Black-water; Mixed-water is considerably higher than aJJ other samples; and the Terra firme forest is intermediate. Figure 9, also based on extrapolation to 100% of the samples, shows nearly the same for adult Coleoptera, except Black-water is slightly lower than White-water and Mixed-water is by far higher than any other forest type. Figure 6 shows the geographical location of the fogging sites with respect to Mana us and each other. In addition, Fig. 6 presents graphically data given in Tables I and 2. Species restriction to each forest is marked and very few species are shared among combined samples of transects in more than two types of forest. Table 1 shows that fully 83% of the beetle species are restricted to one type of forest in the dry season; only 14% are shared between two forest types and 3% among three or more forest types. Table 2 shows that of the 83% restricted species of beetles, most are found in the Terra firme forest, least in the White-water forest.

With regard to sizes of individuals of beetle species in the different forest types see Fig. lO and Tables 3 and 4. Terra firme forest appears to have more species of smaller size, peaking at I mm, and it also contains the largest species; the larger size classes appear to be rather regular, thus the 'curve' is more or less a regular sigmoid one. All other forest types have a peak at 2 mm with many fewer species in the I mm class, the

•

.I

T. L. ERWIN

Terra flrmt Km 69, INPA

Block-woler Roo Torumii Mlrom

Mo~ed-woler

Porono do Xobottnonho

A. End• "'•< 8. Shared w/ MW C. Shored w/ WW o. Snored w/ ew E. Shor ed w/ MW and 8 W F. Shored w/ WW and MW G. Shored w/ WW and 8W H. Snored w/ all

A. Ende""c

78·2 ,. ~·9 ,.

2·4 "· 7·9.,.

2·6 "· 0•6 " · 0·3,. 2•1 %

8. Shored w/ TF C. Shored w/ MW 0. Shored w/ WW E. Snored w/ TF and MW F. Shored w/ ww and MW G. Snoreo w/ TF and ww H. Snored w/ all

A. Endemic

64·4,. 9·7"'

11 · 9 .,.

3·6 .,.

3·2 "· 4·3 ,. 0 ·4 ,.

2·~ "·

8. $1\0red w/ T' C. $hated w/ MW

Whrte·woltr 0. $hated w/ BW Poronci Costa de E. Shored w/ TF ond MW Jlha de Curari F. Shored w/ ew ono MW

A.Endemte 8. Shared w/ ew C. Snored w/ ww 0. Snored w/ fPI E. Shored w/ BW and ww ' · Snored w/ BW and TF G. Shored w/ ww and TF H, Shored ..t oil

70·6 " • G. Shored w/ TF ond 8W 7· 2 % H. snored ..t oil

11 •3,..

4· 3 "· 2·6 "· 2 · 0 "· 0·4 "· 1•5 ,.

~8·4 ,. 3·6,.

23·5 ,. 4·5 ,. 0·9,. 5·4%

0·4 "· 3-2%

67

Fro. 6. Coleoptera species shared between and amo ng forest types at Mana us. Braz.ll. overlain on genernliz.cd

map o f area showing location of fogging transects. Percentage of total fauna shared among forest types is

represented by pic diagrams which arc to be rclld counterclockwise beginning witb larger shaded area as· A'.

Mixed-water having the most, and White-water the least. Both the Mixed and Black-water forests have a very irregular occurrence of larger size classes; but the White-water is not too irregular. The Mixed-water forest has a higher proportion of 3

mm size class; the White-water has a paucity of species in the 5 mm size class; the Black-water 'curve' flattens out between 5 and 7 mm, unlike that of any other forest.

68 Canopy insects at Manaus

-;; .... g II 000 0 ~

9000 a> ... 10 311 F'ormicidoe

... c:

7000 .. E ·;:;

4845 • ~ ..

