Rev. per. Ent. )2: 71-77. Diciembre 1989 (30 14J01to 1990)

Canopy arthropod biodiversity: a chronology of sampling techniques and results

Terry L. Erwin. 1

RESUMEN

ERWIN T L. 1989. Biodiversidad de artr6podos en copas de irboles: aonologfa y resultados de las t«nicas de muestreo. Rev. per. Ent. 32.- Se han desarrollado varios maodos parallepr a las copas de los irboles y efeauar muestreos de artr6podos, teniendo el convencimiento que en l01 bo.ques uopicales en d.icho habitat se encuentra un inmenso n11mero de espec:ies nuevas para Ia ciencia, lo mismo que tantos procesos ecol6gicos que necesitan ser estudiados. Se hace una revisi6n cronol6gica de Ia literature al respeao y se dan recomendaciones para los estud.ios de artr6podos de las copas de arboles, que estm en sus comienzos. Se incluyen observaciones sobre Ia fauna de Ia copa de los arboles y debajo de ella, efectuados porIa t&:nica de Ia nebulizaci6n.

Palabras clave: Artr6podos de lu copas de los arboles, artr6podos de los bosques, t&:nicas para colecci6n de artr6podos en , .rboles.

SUMMARY ERWIN T L. 1989. Canopy arthropod biodiversity: a chronology ot sampling techniques and results.

Rev. per. Ent. 32.- Following the recognition that the tropical forest canopy held untold numbers of species new to science and ecological processes that needed to be studied, various forms of sampling and canopy access have been developed. The literature on these is reviewed chronologically and recommendations are given for those beginning canopy arthropod studies. Observations on arthropod faunae of canopy and undercanopy sampled by the foainl technique are given.

Key words: Canopy arthropod., forest arthropods, techniques to collect canopies' arthropods.

INTRODUCI10N Henry Walter Bates clearly understood insect

radiation in the tropical forest canopy during his 11 years in the Amazon (1848-1859). He wrote of Agra (Bates 1884) the following: " ... this re· markable genus, in which the arboreal type of the family Carabidae reaches its highest develop­ment ... their head quarters are the great forests near the equator in South America, the numbers diminishing on approaching the tropical limits north and south". William Beebe (1917) was also attracted to the 'unknown' of the tree tops and built platforms to observe jungle life there. Though Bates pondered canopy life from the ground and Beebe ascended to watch verte­brates, it was more than half a century later that serious studies began on anhropod life in the last biotic frontier (Erwin 1983a).

Only in the last three decades has collecting arboreal species or sampling them for various types of ecological studies been other than felling of trees or the use of malaised and sticky traps hoisted into the treetops with ropes. Martin (1966) used hydraulic spraying to sample the in-

I. Netional Musoun of Nerurti History. Smithsonian Institution, Wuhington OC 20,60.

sect fauna of pine trees in Canada, and Kikuza­wan and Shidei (1966) used "smoking" techni­ques on a japanese forest, thus opening an era in which further canopy exploration revealed un­anticipated numbers of species, most of them new to science (Erwin 1982).

In the last 15 years, it has become abundantly clear that tropical tree canopies are extremely rich in insect and related anhropod species (Roberts 1973, Wolda 1979, Erwin and Scott 1980, Erwin 1981, 1982, 1983; Stork 1987a, 1987b; Farrell and Erwin 1988, Noyes 1989). Erwin and Scott (1980) and Erwin (1981) first reponed on the incredible numbers of coleopter­ous species found in tropical canopies based on sampling of the tree species Luehea semannii Triana and Planch (Tiliacea) in 1974 in Panama and Wolda (1979) reported on the richness of Homoptera in these same trees. Subsequent pa­pers by Stork (1987a), Farrell and Erwin (1988 in press), and Kay and Miller (in prep) report on various aspects of tropical canopy insect species using fogging methods refined by Erwin {figs. 1, 2, 3), however, thus far only Erwin (1983) documented the methodology underpinning this technique, though until now only partially so, and Adis et al. (1984) and Stork (1988) have

