TRAP CATCHES OF IN THE RESERVA DUCKE, AMAZONAS

T. R. New ( * )

Abstract yield many additional species or, at least, a different relative abundance of species. Psocoptera captured in emergence traps, However, psocids are commonly captured light traps and a Malaise trap at the Reserva in suction traps in temperate regions, and Ducke in 1977-78 are enumerated and discus­ some are known to be attracted to light. sed. Eighty four species were trapped, mainly in small numbers, and the few common spe­ Wolda (1977) has indicated that light tra­ cies were present throughout the trapping pe­ ps may be used only with severe limita­ riod. Fewer species were caught in emergence tions to indicate the evenness component traps than in other trap types. Distribution of of tropical homopteran diversity, and si­ Brazilian psocids is briefly discussed. milar strictures apply for interpretation of catches of Psocoptera. Wide fluctua­ INTRODUCTION tions between successive trap catches may be expected as a result of daily climatic The diversity and ecology of Pso­ fluctuations, and a sampling programme coptera, a group of small which with only one sample a week (or similar feed predominantly on microepiphytes interval) may thus reflect many such and organic debris on plant surfaces, has short-term climatic variations rather than been little studied in the tropics. In Bra­ phenological change. Such catches must zil, although many Psocoptera have been be interpreted with considerable caution. described from the country, the only su­ Similar restrictions apply to Malaise traps. bstantial biological data are derived from Emergence trap catches may not be in­ a seven month study of a small area of fluenced by as many external factors, but the Mato Grosso (New, 1973a). The Re­ psocid activity may be affected by, for serva Ducke samples which form the sub­ examples, soil waterlogging and disper­ ject of this account are the first to provi­ sion of substrate litter. de more than sporadic data from an Ama­ With these reservations, a simplistic zonian rain-forest. They were taken over appraisal of the trap catches may be ma­ a longer period than the Mato Grosso de. The of the collection is samples, and give, a framework for attem­ treated in part by New (1979, 1980). pting to indicate psocid diversity and ha­ bitat limitation near Manaus. RESULTS AND DISCUSSION The traps and the trapping regime are described in Penny & Arias (1982). 1760 psocids, representing 84 spe­ The three trapping methods (here desig­ cies, were collected. Few species were nated 'emergence traps', 'light traps' and common, and many were represented on­ 'Malaise trap') are inherently likely to ly by single individuals. Emergence traps capture rather different sorts of psocids yielded fewer species and individuals than and, as it is commonly assumed that ma­ the other trap types (Table 1). Table 2 in­ ny psocids do not take to fligtht readily, dicates that there was low species overlap it is likely that beating vegetation would between emergence traps and either other

(*) Department of Zoology, La Trobe University, Bundoora, Victoria 3083, Australia.

ACTA AMAZONICA 13(3-4): 591-599. 1983 — 591 TABLE I. Total numbers of Psocoptera collected in three trap types, Reserva Ducke, September 1977 - September 1978.

Trap No. individuals (N) No. species (S) N/S

Emergence 371 13 28.5 Light 662 56 11.8 Malaise 727 46 15.8

Total 1760 84 21.0

TABLE 2. Species overlap of Psocoptera in different types of traps, Reserve Ducke, September 1977 - September 1978.

No. species in

1 trap type 2 trap types 3 trap types Emergence 3

• " Light 31 4 3 Malaise 22

trap type, and only 3 species were collec­ of related species in most cases) treat the ted in all three trap types. There was con­ above families in sequence. siderably higher species overlap between . Members of this light and Malaise traps. Temporal family are commonly associated with leaf distribution of catches is indicated in Fig. litter, or with dead foliage on trees, and 1. The graphs indicate a strong tendency some are known to fly readily. The spe­ for larger catches to include a greater cies captured are 2 of Echmepteryx number of species; peaks for the different Aaron, one of them very close to E. (Thy- trap types do not markedly coincide. lacopsis) bishopi New and the other Captures in the Malaise trap were relati­ (found only in the Malaise trap) undescri­ vely high from February to April and from late July to September. Light trap bed, and 4 individuals of Soa violácea catches were highest in November and New. Both identified species were descri­ December, with some secondary peaks at bed from series from forest litter in the other times, and emergence trap catches Mato Grosso, but one of the type series were very low after May. Grossly, then, of S. violácea was from a water trap, pos­ 'seasonal' trends for the bulked Psocop­ sibly indicating a tendency to disperse. tera appear to depend on trapping me­ . Two specimens of thod. Psocatropos microps (Enderlein), a bra- Table 3 is a summary of the fami­ chypterous and presumed flightless spe­ lies captured in each trap. Undescribed cies, were captured in emergence traps. It species are included in the species totals. is extremely unlikely that members of The following notes on richness and bio­ this genus would be captured in the other logy (necessarily inferred from knowledge trap types. The species is very widely dis- TABLE 3. Numbers of species of each represented family of Psocoptera caught in each trap type, Reserve Ducke, September 1977 — September 1978.