5000 u j -0 3000 2063

117~7r ' 557939 1934

2 7 8 9 3a 46 sa 12922a ·x -f l other

.. J:> E

1000 " z

e e 0 0 e e e g e 0 e e 0 0 0 0 0 0 ., 01 Qj ~ 01 ., ., 0 ., Qj ., 01 Qj Qj !i Qj Qj Qj Ci. Ci. Ci. Ci. Ci. Ci. c: Ci. Ci. Ci. Ci. Ci. Ci. ~ Ci. Ci. Ci. 0 0 0 ;;; 0 "' 0 0 2 .., 0

~ 0 e 0 0 0 ~ 0 0 u

~ E ... .a Qj 0 :; c: u s: ., c: .. s: ~ ~ ., 0 0 ~ a: ~ Qj ~ E E ., %: "' 0

0 ioJ ~ z .... Cl. %: u v; 0 .. s: .... s:

~ ~ %: w

F10. 7. Insect specimens fogged from four types orrorest at Mana us. Brazil. arranged by Order with numbers o f each• . Hymenoptera was highest due to a great number of ants.

5000

4500

.. 4000 c .. E 3500 u .. Q.

"' ~ ...

3000

.E 2500 0

.8 E:

2000

" z 1500

1000

3825 • - 16% •

3513

2581

} 18 % x•2446 ;;;:

68 ""j} x•1928

1394

ltock- -otet IIO~ot ,_.,. .. I

Whlto - ..,_tor 14~ ~ of _ ... . 1

2044

MIM d • • alet 131% of ........ )

Type of forest

1390

Tt lfO fllft"' t

141 % of ........ ,

29 "·

Fto. 8. Insect specimens fogged from fo ur types of forest at Manaus. Brazil. with range, mean. and percentage of total catch• . White-wa ter fo rest was sampled only along one transect.

T. L. ERWIN

900r-----------------------------------~

II> c .. E ·;:; ~ 600 "' !:' .. Q. 0 .. 0 u 0 300 ~ .. .0 E ::> z

825 794• - 16 'Yo*

535

.7• 337 15 ~. 4141} 223

..::: }18%

187

O~--~B~Io~ck~-w-o~ltr--~Wh~ilo~--.o~I---~M~Iu~d-~wo~to-r ~T~or-ro~h-rm-o --~ (60%of (4 5%of (37%of (48 %of somptes) somptu) samples) somplu)

Type o f foresr

69

Fro. 9. Adult Coleoptera specimens fogged from four types of forest at Mana us. Brazik with range. mean. and percentage of total catch • . White-water forest was sampled only along one transect.

Table 3 shows that smallest beetles occur in the Terra firme, the largest in the White-water forests . Table 4 indicates that 54% of all species in all forests are 2 mm or less, and 97% of all species sampled are 8 mm or less. Further testing of the methods in Peru will need to be done to see if large species so notable in tropical forest are non-canopy species or whether the scarcity of larger size classes in our samples is simply because fogging does not bring them down,

Tables 5 and 6 provide information on the trophic !evels of adult Coleoptera species of the canopy. Table 6 lists all families in four general trophic classes; within each class families are listed in descending order of mean length, except the subfamilies

TABLE 1. Distribution of adult Coleoptera species among four forest types at Mana us

Restricted. one forest 83%

Shared, two forests 14%

Shared, three forests Shared. four forests 2% 1%

Total species: 1080: total specimens: 24 350.

TABLE 2. Distribution (%) of restricted Coleoptera from Table I

Forest type Restricted species Percentage of total

Black-water forest 179 20 White-water forest 129 14 Mixed-water forest 325 30 Terra firme forest 266 36

Total 899

70

~ 120 g

110

.. .. ·;:; .. Q. .. !! ~ ... 0 .. 8 0 .. o; .Q E ~

z

Block-water

Canopy insects at Mana us

Size of class in mm of lolollenqlh

While-water Mi~ed-waler Terra firme

Forest type

FtG. 10. Coleoptera species fogged from four types of forest at Manaus. Brazil. arranged by size classes and forest types.