-71-

n REVISTA PERUANA DE ENTOMOLOGIA Vol. 32

offered critiques of the method. Gayne (1979) and Southwood et al. (1982b) have documented hydraulic spraying techniques and Hijii (1983, 1989) provides an extensive bibliography on Ja· p~ese efforts at canopy studies using the "smo­king" technique. Althought the hydraulic spraying and smolcing methods are similar to true fogging techniques in that they disperse insectici­de into the canopy, they have severe limitations in acquiring anywhere near the total fauna even with cumbersome aids (Afreh-Nuamah and Thornhill 1988, Hijii 1986). In tropical forest si­tuations, they are not at all useful because of crown architecture, overlapping branches of dif. ferent tree species, and density of epiphytes. Hy­draulic spraying does not penetrate well all the niches and crevices nor even the center of dense­ly foliated trees, and "smolcing" is only good for small, densely-leaved trees such as conifers, and as Hijii (1986) has shown it works better with tented trees, a technique impossible to use in tro· pical forests with 55 to 70 m high trees, usually interdigitated with others.

The purpose of this paper then is to detail the true fogging method and its history, in addition to reviewing other methods that have been used for sampling arboreal arthropods in both the tro· pical and temperate zones. I also provide observations on the composition of the general tropical forest canopy insect fauna and its be­havior when coUected with true fogging methods, as well as documenting uses to which the material from canopy samples have been put, or might be put.

METI-IODS OF SAMPLING CANOPY ARTHROPODS: A CHRONOLOGY

CoUecting from tree canopies previous to Martin's work mostly involved arboreal traps or destructive techniques, including felling of trees. On the other hand, Hartesvdt et al. (1975) built an elevator on the side of a Giant Sequoia in Cali­fornia and worked inside the canopy to study in­sect biology associated with the tree, foreshado­wing Perry's tree-climbing techniques in Costa Rica by 10 years.

In order to assess the flight patterns of patho­gencarrying bark beetles, Merrill and Skelly ( 1968) sent a modified window flight trap (Chap­man and Kinghorn 1955) above the main canopy of an oak-pine forest in Pennsylvania with ropes and pulleys. Eighteen families of Coleoptera were coUected; other material was discarded. In 1981, Novak et al, reported their use of suction traps and human· bait located on a canopy-elevation tower in a forest in Indiana to sample mos­quitoes. Their work was proceeded by others who invented or improved the suction traps

(Taylor 1951,Johnson and Taylor 1955, Horsfall 1961) and continued a tradition of studies of ver­tical stratification of mosquitoes using other means (Bates 1944, Snow 1955, 1975; Love and Smith 1958, Scholl et al. 1979). Removal of pans of the forest canopy to assay the insect species has been used by several investigators, for exam­ple climbing and pruning (Ohman et al. 1983), pole pruning after aerial spraying of insecticide (Martinat et al. 1988), general pruning (Hunter 1987, Verghese et al. 1988), clipping shoots in lower canopy {Kidd and Tozer 1985), or just pic­king leaves with miners (Wanjala 1978). Beating sheets used under lower canopy leaves (Futuyma and Gould 1979) and accumulated debris in lo­wer branches is a common collectin8 technique of many collectors, but Salman and Lowman (1983) restricted their beating to Antarctic Beech buds at night for a study of leaf beetle biology in a rainforest of New South Wales, Australia.

J L Martin (1966) and Gagne and Martin (1968) studied insect ecology in red pine planta­tions of Ontario, Canada. Their samples were obtained through the use of a hydraulic sprayer using pyrethrum insecticide. Martin seems to be the first to systematically study an intact forest canopy arthropod fauna. Cunningham and Har­per ( 1977) applied insecticide to the canopy from an airplane and used suspended funnel trays for sampling the faUout.

SAMPLING WITH INSECTICIDAL FOGGING TECHNIQUES

A brief history of the fogging method is given in Erwin (1983a) and those aspects will not be repeated here. RobertS (1973) was the first to use a machine (1966 and 1967) which dispersed a true insecticidal fog (droplets of 5 microns in dia­meter of dichlorvos [DDVP) with great penetra­tion powers. Knowing that many tropical Acridi­dae (Orthoptera) are arboreal, Roberts collected by using a small Aero-Dyne fogger to disperse insecticide into the tropical forest canopy on the Osa Peninsula of Costa Rica by pulling the machine up on rope with a pulley fastened in a tree. Many Orthoptera were collected and noted by Robens, but the rest of the material was dis­carded or not labelled adequately for retrieval as canopy-derived specimens. In 1971, Gagne used a Oynafogger with pyrethrum synergized with piperonyl butoxide to capture Hawaiian forest arthropods in cloth funnels. His pioneering work was not reported until later in Gagne ( 1979). The problems associated with fogging tropical cano­pies through 1984 for ecological studies were associated with an inability to acquire samples from individual tree species. However, this was