Trap

Family Emergence Light Malaise Total spp

Lepidopsocidae 1 1 3 3 Psyllipsocidae 1 — 1 Musapsocidae -1 — 1 Liposcelidae 1 - — 1 1 - — 1 - 1 1 1 - 6 5 11 Dolabeilopsocidae 2 2 1 2 2 15 5 19 Caeciliidae 2 6 9 10 Polypsocidae 1 3 3 - 1 1 2 1 - 1 1 — 1 1 1 - 1 2 2 — 2 1 2 - 1 1 1 — 1 — 1 1 4 1 4 — 12 11 17

Total spp 13 56 46 84 Total families 10 16 15 20

tributed, and associated with litter and Pachytroctidae. Five individuals, 3 of domestic habitats. nymphs, of an undescribed species of Musapsocidae. The single specimen Pachytroctes Enderlein were collected captured represents a genus (Musapsocus from emergence traps. Pachytroctidae are Mockford) not previously recorded from common in Brazil, but most species are Brazil. Little is known of the biology or undescribed: many are associated with habitat preferences of members of this litter or dead foliage. genus, but types of several species were Amphientomidae. All 12 individuals collected from dead foliage (Mockford, captured were adults of Seopsocus acu- 1967). minatus Roesler, a species apparently not Liposcelidae. Adults and nymphs of recorded since its description from mate­ one species of Liposcelis Motschulsky, rial collected in Santa Catherina. Some of a large cosmopolitan genus of apterous the type series were collected from bark psocids, were collected. Many species are (Roesler, 1940), and many Amphientomi­ known to breed continuously in litter, dae elsewhere are known to be cortico- under bark or in domestic habitats. lous. Ptiloneuridae. The taxonomy of the lopsocus ctenatus (New) was trapped in Ptiloneuridae, and of the other Epipsoce- only small numbers. tae collected, is treated by New (1980). Epipsocidae. More species of Epipso- Most Ptiloneuridae were collected only in cidae were captured than of any other very small numbers, many being represen­ family, and only 2 of the 19 species were ted by single specimens. It is known that earlier known (from other parts of Bra­ males of some Ptiloneuridae are attracted zil). Few species were common and the li­ to light (New 1972a), and that some spe­ mited data available on Epipsocidae in cies are nocturnal. Little is known of Brazil suggest that many are predominan­ their biology, but their general absence tly bark-frequenters. Most were captured from litter suggests that they are cortico- only at light, but some appear not to be lous: New (1972a) suggested that their attracted to light. 53 specimens of Neu- low incidence in beating samples in the rostigma xanthoptera, for example, were Mato Grosso may have been due to a ha­ captured in the Malaise trap, and none in bit of secreting themselves in crevices du­ the light traps. ring the day. The family is very diverse in Caeciliidae. Three individuals of Dyp- Brazil. socus inka (Enderlein) were taken in the . One of the 2 spe­ Malaise trap. The remaining 9 species are cies collected, a new species of Isthmop- all referable to Caecilius Curtis s.l., and socus Eertmoed, was the most abundant several are known from elsewhere in Bra­ psocid collected during this survey (Ta­ zil. Most Caeciliidae can be regarded as ble 4): it comprised more than 80 per foliage-frequenting psocids, but some spe­ cent of the emergence trap catches, and cies are clearly associated with trees and was common also in the other traps. Few others with low vegetation. Four of the nymphs of the family were found, but present species, including both those the abundance of Isthmopsocus in emer­ from emergence traps, are related to grou­ gence traps indicates that it breeds in ps usually found on low vegetation. ground litter. The other species, Dolabel Polypsocidae. The 3 species captured

TABLE 4. Percentages constituted by selected psocopteran taxa in trap catches, Reserva Ducke, September 1977 - September 1978.

