Diversity of Curculionoidea in Humid Rain Forest Canopies of Borneo : A Taxonomic Blank Spot

From 1992 to 2009, 334 trees were sampled by insecticidal knockdown on Borneo, Malaysia. Here, we describe the taxonomic composition of the 9671 specimens and 1589 species Curculionoidea collected (with additional notes on Cerambycidae). We found a largely unknown fauna with an assumed proportion of over 80% of species new to science, including all 33 Apionidae and 26 Ceutorhynchinae species. Specialists could usually identify only a few specimens leaving the remaining beetles for further investigation. The samples contain numerous genera, two tribes (Egriini, Viticiini), one subfamily (Mesoptiliinae) and one family (Belidae) new to Borneo and several genera not recorded west of the Wallace line before. These data show how little is known about canopy diversity. The lack of taxonomic knowledge implies a respective lack of autecological knowledge and is alarming. Some taxa differed conspicuously between primary and disturbed forests. In contrast to common literature, our results let us conclude that current efforts to narrow down the extent of tropical diversity and its ecological importance must consider the enormous species diversity of the canopy.


Introduction
Despite all efforts, the global extent of biodiversity is still insufficiently known and controversial [1,2].This lack is mostly due to poor knowledge of tropical rain forests where large areas and habitats like the canopy remain insufficiently researched.Although scientists have known for decades that arthropod richness is very high in the canopy of tropical trees [3,4], no comprehensive data or long-term studies have yet been published from the canopy of any near-equator rain forest worldwide.Considering the importance of biodiversity for ecosystem function and services [5] research in tropical rain forests-and from our perspective in the canopy-should be prioritized.Primary forests are ruthlessly exploited worldwide and pushed back to a few wildlife sanctuaries [6].Evidence is growing that disturbed forests differ fundamentally from primary forests not only in respect to species diversity but also to system services on which humans depend [7].Understanding the fundamental processes requires working in pristine forests where most species are usually new to science [8][9][10], which makes research even more difficult.Besides preliminary work on morphospecies sorting only the exact identification of this largely unknown diversity ensures advanced future ecological analyses.Moreover, the lack of taxonomic knowledge causes an almost general ignorance about the autecological requirements of species and makes predictions concerning the future of disturbed forests rather imprecise.
There is an urgent need for well processed data documenting the distribution of biodiversity within and between forest types.Our work narrows this lack of knowledge by analyzing the faunistic and taxonomic composition of arboreal beetles which have been collected from 334 trees in primary and disturbed forests of southeast Asia from 1992 to 2009.Here, we present data on the Curculionoidea, one of the most diverse and abundant group of beetles [11].We provide insight into the faunistic-taxonomic composition of the arboreal weevil fauna and report some striking differences in taxa abundance between primary and disturbed forests.

Study Area
All fogging sites are located in Sabah, the northeastern part of East Malaysia, Borneo.Research was carried out in different primary and disturbed forests.Most data were gathered in primary "Mixed Dipterocarpaceae lowland forests" between 300 m and 900 m a.s.l.mainly in different substations of the Kinabalu National Park, Malaysia (06 • 02 54.18 N, 116 • 41 56.34 E) but also on Gaya Island (6 • 0 51.36 N, 116 • 1 13.74 E) and in the National Parks Crocker Range (5 • 24 43.68 N, 116 • 5 24.84 E) and Tawau (4 • 24 1.26 N, 117 • 53 24.66 E) as the most southern area.There were in total 171 trees.In addition, 43 trees were sampled at different altitudes ranging from 1000 m to 2500 m a.s.l. at Mt. Kinabalu and in the Crocker Range.Different types of disturbed forest were also sampled: 10-to 15-year-old pioneer vegetation on the two small islands Bakkungan Kecil (6 • 10 0.66 N, 118 • 6 32.88 E) and Selingaan (6 • 10 31.38 N, 118 • 3 41.10E)-10 trees together.Three secondary forests of 5, 15 and 40 years after natural regrowth which were growing close to the Kinabalu Park (6 • 19 0.84 N, 116 • 44 2.1997 E) and another three isolated secondary forests of 10, 20 and 50 years after natural regrowth in a distance of at least 10 km to the Crocker Range primary forest (5 • 26 0.68 N, 116 • 7 58.78E).Altogether, these were 80 trees.We also sampled 15 oil palms that were growing in direct neighborhood to the primary forest of Tawau and 15 fruit trees of different species in garden areas (6 • 17 36.84N, 116 • 43 5.76 E) in the Kinabalu region.All disturbed forests were found in areas originally covered with lowland forest (between 5 m and 400 m a.s.l.).More details including a map have been published in [9].