TABLE 3. Mean length of adult Coleoptera in relation to forest type

Forest type Black-water White-water Mixed-water Terra firme

Number of species Mean length (mm)

278 3· 1

221 3·2

460 3·0

340 2-8

TABLE 4. Percentage distribution of Coleoptera species adults in size classes in four forest types at Manaus

Size classes Forest type 2 mm or less 4 mm or less 8 mm or less

Black-water 53 78 97 White-water S4 79 96 Mixed-water 49 83 97 Terra firme 60 82 97

Total fauna 54 81 97

TABLE 5. Mean length of adult Coleoptera in relation to trophic level

Number of species Mean length (mm)

Trophic level Herbivore Fungivore Predator Scavenger

795 3·2

29 2·0

145 2·7

100 3·0

•

..

•

T. L. ERWIN 71

TABU 6. Coleoptera families and subfamilies arranaed in descending order of size and trophlc level with numbers of species in each

Trophic group (family: subfamily) Mean length (mm) No. of species

Herbivores Scarabaeidae 11 ·20 5 Cerambycidae 6·52 42 Oedemeridae 5·00 7 La&riidae 5·00 2 Elateridae 4-47 35 Cantharidae 4·00 2 Buprestidae 3·55 29 Ptilodactylidae H5 2 Curculionidae 3·03 337

Nonidentifiable 1·00 4 Anthonominae 1·64 14 Apioninae 3·22 113 Attelabinae 2·33 3 Baridinae 2·32 14 ' Broadnose' subfamilies 5·58 12 Ceutorhynchinae 1·50 2 Chelinae 3·00 I Cossoninae 1·67 3 Cryptorhynchinae 2·72 91 Cylindrorbininae 2·00 I Erirdhininae 1·50 4 Hylobiinae 7-92 12 Hyporinae 5·75 2 Magdalinae 2·33 3 Mynnecinae 3-78 9 Prionomerinae 2-86 7 Rhynchitinae 1·00 Tachyaoninae 1·00 Tychiinae 1·00 4 Zygopinac 2-42 36

Chrysomelidae 3·00 170 Non identifiable 6·00 3 Alticina~ 2-77 60 Cassidinae s-25 4 Chlamysinae 2·50 5 Chrysomelinae 6·00 3 Cryptocephalinae 1·98 20 Eumolpinae 2·68 60

Galerucinae 5·00 9 Hispinae 4·50 2 Lamprosominac 2-25 4

Throscidae 3·00 10 Bruchidae 2-70 10 M ycetophagidae 2·59 9

• Hclodidac 2-46 28 Mordellidae 2-46 37 Rhipiphondae 2·00 I

Monommatidae 2·00 6

72 Canopy insects at Manaus

TABLE 6. (cont.)

Tropic group (family : subfamily) Mean length (mm) No. of species

Scolytidae 1·67 12 Languriidae 1·33 3 • Anobiidae 1·25 36 Phalacridae 1-17 12

Scavengen L T enebrionidae 4-17 57 Ptinidae 2·00 I Nitidulidae 1·94 16 Anthicidae 1·69 13 Dermestidae 1·00 6 Euglenidne 1·00 6 Hydrophilidae 1·00

Fungivores Anthribidae 3·50 6 Platypodidae 2·50 4 Endomychidae 2·17 6 Erotylidae 1·67 3 Lathridiidae to{)() 3 Pselaphidae 1·00 3 Scaphidiidae 1·00 Biphyllidae 1·00 Ciidae 1·00 Eucinetidae 1·00

Predators Lampyridae 9·57 7 Lycidae 6·38 4 C1eridac 5·30 s Eucnemidae 3·33 3 Carabidae 3-23 13 Ostomatidae 3·00 I frogossitidae 3·00 4 Staphylinidae 2-67 35 Melyridae 2·00 2 Rhizophagidae 2·00 I Coccinellidae 1·66 50 Colydiidae 1·60 5 Cucujidae 1·60 5 Histeridae 1·50 2 Corylophidae 1·00 7 Scydmaenidae 1·00 I