Percentage of total catch in Taxon Emergence Light Malaise

Liposcelis sp 7.5 — — Seopsocus acuminatus Roesler ~ 1-2 3.6 Isthmopsocus luridus New 80-8 36.4 25.6 Neurostigma xanthoptera New 2.1 - 7.3 Polypsocus sp. A - < 0.1 5.5 Peripsocus sp. A - 2.8 < 0.4 Scytopsocus sp. A - 2.4 — Lophopterygella cincta New 0.3 3.2 0.1 Blaste sp. A - 2.2 5.7 Blastesp. B - 1.2 2.5 Cerastipsocus sp. A - < 0.1 7.6 Cerastipsocus sp. B - < 0.2 4.1 Thyrsophoririae (3spp.) — — 2.1 Trichadenotecnum continuatum Roesler — 2.0 — are Polypsocus selenius Roesler (Brazil), appears that they are not markedly P. griseolineatus (Enderlein, hitherto attracted to light. Polypsocus spp. are ge­ known only from Peru), and P. sp. n. (A), nerally foliage — frequenting psocids. and a total of 61 individuals were trapped Ectopsocidae. The two species, to­ - 58 of them In Malaise trans. It thus gether represented by only 3 individuals

Emergence

Light

Malaise

\ »

1977 1978 Date

Fig. 1. Numbers of species and individuals of Psocoptera captured in traps at the Reserva Ducke, 1977 - 1978, on each trapping occasion. ( , no. individuals; o—o , no. species). are presumed to be frequenters of dead Malaise trap, as were a high proportion of foliage. Species of Ectopsocus McLachlan Cerastipsocinae, and most other taxa we­ appear to be generally scarce in tropical re relatively scarce in the light trap. Ho­ South America, and few species have wever, 2 species of Trichadenotecnum been recorded. Enderlein s.1. and 2 of Steleops Enderlein Archipsocidae. The family was re­ were collected in the light trap, and these presented by 5 individuals of Archipsocus genera were absent from the Malaise trap minutiilus New. The scarcity of Archipso­ catches. cidae is surprising, as the family is extre­ Major differences in captures betwe­ mely well-represented in Brazil (New en the 3 trap types are summarised in Ta­ 1973b, Badonnel 1978), with 22 species ble 4, in which all the more abundant ta­ recorded. Most, if not all, species are co­ xa are included. In the absence of direct lonial and live under silken webs on bark, samples from vegetation over the same but some are known only from-litter. period, only limited appraisal of these dif­ Lachesillidae. Twenty-two individua­ ferences is wise. Clearly the litter psocid ls of 1 species of Lachesilla Westwood, 18 assemblage is less diverse than that found of them in the Malaise trap. Several Bra­ on vegetation, and contains a number of zilian Lachesilla spp are known to be taxa not amenable to capture by the associated with dead foliage. other trap types. Several species appear to Peripsocidae. One species of Peri pso­ be attracted to light more than others, ra* Hagen represented by 22 individuals (19 but absences from a light trap may also in light traps), and a singleton of a second reflect differences in diurnal activity species. The genus is not well represented peaks rather than just attraction. in South America, and it is presumed that The catches show substantial varia­ most species present are bark-frequenters. tions in N/S ratios over the trapping pe­ Pseudocaeciliidae.Sixteen individuals riod (Table 5), together with considerable of a species of Scytopsocui Roesler were differences in numbers of species captu­ taken in light traps, together with 5 of red in different months (Fig. 2). Clear Pseudocaecilius sp.. A further Pseudocae- seasonal trends are not evident, but most eilius was captured in the Malaise trap. species were captured in numbers too Habitats of these taxa are unknown, but small to reveal any such trends present. are likely to be dead foliage or bark. The most abundant species (I. luridus New) occurred throughout the period Elipsocidae. Three individuals of an and, in common with many litter psocids, undescribed species of Nepiomorpha Pe- appears to breed continuously. The other arman. Elipsocidae, again, are scarce in S. more common species (N. xanthoptera, America, and most appear to be bark-fre­ Polypsocus A, Blaste A, Cerastipiocus A, quenters. B) also occurred over much of the trap­ Philotarsidae. Six specimens of Aaro- ping period. nietta recta New and Thornton from the light traps. Daily records of rainfall, temperatu­ Myopsocidae. Thirty one specimens re, insolation and relative humidity were of 4 new species (New 1979) taken in examined, but no obvious correlation bet­ iigth traps, and singletons of the most ween climate and numbers of Psocoptera common species in the other trap types. caught were found. The largest Malaise Probably bark-frequenters. trap catches occurred after weeks with Psocidae. This family is very diverse rainfall ranging from about 30 mm to in Brazil, and most species appear to be about 103 mm, with 0-2 fine days, and associated with bark of trees. Thyrsopho- the smallest catches immediately follo­ rinae (3 spp.) were collected only in the wing a peak also sometimes occurred TABLE 5. N/S ratio for bulked monthly Catches of Psocoptera, Reserva Ducke, 1977- 1978.