Collecting and Sorting of Canopy Beetles, Genus and Species Identification
During 1992 and 2009, arthropods were collected by means of insecticidal knock-down ('fogging').Natural pyrethrum diluted in a highly purified white oil was used as insecticide because it degrades within hours in sunlight leaving no poisonous substances in the trees.Fogging was carried out in the early morning or late in the afternoon when there was little wind.All arthropods that dropped into the collecting sheets installed beneath a tree crown after two hours following fogging were transferred into vials filled with ethanol.More technical information has been published elsewhere [12].Here, we focus on weevils following preferably the classification proposed by Alonso-Zarazaga and Lyal [13], with slight modifications; see Oberprieler [14].We are aware that there are also several changes at the subfamily level in between, but did not follow these proposals for practical reasons.Beetles were sorted to morphotypes (morphospecies) and specimens were distinguished by optical means.Morphotypes were treated as real species when sorted by taxa specialists.A dissection of the genitalia was usually not carried out and left to specialists.
Besides specialist knowledge, current taxonomic papers and revisions were used for species determination.Furthermore, the SDEI museum collections in Eberswalde (now Müncheberg) and Dresden were helpful, especially for Conoderinae, Dryophthoridae and Entiminae.

Suitability of Fogging for Collecting Canopy Weevils
From all 334 trees, we collected 9671 individual Curculionoidea in 1589 species.Due to the high number of species from the groups Curculionini, Ochyromerini and Rhynchitidae, as well as smaller groups like Ceutorhynchinae, Nanophyidae or Rhamphini, the data show that canopy fogging is well-suited to collect arboreal species that inhabit leaves, flowers and fruits of trees.This seems to hold also for several groups that are usually regarded as saproxylic, like Anthribidae, Conoderinae, Cossoninae, Cryptorhynchinae, Molytinae and Scolytinae which were also represented by more than 100 species each.Lower numbers than expected were found in Attelabidae with one species apart from genus Euops and Dryophthoridae (15 species in each family).

Faunistic-Taxonomic Classification of Canopy Beetles
The fogging results are presented in Table 1.It contains all relevant taxonomic weevil groups found in the samples, the number of species and specimens, the experts included in the analyses and the number of identified genera and species before 2013.

Belidae-Oxycoryninae
There is one specimen of the genus Metrioxena in the sense of Alonso-Zarazaga and Lyal [13].According to Legalov [16], it is now placed in Vladimirixena (Figure 1).Metrioxena is distributed in Malaysia and Indonesia, and Vladimirexena in Indonesia.Metrioxena sensu lato was also detected in Baltic Amber indicating a minimum age of 35 million years.It is the first record on Borneo, also of the entire family, even though its occurrence was to be expected there.

Attelabidae
In our data, we found one Allolabus species and 14 of the genus Euops, mainly subgenus Suniops (Figure 1).Five morphospecies were identified to species level but all are in need of confirmation by the study of type specimens.Thus, at least 10 species are undescribed.

Apionidae
Our samples contain 33 species of which 21 are in the tribe Piezotrachelini, 11 in the tribe Ixapiini and 1 in the Rhadinocybini.None of the specimens could be identified to species and thus all of them may be new to science (Wanat, pers. comm.).Only three genera could be identified: Microconapion, Ommatocybus and Piezaplemonus with one species each (Figure 1).The last genus, described by Wanat [17], was hitherto known only from Thailand and the genus Microconapion from Japan, Taiwan and Vietnam.Ommatocybus was erected by Wanat [18] during his study on Australo-Pacific Apionidae.This means that all the genera were not previously recorded from Borneo and Malaysia, and this is also true for the tribe Rhadinocybini (see Alonso-Zarazaga and Lyal [13]).According to Wanat, the remaining 30 species represent four as yet undescribed genera.At least these data show that the Bornean Apionidae canopy fauna is completely unexplored.

Brentidae
There are 32 species and, apart from the circumtropical Cylas formicarius (Fabricius, 1798), eight of them were identified to the species level.A further species was described as new, Microtrachelizus floreni Mantilleri, 2012 [19], and 11 species were identified to a genus level of which three will be described later (Cordus sp., Homophylus sp., Hypomiolispa sp.) (Figure 2).The samples contained 10 species, mainly of the subfamily Cyphagoginae, which remained unidentified (without genus diagnosis).The genus Cordus was up to now only known from Australasia (Australia, New Guinea, Solomon Islands) and the species Calodromus insignis (Senna, 1895) only from Sumatra (Figure 3).The checklist of Brentidae includes 204 species from Borneo including three mentioned species known only from "Indonesia" without further location [20].From all species fogged, only nine (28%) could be determined to the species level, representing 4.4% of the known Bornean fauna of Brentidae.If the difficult genus Cyphagogus is excluded, the values change to 32% and 4.9%, respectively.The data demonstrate that, within a group for which all available data are summarized at the present time (see the world catalogue of Sforzi and Bartolozzi [20]), the tree crowns apparently harbor a rather little known Brentidae fauna on Borneo.
Diversity 2018, 10, x FOR PEER REVIEW 5 of 20 The checklist of Brentidae includes 204 species from Borneo including three mentioned species known only from "Indonesia" without further location [20].From all species fogged, only nine (28%) could be determined to the species level, representing 4.4% of the known Bornean fauna of Brentidae.If the difficult genus Cyphagogus is excluded, the values change to 32% and 4.9%, respectively.The data demonstrate that, within a group for which all available data are summarized at the present time (see the world catalogue of Sforzi and Bartolozzi [20]), the tree crowns apparently harbor a rather little known Brentidae fauna on Borneo.
The genera Centrinertus, Hollisiella, Lophobaris and Nespilobaris are new to Malaysia and Borneo (Figure 3).The first was formerly known only from the Philippines, the second from China, the third from Taiwan to Java and the Philippines, so that its occurrence on Borneo could be expected.This is also true for Nespilobaris, which was known from Eastern Siberia, Japan, China, Australia and Africa [13,22].The considerable number of Baridinae genera that had not been recorded from Borneo demonstrates the poor knowledge on the distribution of this group of weevils in Southeast Asia.