Dicicmb~ 1989 ERWIN: MUESTREO DE INSECfOS EN COPAS DE ARBOLES 73

FIG. I : The Oynafogger used from the ground at down puts fog into the tops of 30 m high trees ~gularly due to 1 he warmth of the insecticide and cool temperature near the ground.

not the purpose of the studies in Tambopata, Peru, the most ambitious undenaking thus far using the fogging technique (Erwin 1983). Be­cause the samples were not made in a way that species restricteJ to specific plant species could be detected, their use for analysis is limited, that is they are similar to undercanopy or field sweep samp.Jes Uanzen, pers. comm.). In attempting to analyze the data from those studies, it became dear that statements about the fauna and its rela­tionships with the environment were going to be limited, but generally useful (Farrell and Erwin 1988, Farrell and Erwin, in press) for faunal des­criptions. Earlier attempts to fog single tree spe­cies (for example, in 1974, Lubin and Montgo­mery's fogging of Luehea seemannii. (see Erwin and Scott 1980, Wolda 1979) were difficult to control for specimens from leaves of other trees because of the use of large plastic sheets suspen­ded near the ground that collected general insect rain. Interdigitated branches of Luehea and other trees likely yielded mixed samples. Erwin and Fa­rrell (in prep.) eliminated this problem with new­ly designed sampling funnels and careful place-

FIG. 2: The use of pulleys and ropes in the past allowed fogging of trees higher that 30m, but control of the fogger was difficull.

FIG. 3: Cloth treys suspended from ropes tied berween trees we~ used in studies in southeastern Peru.

ment of these under branches of target species which did not interdigitate with other trees nor have epiphytes (see below and fig. 1). Thus, enough pure samples for several tropical tree spe­cies were obtained and these can be treated sta­tistically.

Another problem faced by canopy researchers has been accessibility to high canopies and super­emergents such as Ceiba, Clarisia, Calophyllum, and Bertolettia. Stork (1987), using the pulley methods of Erwin in Peru (fig. 2), collected from the tops of high Dipterocarps in Suluwesi, but problems of breezes drifting the insecticide and mixed samples from epiphytes with the trees re­mained a problem. Erwin, Farrell and Tobin eli­minated this problem in 1989 at Pakitza, Peru, by establishing climbing ropes in emergent trees, climbing the ropes and dispersing the fog from within the upper canopy itself at dawn (fig. 4, and see below) and later at Pacaya-Samiria National Resetve, Peru during the day. With the use of thl: funnd collectors described below and rope-climbing techniques, it is possible to collect samples from the leaves, branches, and trunks of

74 REVIST A PERU ANA DE ENTOMOLOGIA Vol 32

FJG. 4: Rope-dunbin& teduuqucs allow KCeSS to trees of ~ny height and control of the fot~ ooc branch 11 1 ume af desired.

any tree and have virtually uncontaminated sam­ples. In this way, host specificity of insects can be ascertained with cenainry given a rigorous re­search plan and sufficient samples. Rope­climbing techniques even provide access for habitat manipulation experiments such as con­tinuous removal of arthropod faunae from single branches, but not adjacent ones. A combination of good canopy access and directed fog dispersal must be combined with efficient collecting trays.

FJG. ' ' Aluminum funnds on plasuc Slake S~andatds in 1 forat in Boliv11 with cg-conon wau for C'OI1ecuna cerwn groups dzy.

