Overview
Program coordinator
Yves Basset, Smithsonian Tropical Research Institute (STRI), bassety@si.edu
Background
The ForestGEO ‘Arthropod Initiative’ aims at monitoring key arthropod assemblages over long-term and studying insect-plant interactions over the network of the Smithsonian Forest Global Earth Observatory. The Initiative integrates with ongoing monitoring of plant dynamics within the ForestGEO network, causes minimum possible impact to the plots and focus on a priority set of assemblages chosen for their ecological relevance, taxonomic tractability and ease of sampling. At each participating ForestGEO site, the first year of the program is devoted to a ‘baseline’ survey to serve several purposes, notably to refine the methodology and the definitive choice of assemblages. The baseline survey is followed by longer-term programs of field work and analysis, organized into two main sub-programs: monitoring, and key interaction studies. The monitoring sub- program is directed to detecting long-term changes, as reflected in priority assemblages, driven by climatic cycles, climatic change and landscape scale habitat alteration. Monitoring protocols are derived from those used during the baseline survey. The food web approach of interaction studies targets interactions between plants and specific insect assemblages, with different protocols than those used for monitoring. So far, the Arthropod Initiative involves three SIGEO sites: Barro Colorado Island (BCI) in Panama, Khao Chong (KHC) in Thailand and Wanang (WAN) in Papua New Guinea. At both BCI and KHC, four full-time research assistants are in charge of arthropod protocols. At BCI, the program coordinator, YB, doubled as BCI site supervisor, working at 66% FTE. At KHC, supervision was assured by six short visits of W. Sakchoowong, R. Eastwood and YB. At WAN, no arthropod protocol has formally been initiated.
Annual Report 2009
Monitoring in BCI: Year 2009 represented the first year of monitoring at BCI. The five main insect monitoring protocols derived from last year baseline were successful in estimating population indices for the most common species collected in our protocols. We will probably abandon one of our bee protocol next year and replace it by monitoring nocturnal bees with light traps, as this strategy appears to be feasible at all participating ForestGEO sites. Our local database indicates that so far, 87,991 arthropods were collected on BCI, including 1,040 species of focal taxa representing 16,177 pinned specimens in collections. Taxonomic studies have been initiated for all focal taxa. In addition, we secured an agreement worth ca $50,000 with Paul Hebert (University of Guelph) to DNA barcode for free 5,000 insect specimens from Panama and Thailand. An internal SI grant worth $16,700 allows the transfer of collateral information, including insect pictures, on the web site of the International Barcode of Life Project (http://www.barcodinglife.org/), where the first results are available for BCI butterflies. The resulting quality insect pictures are additionally used as a taxonomic tool in our local database and to produce local guides to improve our protocols. DNA barcodes will help us refining the taxonomy of our material and simplifying some protocols. We are already implementing this strategy with butterfly transects, where we collect only specimens for taxonomically difficult groups.
In 2009, we collected 330 insect samples (80 light trap samples, 50 Winkler samples, 120 butterfly transects, 40 termite transects and 40 bee bait samples) and an additional 60 insect samples as part of a minor re-adjustment of our protocols. We are now able to process, sort and database efficiently all insect samples each year (representing 52,265 specimens in 2009). These data allowed us to estimate annual indices of population for 2009 for all focal groups, and for the most common species within these groups. These estimates were calculated differently for non-social and social insects, due to the spatial aggregation of the later. All together our five protocols were able to provide annual indices for 46 species from our focal groups (6.5% of the total 712 focal species), which represented 47.1% of individuals belonging to our focal groups (6,373 individuals out of 13,519). The high species richness and corresponding low share of species amenable to analysis may not surprise tropical entomologists. However, the few common species amenable to analysis represent nearly half of the total abundance of focal taxa, indicating that our protocols characterize well our focal taxa. Regarding the rest of focal species not amenable to analysis, it must be noted that our protocols showed that their abundance was rather low in 2009, and these results may be compared with those for future years. Next year, we will consolidate protocols and improve the taxonomy of our collections.