Curculionidae-Ceutorhynchinae
The morphospecies sorting revealed 24 species, but Enzo Colonnelli and Hiraku Yoshitake, who checked it in part, both separated one more species each, resulting in a total number of 26 Ceutorhynchinae species.The tribe Mecysmoderini is the main group, represented by 17 species, followed by Ceutorhynchini with five species, Hypohypurini with three and Egriini with one, a Megahypurus species (det.E. Colonnelli by photo).None of them could be identified to species level suggesting that all of them are new to science.Besides Coelioderes (one species) and Watanabesaruzo (two species), two genera just recently described by Yoshitake and Yamauchi [23] and Yoshitake and Ito [24], there are two Coeliosomus and one Cysmemoderes species (aff.C. gibbicollis (Hustache, 1925),
The genera Centrinertus, Hollisiella, Lophobaris and Nespilobaris are new to Malaysia and Borneo (Figure 3).The first was formerly known only from the Philippines, the second from China, the third from Taiwan to Java and the Philippines, so that its occurrence on Borneo could be expected.This is also true for Nespilobaris, which was known from Eastern Siberia, Japan, China, Australia and Africa [13,22].The considerable number of Baridinae genera that had not been recorded from Borneo demonstrates the poor knowledge on the distribution of this group of weevils in Southeast Asia.

Curculionidae-Ceutorhynchinae
The morphospecies sorting revealed 24 species, but Enzo Colonnelli and Hiraku Yoshitake, who checked it in part, both separated one more species each, resulting in a total number of 26 Ceutorhynchinae species.The tribe Mecysmoderini is the main group, represented by 17 species, followed by Ceutorhynchini with five species, Hypohypurini with three and Egriini with one, a Megahypurus species (det.E. Colonnelli by photo).None of them could be identified to species level suggesting that all of them are new to science.Besides Coelioderes (one species) and Watanabesaruzo (two species), two genera just recently described by Yoshitake and Yamauchi [23] and Yoshitake and Ito [24], there are two Coeliosomus and one Cysmemoderes species (aff.C. gibbicollis (Hustache, 1925), det.Yoshitake) and several species from further undescribed Mecysmoderini genera in the samples.The Ceutorhynchini are represented by Hainokisaruzo and the Hypohypurini by the single genus Hypohypurus.Unfortunately, Yoshitake, who was interested in receiving all Ceutorhynchinae weevils, has not described any to date.The insufficient knowledge about this group in Malaysia and Indonesia is shown by the fact that Watanabesaruzo was found until now only in Bali (Indonesia), Yunnan (China) and Perak (Malaysia) [23,25], Coelioderes in Eastern Siberia, Japan, Korea and China [24,26]) and Hainokisaruzo by several species in China, Taiwan, India and Japan [27,28].The tribe Hypohypurini, erected by Colonnelli in 2004 [26], was recorded from China and Vietnam, North Australia, Madagascar and Central Africa, and the tribe Egriini was also not reported from Malaysia or Borneo before 2012 [29].Megahypurus Korotyaev, 1989 was described from Vietnam (Figure 3).As previous collecting in Indonesia revealed a rather small number of Ceutorhynchinae (see, for example: [26]), fogging was very successful in collecting arboreal Ceutorhynchinae.In a current fogging study on Java, there were also a few Ceutorhynchinae of the genus Watanabesaruzo and other Mecysmoderini genera (A.Floren, own data).The insufficient knowledge about this group in Malaysia and Indonesia is shown by the fact that Watanabesaruzo was found until now only in Bali (Indonesia), Yunnan (China) and Perak (Malaysia) [23,25], Coelioderes in Eastern Siberia, Japan, Korea and China [24,26]) and Hainokisaruzo by several species in China, Taiwan, India and Japan [27,28].The tribe Hypohypurini, erected by Colonnelli in 2004 [26], was recorded from China and Vietnam, North Australia, Madagascar and Central Africa, and the tribe Egriini was also not reported from Malaysia or Borneo before 2012 [29].Megahypurus Korotyaev, 1989 was described from Vietnam (Figure 3).As previous collecting in Indonesia revealed a rather small number of Ceutorhynchinae (see, for example: [26]), fogging was very successful in collecting arboreal Ceutorhynchinae.In a current fogging study on Java, there were also a few Ceutorhynchinae of the genus Watanabesaruzo and other Mecysmoderini genera (A.Floren, own data).