Collecling Trays

Peterson (1934), Connola et al. (1966), and Southwood (1966) all described drop traps for use in collecting arthropods. The evolution of collecting surfaces for fogging began with Gagne's cloth funnels in 1971, but plastic sheets laid out on the ground were used by Robens (1973) and the same, tied with strings to trees, by Lubin and Montgomery (cf. Wolda 1979, Erwin and Scott 1981). Subsequently, suspended funnel trays were designed (Cunningham and Harper 1977, Erwin 1983) and these evolved to maxi­mize efficiency and minimize hazards to the specimens collected. Several generations of sus­pended tray design and tests have past (sec also Kay and Miller, in prep). The present system used by Erwin and Farrell (fig. 5) employs alumi­num sheets cut so they can be folded into funnels with a collecting area of 0.25 m2; they are also pre-drilled so that a three string harness can be tied on for suspension over water for sampling flooded forests (e. g. fig. 3, although the tray in the photo is of an earlier design). After folding the aluminum sheet funnel -form and interlocking the edges, a nalgene bottle cap with the center bored out is attached to the neck. These funnels are easily stacked and carried in l O's into the forest site where trees have been selected for sampling. For installation, plastic stakes with flat heads are center-drilled and a plastic cup is anached with a screw. This cup fits the nalgene bottle attached to the bottom of the funnel These stakes are hammered or pushed into the ground at desired locations. Just before fogging, noneck bottles are attached to the funnels, filled to one third with 70% ethanol using a garden pressure sprayer, and the bottle with funnel is in­serted into the cup on the stake (fig. 5) . If dry material is desired, an egg carton (24 egg type) can be laid across the mouth of the funnel (Pogue, per comm; fig. 5). After the usual 2 hours of drop time, desired dry specimens are ta­ken off the egg catton with forceps or puter, the rest dumped into the funnel. Funnels are then washed down with the garden sprayer (70% ethanol) and the bottles detached and capped for study in the lab. This method results in sped­mens appropriate both for the ecologist and the taxonomist leaving the specimens in excellent condition for museum collections at the termina­tion of the study.

Insecticides

The number and types of insecticides used by canopy researchers in nearly as numerous as the studies, however, except for experimental studies (e. g. Martinat et al. 1988, Afreh-Nuamah and Thornhill 1988), coUectors and ecologists have used pyrethroid-based chemkals with knock

Diciembrc 1989 ERWIN: MUESTREO DE JNSECrOS EN COPAS DE ARSOLES

down agents. Pure pyrethrum with piperonyl butoxide as a knock down agent was used by Robens (1973), Gagne (1979), Erwin and Scott (1981), Erwin (1983b), Adis et al. (1984). In my studies, begun in Peru in 1981 (Erwin 1983), a change was made to usc the cheaper and highly effective synthetic pyrethroid chemical, Res­mythrin, or "Cross Fire" trademark (now ca11ed "Respond"). Stork (1987) and Paarman and Stork ( 1987) used Reslin E, similar to Res­mythrin, but with high knock down and low kill properties. They even recovered and bred fogged specimens. Observations from fidd tests in Peru between the two versions of synthetic pyrethroid sprayed on adjacent trees at the same time on the same day indicated that Resmythrin was the more effective of the two chemicals and more reliable for extracting canopy faunae.

DISCUSSION ON THE COMPOSITION OF THE GENERAL CANOPY INSECT FAUNA

Based on many years of fogging experience, I offer here observations of a general nature on the canopy arthropod fauna that might be useful to those planning to usc such techniques in the tro­pics.

As noted by Robetts (1973), wind currents are an important factor in the effect of insecticide on the canopy fauna. It is necessary to fog at dawn, wben there is usually no air movement, for effec­tive penetration of all pans of the canopy and consistently reliable results. Near rivers or on ridges, however, even at dawn there can be prob­lems. In 1990, in Pacaya-Samiria National Re­serve, I made effective usc of the fogger at all times of the day and before dawn in the lower canopy, but only with tree climbing techniques was I able to fog high canopies during the day. On cold mornings, I have put fog into tree crowns up to 55 m from the ground, but mic­roclimate is fickle. This is not recommended as a usual method of sampling.

lmmediatdy after the fog strikes the leaves or trunk area, ants begin to drop, as they are highly affected by "Respond." Carabidae and other beetles drop over a period of 30-45 minutes, lar­ger Orthoptera usually take longer to succumb to this insecticide. After two hours, there is virtually no arthropod left in the fogged area. In Tambo­pata (1982-1885), various follow-up tests were made (branch shaking, refogging, second two hour drop collections, etc) and almost nothing was recovered by anx tests. Therefore I conclude that the twO hour effectiveness of the insecticide as advertised is in effect in tropical faunas as wdl as temperature. Interestingly, both at Tambopata and Pacaya-Samiria, refogging after 10-30 days resulted in as big an insect rain as the original