Monitoring in KHC: Year 2009 at KHC represented a baseline survey and fine-tuning of insect protocols. This worked reasonably well for a complex project running without on-site supervisor. Our local database so far includes 3,407 specimens representing 400 species and 1,788 pinned specimens in collection. However, far more specimens have been collected but not processed or databased. Our five main monitoring protocols have been initiated successfully. Most of specimens collected in butterfly transects in 2008 and 2009 have been processed and databased. Part of tephritids collected in McPhail traps have been studied but many samples are yet to be analyzed. Most focal moths collected in light trap samples during 2009 have been processed and databased. However, other focal taxa collected with light traps are yet to be processed. Most of Winkler samples and termite transect samples have not been processed yet. Ideally, 2010 should represent another year of baseline survey at KHC, so that we can improve routine insect processing and the taxonomy of our collections, and be ready for the first year of monitoring in 2011.
Monitoring in WAN: At the moment, all efforts at WAN are directed towards delineating the 50ha permanent plot. This should be finished in July 2010. However, pilot studies of different insect protocols have been initiated near the plot and the baseline study will formally start in late 2010. We expect little delays in setting insect protocols in routine mode at WAN, since most have already been used by the Binatang Research Center. Funding from the University of Minnesota allowed our Wanang colleagues to organize a field workshop in February 2009 in Panama, whose aims were to maximize the establishment of the new plot. Two days of the workshop were dedicated to insect protocols, with the participation of representatives of WAN, BCI and KHC.
Interaction studies: At BCI, a small grant form the John Fell Fund will allow O. Lewis and S. Gripenberg (Oxford University) to team up with J. Wright and YB (STRI) in 2010, in order to develop practical methods of quantifying insect seed predation. A small ForestGEO grant to our colleague W. Sakchoowong will allow him initiating studies on the effect of leaf litter composition (derived from different tree species) on litter ants in early 2010 at KHC. We also hope that a protocol for quantifying the abundance of leaf herbivores may be developed at WAN in 2011, as our colleagues at the Binatang Research Centre have ample expertise relevant to this subject.
Expansion of our mini-network: Butterfly transects will initiated in early 2010 at Agua Salud in Panama, in different habitats. From discussions with interested colleagues in 2009, it is possible that one or several SIGEO sites may adopt some or all insect protocols future years, possibly in China, Malaysia, Thailand and Brunei. In 2010, we will look specifically to secure the participation of SIGEO sites in the Neotropics and Africa.
Scientific output: Eight assistants and two interns were trained at BCI and KHC during 2009. The insect collections obtained during 2009 at BCI and KHC are considerable. It is apparent that our BCI collections, notably of moths and termites, will represent a real asset for refining the extent of biodiversity in Panama, even if this is by far not the goal of our studies. Several scientific papers are planned for 2010, including a detailed comparison of butterfly transects at BCI and KHC, and a methodological paper on calculating annual indices for social insects in tropical rainforests. The high quality pictures generated during the extraction of insect parts for DNA barcoding will also serve to develop local field guides. We hope to see other ForestGEO sites join their efforts with ours to monitor and study the largest share of biodiversity on Earth, tropical rainforest arthropods.
Plate I. Representative activities for the ForestGEO Arthropod Initiative in 2009, clockwise: setting up the new ForestGEO light trap model at KHC; impression of the partial phylogenetic tree of butterflies collected at BCI; cover of the ForestGEO termite protocol; Wanang field camp; new Bactrocera species collected at KHC; a shot of the tephritid reference collection at KHC; example of picture used for DNA barcoding and local database at BCI (Prepona laertes octavia).
Annual Report 2010
Taxonomy: The taxonomic study of our material at BCI and KHC has significantly improved over 2009, notably for termites and ants. Our aims are (a) to prepare reference collections supported by pictures and DNA barcodes, and (b) to invite interested taxonomists to study our material when (a) is reasonably accurate. In collaboration with Paul Hebert (University of Guelph), insect legs are harvested to provide DNA barcodes for each of our focal species. This greatly helps refining insect taxonomy and the project is now well underway at BCI (38% of focal species sequenced) and KHC (27%). A new agreement with Paul worth $25,000 will allow further sequencing of ca 2,500 insect specimens in 2011. All information, including insect pictures, is stored on the web site of the International Barcode of Life Project (BOLD, http://www.barcodinglife.org/) and will become public as soon as data are deemed to be reasonably clean.