Curculionidae-Curculioninae (Figures 4-7)
This group of weevils is represented by a great number of morphospecies in the samples.Due to the extensive work of mainly Japanese authors (see citations below), there is current information about the occurrence of some tribes in southeast Asia (especially Acalyptini, Anthonomini, Ochyromerini and Rhamphini).Here, we mainly present results for Acalyptini, Ochyromerini, Rhamphini and some species-poor tribes.
The following tribes could be identified (number of morphospecies in brackets): Acalyptini (31), Anoplini (6), Anthonomini (5), Curculionini (122), Demimaeini (3), Derelomini (1), Storeini (2), Rhamphini (52; see below), and Ochyromerini (previously Tychiini, subtribe Ochyromerina) (133).Morphotype differentiation in Acalyptini and in Ochyromerini is still a matter of some uncertainty because there is a rather large proportion of small yellowish weevils of great similarity (in Acalyptini, about two thirds of the total; Figures 4, 5 and 7).Hence, an analysis of the morphospecies numbers, including the dissection of the genitalia, could change the numbers.Usually, dissections were not made to avoid incidental destructions, as most species were present in only small numbers.Curculionidae-Curculioninae (Figures 4-7) This group of weevils is represented by a great number of morphospecies in the samples.Due to the extensive work of mainly Japanese authors (see citations below), there is current information about the occurrence of some tribes in southeast Asia (especially Acalyptini, Anthonomini, Ochyromerini and Rhamphini).Here, we mainly present results for Acalyptini, Ochyromerini, Rhamphini and some species-poor tribes.
The following tribes could be identified (number of morphospecies in brackets): Acalyptini (31), Anoplini (6), Anthonomini (5), Curculionini (122), Demimaeini (3), Derelomini (1), Storeini (2), Rhamphini (52; see below), and Ochyromerini (previously Tychiini, subtribe Ochyromerina) (133).Morphotype differentiation in Acalyptini and in Ochyromerini is still a matter of some uncertainty because there is a rather large proportion of small yellowish weevils of great similarity (in Acalyptini, about two thirds of the total; Figures 4, 5 and 7).Hence, an analysis of the morphospecies numbers, including the dissection of the genitalia, could change the numbers.Usually, dissections were not made to avoid incidental destructions, as most species were present in only small numbers.Forty-two species were determined to genus level: 8 Imachra, 2 Indodinorrhopalus, 2 Morimotonomizo, 23 Orchestes (Orchestes), 4 Orchestes (Nomizo), 1 Rhamphus (Trichorhamphus) and 2 Sphaerorchestes species.In the remaining species, the genus could not be determined without doubt.These species should represent new genera or new subgenera of Orchestes.In Orchestes (s.str.), there were many rather small species that were also assigned here, due to the hollowed metatibiae, the Curculionidae-Mesoptiliinae (Figure 7) Only one species (three specimens) in this subfamily has been collected.Apparently, it is a member of the tribe Laemosaccini, a tribe only known from Australia and New Zealand with the exception of the American genus Laemosaccus.This is the first record of a member of subfamily Mesoptiliinae on Borneo and west of the Wallace line.