TABLE L Total Dlllllkn io nnk orda of ldected tao­-M: poupiop end pc:rcaltqa o(. NlllpK o( Arthro­pocll &om- iaolat.ed caoopy oE ~lin&. uVntlra (Ouy­~) and M•a.M corJ.u (Bomheca«ae) with 11 UDWentified lianu (lin cWaa of atema • 0.02 to 0.09 em dhh) rachiaa the intft1Wioed tift crowna. No epiphy­ta wae n.iWr &o.. the pouod with 8so40B Leitz binocu. len. App~ 2/) oE the total catdl arc reported io em. cable.

Tuon Number of % of specimens toW

Fonnicidae 57)22 69.6 Coleoptera 7)2.3 9.1 Psocopten 3,314 4.0 Dipteta 2,062 2.5 Collembola 1,8)9 2.2 Araneae 1,647 2.0 Homoptera 1)81 1.9 Acari 1,405 1.7 Orthoptera 1,285 1.6 other H ymenoptera 1,207 1.5 all larvae (not nymphs) 805 1.0 Hemiptera 770 0.9 Thysanoptera 704 0.9 Pseudoscorpiones 614 0.7 Lepidoptera 147 0.2 Neuropten 5 1 0.06 Myriapoda .30 0.04 Dennaptera 26 0.03 Scorpiones 15 O.Ql8 Embioptera 14 0.016 lsopten 14 0.016 Thysanura 11 0.01) Trimptera 4 o.oos Odonata 2 0.002

Total 82,391 100.000

fogging (U tests of large trees, 5 tests of sus· pended dry leaves in the undercanopy). My con­clusion is that the arboreal fauna is highly mobile, at least on a locallevd, and the canopy and sub­canopy are virtual horizontal highways across which mass daily movement takes place.

All groups of insects fall using "Respond" with the noted exception of female scale insects (Coccidae). The nature of the tropical beetle fauna has been characterized by me elsewhere and need not be repeated here (see above). Adis et al. (1984} and Stork (1987} have contributed information also on the general nature of tropical canopy arthropod faunae. Through seasonal sam· pllna regimes it is possible to register species nor­inalfy living inside the wood or bark because at some point in time they disperse and lay eggs on the surface. Large butterflies are uncommon in dawn foggings of the canopy and for years I thought they were escaping the fog. With recent undercanopy fogging experience in Pacaya-

76 REVISTA PERUANA DE ENTOMOLOGIA Vol 32

Samiria, it can now be stated that large butterflies do not escape the fog, rather they roost in the undercanopy at night and are not subjected to canopy fogging at dawn. This is not true for Lycaenidae, which are common in dawn fogging samples from the canopy.

Myrmecophilous insects are plentiful in all tree sampled which had large ant populations; fogging is a recommended method for obtaining these. Arthropods which occur in bromeliads are also easily obtained; the action of the insecticide kills by interrupting synapses and the spasms during this process virtually drive the dying anhropod up and out of the leaf axil. This seems to be true for all crack and crevice arthropods in the canopy too. Table I shows percentages of art· hropods collected from an entire canopy foot­print ( 1000 sq feet) of a single isolated canopy (rwo tree crowns interdigitated with II lianas at height of 33 m and a diameter of 11 m) at Pakit­za, Peru (Erwin, in prep). 1t is clear that all ex­pected groups were sampled. Further analysis of this material and others will indicate whether or not faunal balance is consistent among canopies of different species and different individuals of the same species (Erwin and Farrell, in prep.). Certainly, when compared with faunae of the mid Amazonian Jgapo near Manaus, where va­rious taxa annually migrate into tree canopies from flooded soils (Adis 1988), dominance of groups vary considerably from site to site.

ln all studies to date, it has become abundant­ly clear that biodiversity is far more than a list of species. Patterns of occurrence and dominance even on a local level are highly variable. Many areas need to be sampled before we will have estimates within an order of magnitude of arthro­pod biodiversity on this planet. However, current modem methods of canopy sampling will aid greatly in this endeavor, as the canopy is clearly where maximum tropical biodiversity occurs.