Monitoring in BCI: Year 2010 represented the second year of insect monitoring at BCI. So far, the BCI database contains data on 159,591 arthropods, including 1,360 species (57% of which with pictures) and 21,685 pinned specimens in our collections. With the improvement of our collections, our routine has likely greatly improved. Many insect specimens that belong to common species well recognized morphologically and described by DNA barcodes are not pinned anymore, unless they are in a pristine state. We introduced barcode labels for our pinned insect specimens, which speeds up data typing and also prevents mistyping sample and specimen codes. We hope that an improved version of our database, currently being developed by ForestGEO, will be available to us next year. We estimated annual indices of population in 2010 for all focal taxa, and for the most common species within these groups. These estimates were calculated separately for non-social and social insects. Although we cannot comment on annual differences with just two years of data, we note that the precision of our estimates was good and will allow accurate estimation of annual population levels for species amenable to statistical analysis. All together our five protocols were able to provide annual indices for 65 species from our focal groups (7% of the total 896 focal species collected in 2010). These 65 species represented 64% of the total abundance of focal taxa, indicating that our protocols characterize well our focal taxa. Further, species accumulation curves suggest that most common species collected in our protocols were already recognized, if not identified (with notable exceptions for ants and termites).
Baseline studies in KHC and WAN: In 2010 at KHC, we consolidated our collections and staff training so that we can initiate proper insect monitoring in 2011. So far, our database includes 58,508 specimens (6,154 pinned specimens in collections) and 1,045 focal species. Our collections at KHC are currently stronger in butterflies, Geometridae, Crambidae and Tephritidae. The other notable development at KHC was the move from our existing laboratory to a larger house which was until now hosting the headquarters of the KHC Botanical Gardens. This will provide us with more space, improved storage facilities for our collections, as well as improved internet access. All efforts at WAN during 2010 were directed towards delineating the new 50ha permanent plot, surveying its topography, and tagging and identifying tree species. Progresses were considerably slower than previously thought. We nevertheless expect the survey of the plot to end in early 2011. These delays, plus a 20% cut in the original budget of a National Science Foundation grant, mean that funds will not be available for insect protocols, as originally planned. We will be seeking additional funding to initiate insect protocols in 2011 at WAN, either in full or limited temporarily to butterfly transects. On the positive side, the Wanang Research Station (Swire Research Centre) has been completed. Interaction studies. Pilot studies at each of the participating plots are intended to develop protocols related to interaction studies that may later be replicated at other ForestGEO plots. At BCI, a project aiming at quantifying insect seed predation was initiated in July 2010 and led by O. Lewis, S. Gripenberg, S.J. Wright and YB. The project is funded through a grant from the John Fell Fund to O. Lewis. During the initial stage of the project, we have collected 21,600 seeds and fruits from a total of 165 woody plant species and reared more than 2,000 insect individuals representing 180 morphospecies. This project will continue in 2011-2013 with funding through a postdoctoral research fellowship from the Academy of Finland granted to S. Gripenberg.
Monitoring in KHC: At KHC, a project led by W. Sakchoowong is studying the influence of leaf litter composition and canopy leachate from dominant tree species on the diversity of ants and pselaphine beetles. The project is funded by the ForestGEO small grant scheme. During 2010, staff were trained to extract arthropods from the leaf litter with Winkler extractors, to identify soil beetles and other insects, and to manage the collections and database relevant to the project. Sampling will continue in 2011.
Scientific output: The BCI team continues to prove that our model is viable and that with a minimum of organization and personnel, insect population can be monitored with high accuracy that will allow statistical analysis of annual data and, therefore, of long-term changes in insect populations. Next year, we will initiate this model at full speed at KHC and also hope to consolidate funding to start insect protocols at WAN. In 2010, the ForestGEO Arthropod Initiative trained 9 assistants and one intern at BCI and KHC. The value of our now considerable insect collections (for example, 100 insect drawers and > 20,000 pinned specimens at BCI) is augmented by collateral information such as sampling data, pictures and, for some specimens, genetic sequences (DNA barcodes). In 2010, our research team published one article about a rare and large lycaenid butterfly species at KHC, and submitted two manuscripts comparing butterfly assemblages in the forest understory at BCI, KHC and WAN. Recent exchanges with several ForestGEO plot PIs in China, Ecuador, Columbia and Gabon suggest that the ForestGEO Arthropod Initiative may be able to expand arthropod studies at other ForestGEO sites in the near future. We look forward to that, as well as consolidating our arthropod protocols and activities at the three current participating ForestGEO sites.