Curculionidae-Scolytinae
These species were determined by Roger Beaver, e.g., [49].He distinguished 111 species that makes this taxon together with Anthribidae, Rhynchitidae, Conoderinae, Cossoninae, Curculionini, Cryptorhynchinae, Molytinae and Ochyromerini (each over 100 species), one of the commonest in the fogging data.From all 111 species, 50 could be identified to species level, the remainder at least to the genus level (two species with some doubt).The state of knowledge in this potentially The following genera could be identified by the following works: [30][31][32][33][34][35][36][37][38][39][40][41][42][43] and museum collections: Amorphoidea  Kojima (1997) from Sabah.Elaeidobius kamerunicus (Faust, 1898) from the tribe Derelomini is an introduced species inhabiting oil palms [41,44].In Curculionini, the genera Carponinus and Labaninus were identified provisionally.Most species with a tumid base of the rostrum could not be assigned with certainty to Indocurculio due to insufficient illustration of the key of Pelsue and O'Brien [42].The lack of well identified specimens which could be used for comparison is another serious obstacle.
In Acalyptini (Figure 4), there is a large proportion of species unidentified to genus.Most of them are very small, 1-2 mm long, and of a yellowish color.Apart from their small size, they resemble the Derelomini from palm flowers and may inhabit flowering trees.These species are not considered as members from the Acalyptini genus Derelomorphus because they have edentate femora.Two further species of a small size may represent the genus Niseida or a similar undescribed genus.Niseida is up to now restricted to Aru Island [41].These results are of special interest because there are the only data for two Acalyptini genera (with one species each), Amorphoidea and Eudela, recorded from Borneo [41].In the catalogue of Alonso-Zarazaga and Lyal [13], not a single genus of Acalyptini was listed for Borneo.This means that Parimera from our samples is another genus new to this island.
Three species of Anthonomus (Tachypterellus) have been recorded from Borneo [44].In our samples, there are also three, but, without dissection of the genitalia, we cannot state if they are identical.With the availability of the paper of Kojima and Idris [43], the identification of Usingerius parvidens Zimmerman, 1946 was enabled, a genus that we regarded at first erroneously as Ochyromerini.The former Entiminae tribe Viticiini, represented by two small distinct species, with the genus Viticis well characterized by the lack of tarsal claw segments [14], is now represented by two species.The genus is new to Borneo and at the same time the westernmost occurrence of this mainly Pacific genus (see [45]).These are the first records of the genus Viticis west of the Wallace line.
Despite current taxonomic work, the identification of Ochyromerini is difficult.Most of the many species are of small size (usually smaller than 3 or 4 mm, especially in Endaeus, Endaenidius, Eugryporrhynchus, Lepidimerodes and Morimotozo).Reliably identified species from other collections are lacking.The proportion of undescribed species is high.The identification of only three species is considered trustworthy: Eugryporrhynchus malayanus Kojima and Morimoto, 1995, Morimotozo ovipennis (Kojima and Morimoto, 1995) and M. rotundicollis (Kojima and Morimoto, 1995).However, a few species were also doubtfully assigned to this tribe: on first examination, three species looked like Cryptorhynchinae, but they have the abdominal segment sutures directed abruptly posteriad close to the margin.This is usually a good character to distinguish Ochyromerini from Anthonomini and Acalyptini.The number of weevils that could not be assigned with certainty to a higher taxon did not exceed 20 or 30 species and thus cannot really affect the data given in Table 1.(In some groups, for example, Anthribidae, Cryptorhynchinae, and in part Molytinae, an assignment to certain tribes proved to be impossible without extensive study and has been left to experts.)Curculionidae-Curculioninae-Rhamphini (Figure 6) This group was studied in detail by one of the authors (P.S.) based particularly on the papers on East Asian Rhamphini by Kojima [38], Morimoto [46], Kojima and Morimoto [47], and Morimoto and Miyakawa [48].The morphospecies sorting revealed 52 species.These are far more than the total 35 described and three undescribed Rhamphini species from the entire area of southeast Asia from India to South China (Fujian included), Taiwan and Thailand to Indonesia and the Philippines [47,48].
Forty-two species were determined to genus level: 8 Imachra, 2 Indodinorrhopalus, 2 Morimotonomizo, 23 Orchestes (Orchestes), 4 Orchestes (Nomizo), 1 Rhamphus (Trichorhamphus) and 2 Sphaerorchestes species.In the remaining species, the genus could not be determined without doubt.These species should represent new genera or new subgenera of Orchestes.In Orchestes (s.str.), there were many rather small species that were also assigned here, due to the hollowed metatibiae, the outer margin of hind femora with denticles, and the raised setae on the elytra and pronotum.The genus Indodinorrhopalus has not been recorded from Borneo or Malaysia previously, Morimotonomizo not from Borneo and Rhamphus only by an undescribed species from East Malaysia [49].All 52 species are rather small; there is not a single one measuring more than 3 mm.The remarkable southeast Asian Rhamphini genera Dinorhopala and Ixalma with body size usually >3.5 mm were not represented in the samples.
Only three species could be identified to species level: Imachra bifasciata Morimoto & Miyakawa, 1996, I. ruficollis Pascoe, 1874, and I. siamensis Morimoto & Miyakawa, 1996.I. ruficollis is widespread in the Oriental region, and its range is extended in one area just in the Palaearctic region (China: Fujian).I. siamensis was described from South Thailand, and I. bifasciata was re-discovered at the type locality (Kinabalu).Except for the three Imachra species and Morimotonomizo (with one species that could be identical with M. sphinxioides (Morimoto and Miyakawa, 1996) from West Malaysia), all remaining 48 Rhamphini species are very probably undescribed (see [47,48]).
Curculionidae-Mesoptiliinae (Figure 7) Only one species (three specimens) in this subfamily has been collected.Apparently, it is a member of the tribe Laemosaccini, a tribe only known from Australia and New Zealand with the exception of the American genus Laemosaccus.This is the first record of a member of subfamily Mesoptiliinae on Borneo and west of the Wallace line.

Curculionidae-Scolytinae
These species were determined by Roger Beaver, e.g., [49].He distinguished 111 species that makes this taxon together with Anthribidae, Rhynchitidae, Conoderinae, Cossoninae, Curculionini, Cryptorhynchinae, Molytinae and Ochyromerini (each over 100 species), one of the commonest in the fogging data.From all 111 species, 50 could be identified to species level, the remainder at least to the genus level (two species with some doubt).The state of knowledge in this potentially economically relevant group is therefore much better than in all other weevils (except Platypodinae, see the next section).However, even in Scolytinae, 61 species could not be identified and are very probably new to science.

Curculionidae-Platypodinae
Species in this group were sorted and identified by Roger Beaver.However, he received only 50% of the species.From these, 17 species, 14 genera and 11 species could be identified.Six species are new to science.

Remaining Weevil Groups
There is only very scattered information about all the remaining weevil groups (see Table 1): Anthribidae, Nanophyidae, Rhynchitidae (11 Auletini, 36 Deporaini and 63 Rhynchitini species) and the Curculionidae subfamilies Entiminae, Conoderinae, Cossoninae, Cryptorhynchinae and Molytinae.In Figure 8, some long-legged weevils from the subfamilies Conoderinae and Molytinae, which may be typical for tree crowns, are shown.In Nanophyidae, one species could be identified with the key of Lyal and Curran [50] as Damnux tenebriosa Lyal, 2003, but all others remained unidentified.
Diversity 2018, 10, x FOR PEER REVIEW 13 of 20 economically relevant group is therefore much better than in all other weevils (except Platypodinae, see the next section).However, even in Scolytinae, 61 species could not be identified and are very probably new to science.

Curculionidae-Platypodinae
Species in this group were sorted and identified by Roger Beaver.However, he received only 50% of the species.From these, 17 species, 14 genera and 11 species could be identified.Six species are new to science.