Acknowledgements Thanks go to the many persons who have sup­

ported parts of the fogging program over a pe­riod of some 16 years, including Smithsonian staff and Earthwatch volunteers. In preparing this manuscript, Linda Sims and Micheal Pogue were very helpful in bibliographic searches and Abelardo Sandoval supported efforts in the field. John Tobin and Grace Setvat taught me to climb. To all these and George Venable who drew fig. 4, I extend my heartfelt thanks.

Funding for canopy studies comes from the Scholarly Studies Program, Neotropical Lo­wands Research Project, and Department of En­tomology, National Museum of Natural History, all of the Smithsonian Institution.

,_.__cited

I. Adisj, Lubin YD. Monccomay G G. 1984. Arthropods £rom the canopy of inundated and terra fume forau near M111· •us, Brazd, wilh critial consideration of the Pymlthrum­fouina ~~- Studies on Neoc:ropieal F1un.1 and En­vironment. 19:22).236.

2. Afrch-NUim.h K, Thornhill E W. 1988. ULV apptication to • rree crop canopy: CYaluation of physical performance and pcnctfltion. Insect Sd. Applic. 9(5):629-634.

3. BuS(! Y. 1988. A composite interception rrap for sampling arthropods in tr~ canopies. J. Aust. ent. Soc. 27:213·219.

4. B1tcs M. 19-14. Obsavaliom on the distribution of diurnal mos. quitocs in the tropical forest. Ecology 25: IJ9-170.

5. 0\apman J A, Kqhom J t.l 1955. WID<Iow ru,hc trips for iNccu. c.n. ent. 87:46-47.

6. Connob D P, Ln.ois F B, Mc.Donousb- 1966. Expttuncntal fldd tedwques used 10 CYaiUitc gypsy mcxb, PorthtiTI4 Jt~, conrrolln Nev.• York. J . Econ. Ent. 59:284-287.

7. CunninR)ltm H B, Harper J D. 1977. A funnel trip for monitor· illfl fallout &om forest canopies after insecticide 1ppliatio1U. The F1oridl En1omolo8"1 60(4):263-266.

8. Downs W, Go<N.<in P A. 1987. Forest canopy collect.ion of Trichoptcn. l11, Boumaud, M. and H. Rachct, eds. ~­illfls of the Fifth Inu=alional Symposium on Trichopterl, Lyon. Fr1ncc, 21-26 Jul)' 1986:207-209. Dr. W. Junk pub­lishers, Dordrccht, HoUand.

9. Erwin T L. 1982. Tropic.J Forests: Their richness in Colcoptcrl IOd other anhropod species. Colcopt. Bull. 36(1)·74-75.

10. --- 198}L TropiCII forest canopies, the lase biooc fron· ucr. BuD. EoL Soc. Amcnca, 29(1):14-19.

11.--- 198}b. lXales and othcr arthropods of the tropical forest aoop~a • Man.us. Bruil. sampled with IIII«UCCdaa fouina techniques. '"· Tropical Rain Fomu: EcolotY and MIJl&mlent. Sunon, SL. TC Wluunorc, and AC Chldwick (eds.). Blackwell Scien1iflc Pubtiations. Oxford, United IGllfldom:59-75.

12. Erwin T L, Scon J S. 1981. Seasonal and size p111em1, trophic wuaure. ~nd richness of Coleoptera in the tropical uboiUI c<oty~~em: The ftuna of chc tree Lu~h&~ umt.,.lltt Trilllt llld Planch in the Canal Zone of Panama. Colcopt. Bull. 34(3):305-322.

13. Furdl B D, Erwin T L. 1988. ulf-bccdcs (Chrysomehdacl of • forest anopy in Amazonwt Peru: synoptic list of ttn, IC&JOflliiry ~nd host-lfftlitlions. lll, }olivet P., E. Pctitpicttc, IOd T. Hsiao, (Eds.), Tht Bio/oo of th~ Chrvso,~ftJ,u Dr. W Junk bv Publishcn. The Haguc-:73-90.