Month 1977 1978 Trap IX X XI XII I II III IV V VI VII VIII IX Emergence 4.5 11.4 18.7 12.2 15.5 1.7 24 8.2 4 1.5 1.2 1 Light 2.4 2.6 7.2 6.1 3.2 2.2 4.3 1.7 2.7 1.9 2.3 2.6 5.1 Malaise 2.4 1.7 3.1 2.8 1.7 3.9 4.4 4.8 2.8 3.1 2.6 5.2 4.3

after weeks with considerable variation in ples, but also contained a high proportion precipitation (rainfall ca. 13-151 mm, (ca 90/137 (115 'forest' species) or 66 fine days 1 - 3), with similar very small percent) of new species. temperature variations. Only 3 of the 13 emergence trap spe­ The light traps, set to capture over 1 cies are possible 'primary litter frequen­ night per week, also gave inconclusive da­ ters' (New 1969) - Psocatropos, Lipos- ta on correlation between catches and celis and Pachytroctes, and the others are 'weather'. The three highest catches all likely to be mainly casually present after occurred on nights following a day wi­ dropping on/from vegetation. A possible thout rain or with little (1.7 mm) rain, exception is I. luridus, the large numbers but many similar days preceded low trap of which suggest a more regular associa­ catches. No correlation with temperature tion with litter. However, as it was also was evident, but median temperatures va­ common elsewhere,and very few nym­ ries over only a narrow range in the Ma- phs were collected, it is probably a 'gene­ naus area. Data on wind speed, lunar pha­ ral dead-vegetation' frequenting species. ses and cloud cover are not available. As in the.Mato Grosso (New 1973a), the Elton (1973, 1975) has indicated that samples reveal a greater number of pro­ population densities of many Neotropical bable 'bark-frequenters' than 'foliage-fre­ rain forest insects are very low, and deri­ quenters'. Including presumed dead-folia­ ved his conclusions from counts from the ge-frequenting forms, as well as groups field layer and low to middle canopy, the more usually associated with living folia­ region 'sampled' by the present traps. Al­ ge, about 3 times as many bark-frequen­ though the trap catches are selective, it is ters as foliage-frequenters were captured. clear that many psocid species may be The greater number (104/116) of forest scarce. Clearly, there are many species in bark frequenters from the Mato Grosso the area and the high proportion of un­ samples reflects, in part, (a) the greater described species (probably 52/84 or 62 number of Archipsocidae and (b) a con­ percent) suggests that exhaustive collec­ siderable diversity of Liposcelidae, which ting would reveal many more taxa. It is are not captured by traps such as those further likely that many are of very limi­ used in the Reserva Ducke. ted distribution: although several are simi­ Nearly all families were found in the lar to (or identical with) species described two surveys, and (neglecting Liposcelidae, from Peru, relatively few of the Manaus Pachytroctidae and Archipsocidae) simi­ species were found during an intensive lar relative richness levels were found. In survey of part of the Mato Grosso (New, both, were abundant, althou­ 1973a). The latter collection contained gh Ptiloneuridae were more diverse in the many taxa not found in the present sam­ Reserva Ducke than in the Mato Grosso SON D J F MAMJ J A S 1977 1978

Fig. 2. Numbers of species of Psocoptera captured in traps each month at the Re- serva Ducke, 1977-1978. (•—• , emergence traps; #—# , light traps; O—o , Malaise trap)