Remaining Weevil Groups
There is only very scattered information about all the remaining weevil groups (see Table 1): Anthribidae, Nanophyidae, Rhynchitidae (11 Auletini, 36 Deporaini and 63 Rhynchitini species) and the Curculionidae subfamilies Entiminae, Conoderinae, Cossoninae, Cryptorhynchinae and Molytinae.In Figure 8, some long-legged weevils from the subfamilies Conoderinae and Molytinae, which may be typical for tree crowns, are shown.In Nanophyidae, one species could be identified with the key of Lyal and Curran [50] as Damnux tenebriosa Lyal, 2003, but all others remained unidentified.The Dryophthoridae were represented by only four sizable specimens of the genera Aplotes, Cryptoderma and Laogenia, probably a species of Diocalandrini, and mainly by several Litosomini species of the genera Sitophilus (six species) and Myocalandra or a closely related genus (four species).In Entiminae, there are single Dermatodes (Dermatodini) and Eugnathus (Sitonini) species, some Ottistirini and many Cyphicerini, but there is also a rather large number of species of unidentified tribes (perhaps in part Myllocerini).Apocyrtidius chlorophanus Heller, 1908, is the only member of The Dryophthoridae were represented by only four sizable specimens of the genera Aplotes, Cryptoderma and Laogenia, probably a species of Diocalandrini, and mainly by several Litosomini species of the genera Sitophilus (six species) and Myocalandra or a closely related genus (four species).In Entiminae, there are single Dermatodes (Dermatodini) and Eugnathus (Sitonini) species, some Ottistirini and many Cyphicerini, but there is also a rather large number of species of unidentified tribes (perhaps in part Myllocerini).Apocyrtidius chlorophanus Heller, 1908, is the only member of Pachyrhynchini on Borneo, and it is represented by 47 specimens in samples from the Mesilau region of Kinabalu in the montane forest.
In Conoderinae (Figure 8), the bulk of the 171 morphospecies probably belong to Othippiini and a rather significant number to the usually black-and-white-coloured Menemachini, the remaining to Coryssomerini, Mecopini (s.l.), Campyloscelini and unidentified tribes.We are aware that the taxonomy of the higher tribes is rather uncertain especially in this subfamily [51].A few genera could be identified with little doubt by the keys of Hustache [52], the Intkey CD [53], and museum collections: Phaenomerus, Talimanus, Telephae and Tomicoproctus-and with some doubt: Agametis, Metialma, Odoacis, and Osphilia, but none of the probably largest tribe Othippiini, which has mainly small species.According to Alonso-Zarazaga and Lyal [13], the well-characterized genus Tomicoproctus, represented by one species, was not recorded from Borneo previously, only east of the Wallace line and in Africa.
In Molytinae, there was a great number of Trachodini (63 species), including two Pseuderodiscus species, previously placed in Curculioninae-Erodiscini, followed by Ithyporini (37), Hylobiini (16), Mecysolobini (14), Lithinini (about eight species, one of them identified as Seleuca by C. Lyal by photo) and Trigonocolini (1), but, among Ithyporini, there may also be other tribes with small species, one of them, for example, looking like Seticotasteromimus (tribe Pissodini), a genus recently erected by Germann [54].In "Ithyporini sensu lato", the differentiation from Cryptorhynchinae was not clear in several cases.Some species from these groups and from Ochyromerini, with more or less doubtful assignments, are depicted in Figures 9  and 10.The genus Phaeopholus (Figure 9a) is the only representative of Hyperinae; prior to this study, it has only been recorded from China and Japan and is thus another genus new to Borneo.Pachyrhynchini on Borneo, and it is represented by 47 specimens in samples from the Mesilau region of Kinabalu in the montane forest.In Conoderinae (Figure 8), the bulk of the 171 morphospecies probably belong to Othippiini and a rather significant number to the usually black-and-white-coloured Menemachini, the remaining to Coryssomerini, Mecopini (s.l.), Campyloscelini and unidentified tribes.We are aware that the taxonomy of the higher tribes is rather uncertain especially in this subfamily [51].A few genera could be identified with little doubt by the keys of Hustache [52], the Intkey CD [53], and museum collections: Phaenomerus, Talimanus, Telephae and Tomicoproctus-and with some doubt: Agametis, Metialma, Odoacis, and Osphilia, but none of the probably largest tribe Othippiini, which has mainly small species.According to Alonso-Zarazaga and Lyal [13], the well-characterized genus Tomicoproctus, represented by one species, was not recorded from Borneo previously, only east of the Wallace line and in Africa.
In Molytinae, there was a great number of Trachodini (63 species), including two Pseuderodiscus species, previously placed in Curculioninae-Erodiscini, followed by Ithyporini (37), Hylobiini (16), Mecysolobini (14), Lithinini (about eight species, one of them identified as Seleuca by C. Lyal by photo) and Trigonocolini (1), but, among Ithyporini, there may also be other tribes with small species, one of them, for example, looking like Seticotasteromimus (tribe Pissodini), a genus recently erected by Germann [54].In "Ithyporini sensu lato", the differentiation from Cryptorhynchinae was not clear in several cases.Some species from these groups and from Ochyromerini, with more or less doubtful assignments, are depicted in Figures 9 and 10.The genus Phaeopholus (Figure 9a) is the only representative of Hyperinae; prior to this study, it has only been recorded from China and Japan and is thus another genus new to Borneo.   ) is provisionally regarded as Ithyporini.Species with a similar habitus look like inhabitants of the soil, but they were found in rather large numbers in the tree crowns; usually, they have a rostral furrow.In the last small species, such a furrow is absent, and this species is regarded provisionally as Seticotasteromimus (Pissodini) or a closely related genus (c,d).