1-4. Hijii N. 1983 Arborul arthropod £.uoa in 1 forest. I Prdunm· uy obtavatioo on -.-1 fluctllllions in densily. biomass. IJld flwW ~lion in I ~ obiMU plantl· tion. )tp. ) . Ecology. }3:-135-444.

U. --- 1984. Arborell anhropod f1un1 in • forest. II. Pre­sumed community "ructures btsed on biomass 1nd number of trthropods in 1 Cha~~tllecypllriJ obtMJII pllntttion. J tp. J. Ecol }-4:187-193.

16. --- 1989. Arthropod communities in • J•p- ccdu (C/uM•«y~ris obtM.U D. Don) planution: .bundtncr, biomtss 1nd some propm~. Ecol. Ro. 4:2-1) 260

17. --- 1986. Dms1ty, biOfi\IU, IOd gwld llructurc: of uboruJ arthropods II related IO their inhabited tree IIZC in I CryptomMII jqo11iC1 pltm1tion. Ecol Ro. I 97 118.

18. Gllfle W C. 1979. Clnopy-IIIOCiated anhropods in AQHU Ito. IOd MmosiJmn u-ee comm\lllll:ics .Jon, 111 alticudinaltfll'l· sea on Hawlii ldand. PICilic Inseas 21(1):56-82.

19. ---.Maron J L 1968. The insect ccolocY ol red pii>C plant•uons in central Ontano. V. The Coccinclhdte (Col· c:optert). C.n. Ent. 100:835-846.

20. Horsfall W R. 1961. Traps for determining dir«~ion of llight of insect~. Mosq. News 21:296-299.

2 I. Hunter MD. 1987. Opposi113 df«ts of sprillfl defoliation on late ICIIOf\ Olk caterpillars. Ecol. ent. 12:373-382.

22. ]ohn10t1 C G, T1ylor L R. 1~5. The dcvdopmc:n1 of large suction lrlp$ for lirbomc iNccu. Ann. Appl Btol 4) .51-62.

Didembrr 1989 ERWIN: MUESTREO DE INSECTOS EN COPAS DE ARBOLES n

23. KJdd N A C, Tozer DJ. 198,. ~distriburionofthe largepiM aphid, Cmuil pin~il (Mordv.) within the canopy o( S.:0U pme. PmuJ ryk~11ru L Z. q . Em. 99:3-41·3,0.

24. Love G J , Smith W W. 1~8. ~ strati6ation o( mosquttoes. Ma.q. Nnn 18:279-283.

V . Martan J L. I 966. ~ insect ecolotY oC rrd piM plantatioos in cmtr.J Ontario. IV.~ crown fauna. Can. Enr 911.10-27.

26. Martinat P J, Coffman C C. Dodfle K, Cooper RJ , Whnmore R C. 1988. Effects of Diflubenzuron on the canopy arthropod community in a cenrral Appalachian forest. J . Econ. enl. 81(1):261-267.

27. McmD W. Skdly J M. 1968. A window trap for coUection of UU«U above the Cores~ canopy. J. Econ enr 611'): 1461-1~.

28 Moran V C. Southwood T R E. 1982. ~guild composihoo oC anhropod communities in rrea. J. Anim.. Ecol R289-306.

29. Novak RJ, PdoquinJ, ROOm W. 1981. Vcrucal distribution oC adult mosquitos (Dip«cra: Culicidae) an a northern de­ciduous fo~r in Indiana. J. Mcd. ent. 18(2)· 116 12.2.

30. Noyes. 1989. A srudy of five methods of samplina Hymenop­tera (lns«u) in a aopical ralnfores~, with lpCcial rmrmce to the parasites. J. oar. hist. 23:2~-298.

Jl. Ohman C P. Stcwan L G. Thomas] R. 1983. PhytophiiJOUS insect communities in the canopies of three EuCJtluplM! Cores~ rvpa in south·c-•srern Australia. Ausr. J. Ecol 8:39,·403.

32. Paannann W, Stork N. 1987. Canopy f<llll!inl!. a method of coUecti113 living insects for investigations of ltfe history W'ltC8ics. J . nat. lUst. 21:,63-'66.