Plate II. Representative activities for the ForestGEO Arthropod Initiative in 2010: (1) barcode label at BCI; (2) extraction of insect legs for DNA barcoding at KHC; (3) topographic survey at WAN; (4) pre-sorting light trap samples on BCI; (5) hemispherical canopy picture at BCI, to help in the interpretation of butterfly transects; (6) Common reduviid and flatid species at light trap samples on BCI; (7) extracting litter arthropods with Winkler sacks at KHC; (8) aerial view of the Swire Research Station at WAN and its inauguration; (9) pictures and illustrative barcode for one BCI geometrid, as available on the BOLD web site; (10) educational poster at KHC; (11) identifying fruit flies at KHC; and (12) fruits damaged by one species of Sesiidae at BCI.
Annual Report 2011
Taxonomy: The taxonomic study of our material at BCI and KHC has significantly improved over 2010, notably for moths and ants. Our aims are (a) to prepare reference collections supported by pictures and DNA barcodes, and (b) to invite interested taxonomists to study our material when (a) is reasonably accurate. In collaboration with Paul Hebert (University of Guelph), insect legs are harvested to provide DNA barcodes for each of our focal species. This greatly helps refining insect taxonomy and the project is now well underway at BCI (67% of focal species sequenced) and KHC (27%). All information, including insect pictures, is stored on the web site of the International Barcode of Life Project (BOLD, http://www.barcodinglife.org/) and are gradually becoming public as soon as data are deemed to be reasonably clean.
Monitoring in BCI: Year 2011 represented the third year of insect monitoring at BCI. So far, the BCI database contains data on 221,594 arthropods, including 1,374 species (72% of which with pictures) and 25,966 pinned specimens in our collections. With the improvement of our collections, our routine has likely greatly improved. Many insect specimens that belong to common species well recognized morphologically and described by DNA barcodes are not pinned anymore, unless they are in a pristine state. An improved version of our old Access database, currently being developed by ForestGEO, should be available to us next year. We estimated annual indices of population in 2011 for all focal taxa, and for the most common species within these groups. These estimates were calculated separately for non-social and social insects. All together our five protocols were able to provide annual indices for 56 species from our focal groups (6% of the total 910 focal species collected in 2011). These 56 species represented 55% of the total abundance of focal taxa (16,662 individuals collected in 2011), indicating that our protocols characterize well our focal taxa. Out of 20 common species that were tested using simple models, 40% showed significant changes in abundance since the inception of the project at BCI. This small time series does not allow further discussion but it confirms that the precision of our estimates is good and suitable to interpret long-term changes.
Monitoring in KHC: Year 2011 represented the first year of insect monitoring at BCI, which went well with 70-100% of focal material already processed by the end of 2011(the remaining backlog will be absorbed before the first survey in March 2012). So far, our database includes 71,670 specimens (9,729 pinned specimens in collections) and 1,329 focal species. We finalized our move into the previous headquarters of the KHC Botanical Gardens and can enjoy now larger space and improved storage facilities. A volunteer student from Oxford University helped the project during five weeks and we received a generous donation of SF5,000.00 from the Swiss bank Pictet to defray the costs of the project at KHC.
Baseline studies in WAN: All efforts at WAN during 2011 were directed towards mapping and identifying tree species in the new 50ha permanent plot. The Wanang Research Station is now operational and different insect projects from our colleagues of the Czech Academy of Science and Binatang Research Centre (Papua New Guinea) have been initiated at the permanent plot. All of these studies will be invaluable for developing future insect monitoring at WAN, and to interpret soundly the results of such protocols. Interaction studies. Pilot studies at each of the participating plots are intended to develop protocols related to interaction studies that may later be replicated at other ForestGEO plots. At BCI, a continuing project aiming at quantifying insect seed predation is led by S. Gripenberg, in collaboration with O. Lewis, S. Pinzon, S.J. Wright and YB. To date, we have collected >85,000 seeds and fruits from >300 woody plant species. More than 10,000 insect individuals representing some 850 morphospecies have emerged from these seeds. The project employed two interns in 2011. At KHC, a project led by W. Sakchoowong with one research assistant is studying the influence of leaf litter composition and canopy leachate from dominant tree species on the diversity of ants and pselaphine beetles. The field component of the project is now finished and Dr Sakchoowong is overseeing insect processing and data analyses.