Cerambycidae
For reasons of comparison, the highly diverse longhorned beetles, many of great economic importance and of special interest for collectors, were also included in this work.Like in the other beetle taxa, there are many undescribed species, underlining how little is still known about the canopy fauna of Bornean primary forests.In addition, 219 species were distinguished by Carolus Holzschuh.Nine species could not be identified to genus level and another 11 with doubt, 37 species were identified to species level and another 19 with doubt.Thirteen species from this dataset were described by Holzschuh as new [55][56][57]; and work is still in progress.A current checklist of Bornean Cerambycidae contains 1270 species [58], but the total number is estimated to exceed more than 2000 species [55].The most species-rich tribes in the samples were Apomecynini with 61 and Pteropliini with 31 species, both in subfamily Lamiinae, with genera like Pterolophia, Ropica and Sybra.Only 37 (16.9%) of the 219 recorded species could be identified up to 2012, representing only 2.9% of the Bornean fauna and revealing the conspicuous gaps of knowledge even in a 'collectors' group'.

Brief Summary
The analysis of the canopy weevils provides a first impression on the distribution of higher weevil taxa in Bornean rain forests.Except Brentidae and a few other groups with long narrow body form, there is a very large number of rather small species not exceeding 4 mm.The data suggest that the proportion of species new to science is very high.Several taxa like Apionidae and Ceutorhynchinae contain what appear to be exclusively new species and others mainly new species, such as Baridinae, Brentidae and Rhamphini (Curculioninae).We also verified the existence of at least 15 ) is provisionally regarded as Ithyporini.Species with a similar habitus look like inhabitants of the soil, but they were found in rather large numbers in the tree crowns; usually, they have a rostral furrow.In the last small species, such a furrow is absent, and this species is regarded provisionally as Seticotasteromimus (Pissodini) or a closely related genus (c,d).

Cerambycidae
For reasons of comparison, the highly diverse longhorned beetles, many of great economic importance and of special interest for collectors, were also included in this work.Like in the other beetle taxa, there are many undescribed species, underlining how little is still known about the canopy fauna of Bornean primary forests.In addition, 219 species were distinguished by Carolus Holzschuh.Nine species could not be identified to genus level and another 11 with doubt, 37 species were identified to species level and another 19 with doubt.Thirteen species from this dataset were described by Holzschuh as new [55][56][57]; and work is still in progress.A current checklist of Bornean Cerambycidae contains 1270 species [58], but the total number is estimated to exceed more than 2000 species [55].The most species-rich tribes in the samples were Apomecynini with 61 and Pteropliini with 31 species, both in subfamily Lamiinae, with genera like Pterolophia, Ropica and Sybra.Only 37 (16.9%) of the 219 recorded species could be identified up to 2012, representing only 2.9% of the Bornean fauna and revealing the conspicuous gaps of knowledge even in a 'collectors' group'.