Unknown food requirements Chelonariidae 1·00 Clambidae 1·00 4 Discolomidae 1·00 I • Nonidentifiable families 5

•

T. L. ERWIN 73

ofCurculionidae and Chrysomelidae which are listed in alphabetical order. In addition the number of species is given for each family. Table S shows that herbivores are the largest and fungivores the smallest in terms of mean size of individuals. It should be noted however, that family trophic assignments are based on best-estimates of material collected with full recognition that some families have diverse food preferences among species.

CONCLUSIONS

4 The most important conclusion to be drawn from the data presented above is a substantiation of the prediction (Erwin 1981; Erwin & Adis 1982) that tropical canopy species have low powers of vagility, thus are found restricted to forest types. Of the l 080 species of adult Coleoptera here analysed, 83% are restricted to one kind of forest ('endemics' of Fig. 6), 14% are restricted to two forest types. This clearly supports the prediction that there may be as many as 30 000 000 species of insects in the world (Erwin 1982), given the number of forest types in the Amazon Basin alone. It is also of importance when setting aside tracts of land for conservation to know that in terms of

insects, at least, geographically small biotypes are unique to thousands of species.

Another major finding is that the Mixed-water forest is the richest by far in species of insects indicating its general richness in other biota and its probable higher nutritional state in comparison with other forest types. However, the Terra firme is the richest in terms of restricted species and second richest in numbers of species, and the White-water (based on one transect) is the richest in terms of individuals. The richness of restricted species in the Terra firme forest is perhaps explained partly by the fact that this type of forest seems to carry a higher load of smaller species ( I mm class, see Fig. 10); another possible explanation is the susceptibility of Terra firme forests to refugial-breakup during the Tertiary (Prance I 982.a). The richness of individuals in the White; water is possibly due to the disturbed nature of the habitat and its continuous distribution over large distances (rivers of the Amazon Basin). The White-water system · is in a constant state of flux during flooding and probably acts like a pioneer community, thus a few species produce high numbers of dispersants; note also that

small beetles, those with potentially greater dispersal problems, are relatively few in this habitat, whereas the Terra firme has the highest number of small species. The White-water forest is lowest among the fou r in terms of species, although the numbers recorded here are based on a single transect and must be regarded as tentative. The plants in this transect however, are more continuously distributed over wide areas (i.e. along White-water riverine systems and gallery forests away from Amazonia proper) than are plants of the other types of forest (Prance 1982b); thus, it is likely that insect species of this habitat too a re more widespread.

Among adult Col~optera, first the Curculionidae and then Chrysomelidae are the dominant forms of canopy life. No other groups in any trophic level approach the richness of species of these two families. Tenebrionidae are the dominant scavengers, Anthribidae and Endomychidae the dominant fungivo res, and Coccinellidae and Staphylinidae the dominant predators. In terms of size (mean length), Scarabaeidae

74 Canopy insects at Manaus

are the largest herbivores by far, Tenebrionidae the largest scavengers, Anthribidae the largest fungivores, and Lampyridae the largest predators.

ACKNOWLEDGMENTS

Besides those persons mentioned in Material and Methods, I thank Max Gunther of Explorer's Inn, Peru, for making available facilities and personnel, the Scholarly Studies Program and Amazon Ecosystem Project of the Smithsonian Institution for funding for field work, the Department of Entomology and National Museum of ~

Natural History of the Smithsonian for special equipment, INPA at Manaus for support both in the lab and field, and particularly tbe following people for their efforts in the field: Moira Sucharov, Linda Sims, Nigel Stork, David Pearson, Joachim Adis, Gene Montgomery, Joseph Patt, Whitcomb Bronaugh, Ursula Feller, Marina Gunther, Cris Jensen, Allison Date, Leslie Shimmel, Vera Krischik, Gerardo Lamas, and several Brazilian and Peruvian students. George Venable provided illustrations.