Scientific output: The BCI and KHC teams are proving that our model is viable and that with a minimum of organization and personnel, insect population can be monitored with high accuracy that will allow statistical analysis of annual data and, therefore, of long-term changes in insect populations. Next year, we will consolidate our protocols at KHC and also hope to start insect protocols at WAN. In 2011, the ForestGEO Arthropod Initiative trained 10 assistants and interns at BCI and KHC. The value of our considerable insect collections is augmented by collateral information such as sampling data, pictures and, for some specimens, genetic sequences (DNA barcodes). In 2011, our research team published one article about the result of our monitoring protocols for butterflies at BCI, KHC and WAN, submitted a related manuscript discussing global monitoring of rainforest butterflies, and another one on the conservation of moths and butterflies. Recent exchanges with several SIGEO plot PIs in China, Gabon, Malaysia, Hong Kong and Vietnam attest of the scientific interest in the ForestGEO Arthropod Initiative. Our aims remain consolidating our arthropod protocols and activities at the three current participating SIGEO sites, hoping that we may be able to expand arthropod studies at other ForestGEO sites in the near future.
Plate III: Representative activities/items for the ForestGEO Arthropod Initiative in 2011: (1) Logo of one sponsor for KHC in 2011; (2) topographic map of the Wanang 50 ha plot; (3) one page of the new KHC butterfly field guide; (4) Embiratermes chagresi, a termite species from BCI (photo courtesy M. Leponce); (5) Graphium arycles, a common papilionid at KHC; (6) Winkler samples at KHC; (7) view of the new air-conditioned lab at KHC; (8) insect rearing from seeds and fruits at BCI; (9) part of the crew at KHC in the new local, with insect displays; (10) Elysius sp. 1YB, a collection specimen at BCI; (11) diagram of the new transects at KHC; and (12) Y. Gonzalez looking at BCI butterflies, as published in one issue of STRI News.
Annual Report 2012
Taxonomy: The taxonomic study of our material at BCI and KHC has significantly improved over 2011, notably for moths and ants. Our aims are (a) to prepare reference collections supported by pictures and DNA barcodes, and (b) to invite interested taxonomists to study our material when (a) is reasonably accurate. In collaboration with Paul Hebert (University of Guelph), insect legs are harvested to provide DNA barcodes for each of our focal species. This greatly helps refining insect taxonomy and the project is now well underway at BCI (75% of focal species sequenced) and KHC (25%). All information, including insect pictures, is stored on the web site of the International Barcode of Life Project (BOLD, http://www.barcodinglife.org/) and are gradually becoming public as soon as data are deemed to be reasonably clean. WE received a $49,000 grant from the Smithsonian Institution Barcoding Opportunity FY012 to defray costs of insect DNA barcoding at BCI.
Monitoring in BCI: Year 2012 represented the fourth year of insect monitoring at BCI. So far, the BCI database contains data on 282,838 arthropods, including 1,473 species (69% of which with pictures) and 28,638 pinned specimens in our collections. With the improvement of our collections, our routine has likely greatly improved. Many insect specimens that belong to common species well recognized morphologically and described by DNA barcodes are not pinned anymore, unless they are in a pristine state. An improved version of our old Access database, currently being developed by ForestGEO, should be available to us next year. We estimated annual indices of population in 2012 for all focal taxa, and for the most common species within these groups. These estimates were calculated separately for non-social and social insects. All together our five protocols were able to provide annual indices for 57 species from our focal groups (7% of the total 779 focal species collected in 2012). These 57 species represented 26% of the total abundance of focal taxa (42,114 individuals collected in 2012), indicating that our protocols characterize well our focal taxa. Out of 20 common species that were tested using simple models, 25% showed significant increase in abundance/occurrence since the inception of the project at BCI. This small time series does not allow further discussion but it confirms that the precision of our estimates is good and suitable to interpret long-term changes. Monitoring – KHC. Year 2012 represented the second year of insect monitoring at BCI, which went well with 60-100% of focal material already processed by the end of 2012(the remaining backlog will be absorbed next year). So far, our database includes 91,235 specimens (15,110 pinned specimens in collections) and 1,754 focal species. Two volunteer students from Harvard University and Oxford University helped the project during eight weeks.