Brief Summary
The analysis of the canopy weevils provides a first impression on the distribution of higher weevil taxa in Bornean rain forests.Except Brentidae and a few other groups with long narrow body form, there is a very large number of rather small species not exceeding 4 mm.The data suggest that the proportion of species new to science is very high.Several taxa like Apionidae and Ceutorhynchinae contain what appear to be exclusively new species and others mainly new species, such as Baridinae, Brentidae and Rhamphini (Curculioninae).We also verified the existence of at least 15 genera, one subfamily and one family formerly not known from Borneo.Furthermore, a high proportion of non-identifiable specimens was confirmed even for larger groups of general interest like the bark and ambrosia beetles (Scolytinae) or the longhorned beetles (Cerambycidae).Information on several other taxa confirm a similar poor state of taxonomic knowledge; this refers to Alticinae (Chrysomelidae, processed by the late M. Döberl), Anthicidae (D. Telnov), Pselaphinae (Staphylinidae, V. Brachat), Scirtidae (B.Klausnitzer) and Tenebrionidae (Alleculinae and Lagriinae excluded; R. Grimm, H. J. Bremer), which were sorted and determined by several specialists.Taking all this into consideration, we cautiously estimate that the proportion of Curculionoidea, as one of the most species-rich taxa in beetles that are new to science, considerably exceeds 80%.Comparable results were found by other authors, too: Oberprieler et al. [11] reported similar proportions of undescribed species for litter-inhabiting species in Central America.In a case study on New Guinea, Basset et al. [59] found a ratio of 4.6:1 of undescribed to described species for arboreal weevils in poorly studied groups, such as Rhynchitidae, Entiminae, Molytinae and Curculioninae s.l.Lees et al. [10] found a ratio of 5.7:1 in Gracillariidae moths from the Neotropics demonstrating that similar faunistic and taxonomic limitations are also found in other taxa.
We also observed conspicuous changes in the species frequency distribution between primary and disturbed forests for several groups of arboreal weevils.These are especially noticeable for those groups which had increased in individual and/or species numbers in the disturbed forests despite a much lower number of foggings than in the primary forests (214 versus 120 samples, Table 2).The total number of collected beetles in both primary and disturbed forests was almost identical (4783 versus 4888), but species numbers were much higher in the primary forests (1287 vs. 473 species).Similarity in numbers was largely caused by Scolytinae which were very abundant and-as the only group-collected with more species in the disturbed forests than in the primary forests.Ochyromerini occurred in large numbers especially in the disturbed forests whereas species richness was much higher in the primary forest.Acalyptini and Baridinae were also found in higher numbers, but lower species richness in the trees of the disturbed forests.It is also noteworthy that not a single individual of Ceutorhynchinae (203 specimens in 26 species) or Attelabidae (60 specimens in 15 species) was collected in the disturbed forests.To what extent these differences can be linked to anthropogenic forest disturbance as suggested for canopy spiders [60] is currently being analysed.The mentioned changes in the rank-abundance distribution were often based on a few species that occurred in large numbers.In Scolytinae, for example, one species of ambrosia beetles of the genus Scolytogenes was collected in more than 1500 individuals from Vitex pinnata (Lamiaceae-Verbenioideae).This might indicate an unknown relationship between this weevil and its host tree.In Acalyptini and Ochyromerini, with the many small light yellowish species, the high abundance in the disturbed forests may be partially attributed to flowering and fruit trees.Several Ochyromerini species and many specimens were collected from Melochia umbellata and from V. pinnata.The most abundant species of Acalyptini (341 individuals) was mainly collected from Terminalia catappa L., Premna corymbosa Rottl.and Willd.and Mallotus spec.These few examples are representative for the limited knowledge on the diversity and biological interactions of beetles in the humid tropical forests.
Diversity 2018, 10, x FOR PEER REVIEW 7 of 20 det.Yoshitake) and several species from further undescribed Mecysmoderini genera in the samples.The Ceutorhynchini are represented by Hainokisaruzo and the Hypohypurini by the single genus Hypohypurus.Unfortunately, Yoshitake, who was interested in receiving all Ceutorhynchinae weevils, has not described any to date.

Figure 4 .
Figure 4. Curculioninae I. Parimera spec.(a) and two unknown small genera of Acalyptini (b,c); the latter cf.Niseida (c); a scymnine-like, red-and-black coloured Demimaea species with trifid hairs ventrally and at the base of the elytra (d) and two long-nosed Curculionini species (e,f); one of them (f) a Pseudoculio species.

Figure 4 . 20 Figure 5 .
Figure 4. Curculioninae I. Parimera spec.(a) and two unknown small genera of Acalyptini (b,c); the latter cf.Niseida (c); a scymnine-like, red-and-black coloured Demimaea species with trifid hairs ventrally and at the base of the elytra (d) and two long-nosed Curculionini species (e,f); one of them (f) a Pseudoculio species.

Figure 9 .
Figure 9. Weevils with doubtful assignments I.The first species (a) represents the Hyperini genus Phaeopholus (confirmed by Jiři Skuhrovec).The second species (b) probably belongs to Microstylini, a tribe of rather uncertain placement (Curculioninae or Molytinae).The third species (c) has a very extraordinary habitus: small, flat, shortened elytra, very thick profemora, and a very long antennal club.There is also sexual dimorphism.The long-nosed female is shown.The assignment is unknown, perhaps Acalyptini (M.A. Alonso-Zarazaga, pers.comm.).The next species (d) is probably a Sophrorhinini species.

Figure 9 .
Figure 9. Weevils with doubtful assignments I.The first species (a) represents the Hyperini genus Phaeopholus (confirmed by Jiři Skuhrovec).The second species (b) probably belongs to Microstylini, a tribe of rather uncertain placement (Curculioninae or Molytinae).The third species (c) has a very extraordinary habitus: small, flat, shortened elytra, very thick profemora, and a very long antennal club.There is also sexual dimorphism.The long-nosed female is shown.The assignment is unknown, perhaps Acalyptini (M.A. Alonso-Zarazaga, pers.comm.).The next species (d) is probably a Sophrorhinini species.

Figure 10 .
Figure 10.Weevils with doubtful assignments II.The weevil on the first photos (a,b) is provisionally regarded as Ithyporini.Species with a similar habitus look like inhabitants of the soil, but they were found in rather large numbers in the tree crowns; usually, they have a rostral furrow.In the last small species, such a furrow is absent, and this species is regarded provisionally as Seticotasteromimus (Pissodini) or a closely related genus (c,d).

Figure 10 .
Figure 10.Weevils with doubtful assignments II.The weevil on the first photos (a,b) is provisionally regarded as Ithyporini.Species with a similar habitus look like inhabitants of the soil, but they were found in rather large numbers in the tree crowns; usually, they have a rostral furrow.In the last small species, such a furrow is absent, and this species is regarded provisionally as Seticotasteromimus (Pissodini) or a closely related genus (c,d).

Table 1 .
Canopy weevils from the superfamily Curculionoidea in the samples from Malaysia (Borneo).

Table 2 .
Weevil groups showing the most conspicuous differences in individual and species numbers between primary and disturbed forests.