REFERENCES

Adis, J . (1981). Comparative ecological studies of the terrestrial arthropod fauna in Central Amazonian Inundation-Forests. Amazoniana, 1, 87-173.

Beck, L. (1969). Zum jahreszeitlichen Massenwechsel zweier Oribatidenarten (Acari) im neotropischen Uberschwemmungswald. Verhandlungen der Deutschen Zoologischen Gesellschaft. pp. 535--40.

Beck, L. (1971). Bodenzoologsiche Gliederung und Cbarakterisierung des amazonischen Regenwaldes. Amazoniana, 3, 69-132.

Erwin, T.L. (1978). Techniques. Coleoprerisrs Bullerin, 3~ 372. Erwin, T.L. (1981). Taxon pulses, vicariance, and dispersal: an evolutionary synthesis illustrated by Carabid

beetles. Vicariance Biogeography: A Critique (Ed. by G. Nelson & D. Rosen), pp. 159-96. Columbia University Press, New York.

ErwiD. T.L. (1982). Tropical forests: their richness in Coleoptera and other arthropod species. Coleopterists Bulletin, 36, 74-75. '

Erwin, T.L. It Adis, J. (1982). Amazonian inundation forests: their role as short-term refuges and generator of species richness and taxon pulses.- Biological Diversification In the Tropics (Ed. by G. Prance), pp. 358-71. Columbia University Press, New York.

Erwin, T.L. It Scott, J.C. (1980). Seasonal and size patterns, trophic structure, and richness of Coleoptera in t'be tropical arboreal ecosystem: the fauna of the tree Leulrea seemannii Triana and Planch in the Canal Zone of Panama. Coleopterists Bulletin. 34, 305-22.

Gape, W.C. (1979). Canopy-associated arthropods in Acacia koa and Metroskkros tree communities along an altitudinal transect on Hawaii Island. Pacific Insects, lt, 56-82.

Gape, W.C. It Martia. J .L. (1968). The insect ecology of red pine plantations in Central Ontario. V. The Coccinellidae (Coleoptera). The Canadian Entomologists, 100, 835-46.

Hubbell, S.P. It Foster, R.B. (1983). Diversity of canopy trees in a neotropical forest and implications for conservation. Tropical Rain Forest: Ecology and Management (Ed. by S. L. Sutton, T. C. Whitmore & A. C. Chadwick). pp. 25--41. Blackwell Scientific Publications. Oxford.

Irmler, U. (1975). Ecological studies of the aquatic soil invertebrates in three inundation forests of Central Amazonia. Amazoniana, S, 337-409.

lrmler, U. (1977). Inundation-forest types in the vicinity of Manaus. Biogeographica, 1, 79-99. Martin, J .L. (1966). The insect ecology of red pine plantations in central Ontario. IV. The crown fauna. The

Canadian Entomologis t, 98, 10-27. Prance, G.T. (1979). Notes on the vegetation of Amazonia. III. The terminology of Amazonian types subject

to inundation. Briuonia. 31, 26-38.

•

T. L. ERWIN 75

Pruc:e, G.T. (Ed) (198la). Biological DhJers{fication in tM Tropics. Columbia University Press, New York. Pruce, G.T. (198lb). Forest Refuges: Evidence from woody angiospenns. Biological Dlotrs{fication in tM

Tropics (Ed. by G. Prance), pp. 137- 58. Columbia University Press, New York. Ranzaal, G. (1980). Identifica~ao e caracteriza~iio de alguns solos da Esta~ao Experimental de Silvicultura

Tropical do INPA. Acta Amazonica, 10, 7-41. Roberts, H.R. (1973). Arboreal Orthoptera in the rain forests of Costa Rica collected with insecticide: a

report on the grasshoppers (Acrididae), including new species. Proceedings of the Academy of Natural Sciences of Philadtlphia, 125, 4~6.