Baseline studies in WAN: Most efforts at WAN during 2012 were directed towards identifying tree species in the new 50ha permanent plot. The Wanang Research Station is now operational and different insect projects from our colleagues of the Czech Academy of Science and Binatang Research Centre (Papua New Guinea) have been initiated at the permanent plot. All of these studies will be invaluable for developing future insect monitoring at WAN, and to interpret soundly the results of such protocols. YB spent a month at WAN training 16 local parataxonomists for an altitudinal project on Mt Wilhelm, with the Wanang plot as reference point.
Interaction studies. Pilot studies at each of the participating plots are intended to develop protocols related to interaction studies that may later be replicated at other SIGEO plots. At BCI, a continuing project aiming at quantifying insect seed predation is led by S. Gripenberg, in collaboration with O. Lewis, H. Barrios, S. Pinzon, S.J. Wright and YB. At KHC, a project led by W. Sakchoowong is studying the influence of leaf litter composition and canopy leachate from dominant tree species on the diversity of ants and pselaphine beetles. The field component of the project is now finished and Dr Sakchoowong is overseeing insect processing and data analyses.
Scientific output. The BCI and KHC teams are proving that our model is viable and that with a minimum of organization and personnel, insect population can be monitored with high accuracy that will allow statistical analysis of annual data and, therefore, of long-term changes in insect populations. Next year, we will consolidate our protocols at KHC and also hope to start insect protocols at WAN. In 2012, the FORESTGEO Arthropod Initiative trained 10 assistants and interns at BCI and KHC. The value of our considerable insect collections is augmented by collateral information such as sampling data, pictures and, for some specimens, genetic sequences (DNA barcodes). In 2011, our research team published two articles about insect monitoring in tropical rainforests. YB and colleagues also published an article in Science about a project, initiated before the ForestGEO Arthropod Initiative in Panama, which was instrumental in developing the rationale and some of the protocols used in the ForestGEO Arthropod Initiative. Recent exchanges with SIGEO plot PIs in China and Gabon attest of the scientific interest in the ForestGEO Arthropod Initiative. Our aims remain consolidating our arthropod protocols and activities at the three current participating ForestGEO sites, hoping that we may be able to expand arthropod studies at other ForestGEO sites in the near future.
Plate IV. Representative activities/items for the ForestGEO Arthropod Initiative in 2012: (1) Carmenta sp., a Sesiidae reared from seeds of Prioria copaifera (Fabaceae) on BCI. (2) M. Panmeng performing a butterfly transect at KHC. (3) Identifying tree species at the WAN plot is still not finished yet. (4) Front cover of Science, 14 December 2012, where Y. Basset et al. reported on arthropod diversity in a Panamanian forest. (5) Screenshot of the new, on-line, ForestGEO insect database being developed by R. Reeves. (6) Rearing facility for the seed predator project at BCI. (7) Home made flight interception trap located at WAN for project IBISCA-Niugini. (8) Part of the parataxonomists trained at WAN. (9) Training of parataxonomists at WAN for the beating protocol. (10) Selection of seeds collected on the forest trail to the WAN research station. (11) Atrophaneura coon, a spectacular Papilionidae from KHC. (12) Head of Pyramica sp., a common genus of litter ants at BCI (at least 6 species; photo M. Leponce).
Annual Report 2013
Taxonomy: In collaboration with Paul Hebert (University of Guelph), insect legs are harvested to provide DNA barcodes for each of our focal species. This greatly helps refining insect taxonomy and the project is now well underway at BCI and KHC. All information, including insect pictures, is stored on the web site of the International Barcode of Life Project (BOLD, http://www.barcodinglife.org/) and are gradually becoming public as soon as data are deemed to be reasonably clean. The taxonomic study of our material at BCI has significantly improved over 2012, notably for moths, ants and flatids. A Smithsonian grant facilitated the addition of 3,124 DNA sequences, and several experts revised our collections. Our aims remain to (a) prepare reference collections supported by pictures and DNA barcodes, and (b) invite interested taxonomists to study our material when (a) is reasonably accurate. We received a $19,556 grant from the Smithsonian Institution Barcoding Opportunity FY013 to defray costs of insect DNA barcoding at KHC next year.