33. Pc!enoo A. 1934. A manual of entomolotJcal equipmmt and mahods. Pt. l Edward B-. Inc. Ann Azbor p

J.l Rat~J A. de Sousa Gora)-eb l1982. T~ (Diptcra) da ArnazocU.: l Uma oova armachlha SUSfiC0$8 e pnmciros rq· btroa de muruas de copu de arwrra. Acta Amazonia. 1.2( 1):23.2-236.

" · Roberts HR. 1973. Arboreal Orthoptera in the rain forrst of Cosra Rica collected with insecricide: a rrport on the gru· shoppers (Acrididae) includinl! new species. Proc. Acad. Nat. Sci. Pbila. 12':46-66.

36. Selman B J, Lowman M D. 1983. ~ bioqy and herbivory rates oC Nowaslrol nothc/•&t Selman (Coleoptera: Chry­somebdac), a new genus and species on Notho/•1111-"in AuSiffitan temperate rain forrsrs. Aust. J. Zoot 31:179-191.

37. School P J , DeFolian G R. Nemenyi P B. 1979. Vertical dis­lributloo of bitinc activity by A~ tniffl41M1. Ann. Eot. Soc. Am. 72:.SJ7-,39.

38. Snow W E. 1~'· F-ru. activities of ~ blood-auclrini Dipcera with rrfcrence to vertical distribution in bonorn bnd forcsu. Ann. Eot. Soc. Am. 48:512-,21.

39. --- 197,. ~vertical disttibutioo oC flying mosquuoes (Dip«cra, Culicidae) in W CSl Africa savanna. Bu.1L E111. Res. 63:269·.276.

40. Southwood T R E. 1966. Ecoqical Methods. Methuen and Co., Ltd. London. 391 pp.

41. ---, Moran V C, Kennedy C EJ. 1982a. The richness, abundance, and biomass of the arthropod communities on trees. ] . Anim. Ecol. ":6},~0.

42. ---. ---, ---. 1982b. The assessment of arboreal ansecr fauna: comparisoos of knockdown samplina and faunal lists. Ecol. Ent. 7:JJI -340.

43. Stork N. 1987L Arthropod CawW similariry of 8orneao I'IUI

CorCSlrrea. Ecol -- 12:219-226. 44.--- 1987b. Guild sauaun: of anhropocls from Bomean

rain f'orCSl trees. Ecol cnl. 12:69-80. 4'. --- 1988. Insect diversity: faets, ficrion, and specular ion.

Bioi.). Linn. Soc. " :J21-JJ7. 46. Taylor l R. 1~1. An improved suction trap for inSeCtS. Ann.

Appl Bioi. 38:,82-m. 47. Teuloo D A J, Penman 0 R. .. Vcnical stratifications of atidty

boerd c:atcncs of kalhopper edulu (Hocnoptcra: Cic:adelli­droe) within apple orchards. New Zealand Entomoqist. 9:100..103.

48. Vergheae A P, Taodoo L, Pradala Rao G S. 1988. Sparial dis­tribution pattern and sampling plan for the Blisu~r Midge, Erosomy;. tnJia Grover (C«idomyiidae: Dip«cra) in lndiL Insca Sci Applic:. 9(4)·'1'·'18.

49. Wanjala F M E. 1978. RdMiYc abuodaDtt and widUn canopy clisuibudoo oC the p&nlliles oC the Coftee Leafminu, ~ t~ •"P*i [Lep.: L~] in Kmya. Entomophap lJ:H-63.

,0. Watanabe H. 1983. Efferu of rrpeated aerial appbcarlons of inscctlddcs for pine-wilt discuc oo arboreal arthropods in a fine stand. J . Jap. For. Soc. ~(8):282-287.

'I. Watanabe H, Rua)'IOOII8ncm S. 1989. Estilmtion of arboreal arthropod density in a dry evergreen forest in Northeastern Thailand. ] . Trop. EcoU:"1-"8.

,2. Wolda H. 1979. Abundance and diver'Sity o( Homopcera in the canopv ofo tropical forcst. Ecol. Ent. 4:181-190

" · Hanesvdt R J, Harvey H T, Sbdllwnmtt H S, Stecker R E. J9n. Giani Sequoia~ of tM Sierra Neveda. National Park Service Sdeoti6c Morqnph no. 12. U.S. Prindn« Office, Wllhiaccon D.C., 180 pp.