(11 spp. cf 4 spp.). The subgenus Poeci- gested several phenological traits in the lopsocus of Thyrsopsocus Enderlein (3 Mato Grosso Psocoptera but the present spp) were the only Thyrsophorinae in the samples do not permit meaningful aug­ Mato Grosso samples, whereas the present mentation of these — due in part to the samples include representatives of three selective nature of the trapping and the genera/subgenera: it appears that the low abundance of most species. Probably, Thyrsophorinae are a predominantly 'tro­ however, the greater climatic constancy pical forest' group. Many other genera of of the Manaus region obscures phenologi­ Psocidae are widely distributed in tropical cal traits which could be revealed by se­ Brazil. Seopsocus acuminatus was not col­ ries of beating samples in which psocids lected in the Mato Grosso, and may also are more clearly associated with particu­ be a 'forest' species. lar habitats. It is inferred that many of the species may breed throughout the However, conclusions on distribu­ year, as indicated by their presence in tion of Psocoptera in tropical South Ame­ traps at intervals throughout the whoJe rica must be very tentative. Many areas of the trapping period. have not been surveyed for Psocoptera and, although those of the temperate ACKNOWLEDGMENTS south of South America are now tolera­ bly well-known (New & Thornton s.d.), I am very grateful to Drs Jorge Arias the high proportion of new taxa found in and Norman D. Penny for allowing me the present collection is usual in collec­ to study Psocoptera from the Reserva tions from the tropics. New (1973a) sug­ Ducke. Resumo 1972a. Some Epipsocetae (Psocoptera) from central Brazil. Trans. R. ent. Soc. São enumerados e discutidos os Psocopte­ Lond., 123 (1971): 455-497. ra, capturados em armadilhas de eclosão, ar­ 1972b. Some Thyrsophoridae from madilhas de luz e uma armadilha de Malaise, central Brazil (Psocoptera). Zool. Anz., na Reserva Ducke, durante os anos de 1977 - Leipzig, 189:190-199 1978. Oitenta e quatro espécies foram captu­ 1973a. Local distribution of Psocopte­ radas em número reduzido, e as espécies co­ ra in the Mato Grosso, central Brazil. Pap. muns, poucas estiveram presente todo o pe­ Av. Zool., S. Paulo, 27:115-144. ríodo de coleta. Com as armadilhas de eclosão, 1973b. The Archipsocidae of South menos espécies foram coletadas do que com America (Psocoptera). Trans. R. ent. Soc. qualquer outro tipo de armadilhas. As distri­ Lond., 125:57-105. buições geográficas dos psocôpteros brasilei­ 1979 New and little-known Psocoptera ros são discutidos resumidamente. from the Reserva Ducke, Amazonas. Acta Amazonica, 9(4) : 773-781. References 1980 Epipsocetae (Psocoptera) from the Reserva Ducke, Amazonas. Acta Badonnel, A. - 1978. Complements a 1'étude Amazonica, 10(1): 179-206. des Archipsocidae du Brésil (Insecta, Pso- New, T. R. & Thornton, I. W. B. - s. d. Pso­ coptera). Rev. Brasil. Biol., 38: 177-186. coptera from central and southern Chile. Elton, C. S. - 1973. The structure of inverte­ Pacific Insects. brate populations inside Neotropical rain Penny, N.D. & Arias, J. R. - 1982. Insects of forest. J. Anim. Ecol., 42: 55-104. and Amazon Forest. New York, Columbia 1975. Conservation and the low po­ University press. 269p. pulation density of invertebrates inside Roesler, R. — 1940. Neue und wenig bekannte Neotropical rain forest. Biol. Conserv., Copeognathengattungen. I. Zool. Anz., 7:3-15 Leipzig, 129:225-243. Mockford, E. L. - 1967. The Electrentomoid Wolda, H. — 1977. Flutuations in abundance Psocids (Psocoptera). Psyche, Camb., Mass., of some Homoptera in a Neotropical fo­ 74:118-165 rest. Geo. Eco. Trop., 3: 229-257. New, T.R. -1969. Observations on the biolo­ gy of Psocoptera found in leaf litter in Southern England. Trans. Soc. Brit. Ent, 18:169-180 (Aceito para publicação em 6/7/83).