Monitoring in BCI: The BCI database now displays internet links to the Barcodes Index Numbers (BINs) hosted by BOLD, which allows checking conveniently the occurrence of cryptic species. Year 2013 represented the fifth year of insect monitoring at BCI. So far, the BCI database contains data on 309,292 arthropods, including 1,643 species (78% of which with pictures) and 31,313 pinned specimens in our collections. With the improvement of our collections, our routine has likely greatly improved. Many insect specimens that belong to common species well recognized morphologically and described by DNA barcodes are not pinned anymore, unless they are in a pristine state. We estimated annual indices of population in 2013 for all focal taxa, and for the most common species within these groups. These estimates were calculated separately for non-social and social insects. All together our five protocols provided annual indices for 54 species from our focal groups (8% of the total 704 focal species collected in 2013). These 54 species represented 45% of the total abundance of focal taxa (22,975individuals collected in 2013), indicating that our protocols characterize well our focal taxa. We assessed temporal changes in several assemblages during 2009-2013. Directional changes appear to be few, with the exception of perhaps litter ants. Temporal patterns also appear to be unrelated between taxa. This simple analysis emphasized the need to monitor different arthropod taxa, as their responses to environmental stress may be rather different.
Monitoring in KHC and WAN: Year 2013 represented the third and first years of insect monitoring at KHC and WAN, respectively. At KHC most of the material collected in 2013 has been processed (the remaining backlog will be absorbed next year), and our database includes 114,260 specimens (22,586 pinned specimens in collections) and 1,996 focal species. One volunteer student from Oxford University helped the project during two months. At WAN, butterfly transects are now in routine mode. A motorized balloon flew over the WAN plot to collect insects from the forest canopy in April 2013. This represented the first flight of an air balloon in Papua New Guinea. Interaction studies. A three-year grant from the Grant Agency of the Czech Republic ($323,121) allowed to replicate the seed predation protocol established at BCI in the plots of KHC and WAN. The project aims at comparing quantitative food webs based on seeds and their insect predators and parasitoids at BCI, KHC and WAN. Scientific output. The BCI and KHC teams continue to prove that our model is viable and that with a minimum of organization and personnel, insect population can be monitored with high accuracy that will allow statistical analysis of annual data and, therefore, of long-term changes in insect populations. In 2013, the ForestGEO Arthropod Initiative trained 14 assistants and interns at BCI, KHC and WAN. The value of our now considerable insect collections at BCI and KHC is augmented by collateral information such as sampling data, pictures and, for many specimens, genetic sequences (DNA barcodes). We hope that our mini-network consisting of 5 sites may further expand in 2014 (Gabon, China), and that some of the participating sites, such as WAN, Yasuni or Hong Kong, may also initiate additional protocols than the one currently in use. Our aims remain consolidating our arthropod protocols and activities at the three current main participating ForestGEO sites of BCI, KHC and WAN.
Plate V. Representative activities/items for the ForestGEO Arthropod Initiative in 2013: (1) Thumbnail pictures of fruits/seeds collected at KHC; (2) one of the Phostria species determined by A. Solis; (3) the air balloon inflated near the Wanang plot; (4) screenshot of the BCI database displaying a list of clickable Barcode Index Numbers for each morphospecies; (5) the corresponding BIN window from the BOLD database; (6) a brief story about butterfly transects in the STRI News of July 2013; (7) S. Gripenberg and C. Dahl trekking to the Wanang plot; (8) Details of the molecular tree for the 805 sequences of Arctiinae for BCI; (9) C. Dahl inspecting seed rearing containers at Wanang; (10) draft guide of Wanang butterflies; (11) C. Chung performing butterfly transects at Hong Kong; (12) illustrated guide for BCI Flatidae with genitalia sketches.
Annual Report 2014
Monitoring in BCI, KHC and WAN: During 2014, we replicated butterfly and termite transects near BCI in the Canal Zone as part of an environmental study for Argos (CementoPanama). This boosted our butterfly collections and we will prepare a report for Argos summarizing the results in 2015. In November, we moved from two cubicles in the Maestria de Entomologia 2
(University of Panama) to a large room (81m ; Plate 1) of the Center for Tropical Palaetolongy and Archaelogy (CTPA). Year 2014 represented the sixth year of insect monitoring at BCI. So far, the BCI database contains data on 340,914 arthropods, including 1,792 species (1,628 of which with pictures, 91%) and 34,921 pinned specimens in our collections. Instead of detailing statistics for each protocol performed on BCI during 2014, as we did for previous years, we present in the long version of this report a draft of a manuscript on the Saturniidae of BCI. The abstract is detailed below. Year 2014 represented the fourth and second years of insect monitoring at KHC and WAN, respectively. At KHC most of the material collected in 2014 has been processed (the remaining backlog will be absorbed next year), and our database includes 132,046 specimens (26,303 pinned specimens in collections) and 2,209 focal species. Two volunteer students from Bristol University helped the project for two months. At WAN we implemented termite transects, Winkler protocols for litter ants and monitoring of Tephritidae by McPhail traps this year. This complement the butterfly protocols initiated last year. The ForestGEO insect database contains data on 11,567 specimens, but apart from butterflies, few of these specimens are yet identified.
Interaction studies: A current project is replicating the seed predation protocol established at BCI in the plots of KHC and WAN. The project aims at comparing quantitative food webs based on seeds and their insect predators and parasitoids at BCI, KHC and WAN. Field work is now finished on BCI, while we are still actively rearing insects from seeds at KHC and WAN. At BCI, we reared 23,500 insects from 486 plant species. At KHC, we reared 5,922 insects from 272 plant species, while at WAN we reared 24,965 insects from 296 plant species.
Scientific output: The BCI and KHC teams continue to prove that our model is viable and that with a minimum of organization and personnel, insect population can be monitored with high accuracy that will allow statistical analysis of annual data and, therefore, of long-term changes in insect populations. In 2014, the ForestGEO Arthropod Initiative trained 16 assistants and interns at BCI, KHC and WAN. We hope that our mini-network consisting of 7 ForestGEO sites may be consolidated in 2015 by securing current monitoring programs at the sites which joined recently the network, and by the addition of further monitoring programs targeting different taxa at most sites.
The Saturniidae of BCI: DNA barcodes and recent population trends –Abstract. The Saturniidae are among the largest moths and are best represented in the Neotropics, but there are few recent lists available for Panama and particularly for Barro Colorado Island (BCI), one of the best-studied rainforest sites in the tropics. By collating published lists, and recent rearing and monitoring records, we provide an actualized list of the 60 species of Saturniidae collected on BCI during the period 1958-2014. This list will serve as baseline data for assessing long-term changes of saturniids on BCI in the future, as 87% of species can be identified by their unique DNA Barcode Index Number, including 8 cryptic species not yet formally described. A local species pool of 60+ species breeding on BCI looks plausible, but more cryptic species are bound to be discovered in the future. We use monitoring data with light traps to analyze recent population trends on BCI for saturniid species that were relatively common during 2009-2014, this period representing probably > 20 saturniid generations. The abundance of 7 species, out of 11 tested, could be fitted to significant time-series models. One species showed a strong linear increase in time, whereas six species had significant effects for each time point. Of the latter, one species showed a strong decline in time, two a moderate decline and for the three others the direction of the change was uncertain. Seven out of the 11 species showed a peak in abundance during 2011 but this observation remains unexplained. We conclude that Saturniidae may be recommended as model taxa for studying the long-term effects of climate change on tropical insects, but that their relative sensitivity to these effects must be compared to that of insect taxa popular in the conservation literature.
Plate VI. Representative activities/items for the ForestGEOArthropod Initiative in 2014. (1) Drone flight over KHC plot, see youtube link (S. Milne). (2) Composite view of the new insect laboratory at CTPA in Panama. (3) Extracting litter ants at Wanang. (4) McPhail trap to attract Tephritidae at Wanang. (5) The first butterflies collected at the Xishuangbanna plot in China (A. Nakamura). (6) Extract of the Tephritidae guide prepared by M. Schutze for Wanang. (7) Seeds reared at Wanang showing attacks of Scolytinae. (8) Litterfall traps in construction that will be later set up at Wanang for the seed predation project. (9) Butterflies waiting to have a leg extracted for DNA barcoding at KHC. (10)-(12) Representative pictures of Saturniidae, as featured in the long version of this report: Rhescyntis hippodamia, Periphoba sp. 1YB and Dysdaemoria boreas.
