THE POTENTIAL OF ARTIFICIAL NESTING SITES FOR INCREASING THE POPULATION DENSITY OF THE BLACK CACAO ANTS

Transcription

1 The Indonesin potentil Journl of rtificil of Agriculture nesting sites 3(1),... 21: THE POTENTIAL OF ARTIFICIAL NESTING SITES FOR INCREASING THE POPULATION DENSITY OF THE BLACK CACAO ANTS Meldy L.A. Hosng ), Christin H. Schulze ), Tej Tschrntke c), nd Dmynti Buchori d) ) Indonesin Coconut nd Other Plme Reserch Institute, Jln Bethesd II, Mpnget PO Box 14, Mndo 951, North Sulwesi, Phone: (431) 81243, 85143, Fcs.: (431) 81217, Emil: litk@indost.net.id ) Deprtment of Animl Ecology I, University of Byreuth, D-9544 Byreuth, Germny c) Agroecology, University of Göttingen, Wldweg 26, 3773 Göttingen, Germny d) Deprtment of Plnt Pests nd Diseses, Fculty of Agriculture, Bogor Agriculturl University, Kmpus IPB Drmg, Bogor 1668, West Jv, Indonesi ABSTRACT The lck cco nt (Dolichoderus thorcicus Smith) (Hymenopter: Formicide) is one of the predtors commonly found in cco plnttions nd known to e n effective iologicl control gent of cco pests. One of the limiting fctors for the popultion density of the nts in cco plnttions my e the vilility of dequte nesting sites. The purposes of this study were to evlute which rtificil nesting sites re most successful in ttrcting lck cco nts nd in which nesting sites of D. thorcicus colonies rech the lrgest size. The study ws conducted in two cco plnttions in the vicinity of the Mkmur villge locted t the estern mrgin of the Lore Lindu Ntionl Prk, Plolo vlley, Centrl Sulwesi. At ech cco plnttion, 2 trees colonized y lck cco nts were selected. In every chosen tree, ll five nest types were plced for two weeks. D. thorcicus significntly preferred rtificil nesting sites constructed of undles of dry coconut leflets. Other rtificil nests (undles of dry nn nd cco leves, hollow moo filled with dry leves, nd plstic gs filled with dry cco leves) did contin colonies with less dult nts, lrve, nd pupe. [Keywords: Dolichoderus thorcicus, Theorom cco nesting, iologicl control] INTRODUCTION In Indonesi, iologicl control of the mirid ug y introducing the lck cco nts ws first documented in the erly 19s. There were lrge signords on the min rods of the Siloewok Swngn estte (Centrl Jv), t the time it ws the lrgest cco estte in Jv, sying without lck nts no cco (Gieserger 1983). This emphsizes the importnce of nts in controlling cco pests, ut this prctice ws discontinued s pesticides ecme widely ville (Perfecto nd Cstineirs 1998). More studies on D. thorcicus were conducted in Mlysi in the 198s. Studies y Khoo nd Chung (1989) nd Wy nd Khoo (1989) showed tht the lck cco nts could protect cco pods ginst Helopeltis sp. Furthermore, See nd Khoo (1996) reported tht n incresing undnce of the lck cco nts ws ssocited with significnt reduction of the percentge of unextrctle pods (ll ens stuck to ech other nd to the husk). They lso found tht the percentge of pods infested y C. crmerell ws generlly elow 5% in plots with high nt undnce of D. thorcicus nd mostly exceeded the 5% level in plots where these nts were scrce. While in mny plnttions the lck cco nt is extremely rre; in some plnttions it does not occur t ll. To study if nesting site ws n importnt limiting fctor for Biologicl control y using nturl enemies such s prsitoids, predtors, nd insect pthogens hs gined more interest in the lst ten yers. The use of nts for controlling insect pests ws first reported from Chin (Flint nd vn den Bosch 1981; Hölldoler nd Wilson 199). The lck cco nt (Dolichoderus thorcicus Smith) (Figure 1) is one of the well-known predtors tht hs high potentil to ct s iologicl control gent ginst cco pests. The nt is cple of decresing pod dmge cused y mirid ugs of the genus Helopeltis sp. nd the cco pod orer (Conopomorph crmerell Snellen) (Gieserger 1983; Khoo nd Chung 1989; Wy nd Khoo 1989; Khoo nd Ho 1992; See nd Khoo 1996; Khoo 21) nd lso cn protect cco trees ginst folivorous insects (Hosng 24). Figure 1. Blck cco nt (Dolichoderus thorcicus) ttending white melly ugs (Ctenococcus hispidus Mrrison) on the surfce of cco pod.

2 46 Meldy L.A. Hosng et l. the undnce of D. thorcicus, five different kinds of rtificil nesting sites were plced in two cco plnttions. In prticulr, the purposes of this study were to evlute which rtificil nesting sites were most successful in ttrcting lck cco nts nd in which nesting sites D. thorcicus colonies rech the lrgest size. MATERIALS AND METHODS The study ws conducted t the estern mrgin of the Lore Lindu Ntionl Prk in Plolo vlley, Centrl Sulwesi. The two selected cco plnttions, Mkmur 1 (S ' E ') nd Mkmur 2 (S ' E '), were locted in the vicinity of the Mkmur Villge (Plolo Sudistrict, Donggl District) t n ltitude of 62 nd 61 m sl, respectively. The study ws conducted in Octoer 22 synchronously t oth selected study sites, Mkmur 1 nd Mkmur 2. In oth cco plnttions, five different kinds of rtificil nesting sites were plced, i.e. (1) Coc = undle of dry coconut leflets; (2) Bn = undle of dry nn leves; (3) Cc = undle of dry cco leves; (4) Bm = moo (length: 25 cm, dimeter: 1 cm) filled with dry cco leves; nd (5) Bg = trnsprent plstic g filled with dry cco leves At oth cco plnttions, 2 trees colonized y lck cco nts were selected (2 x 2 trees = 4 trees). In every tree ll five nest types were plced to lck cco nts y tying the rtificil nests on the trunk or min rnch of tree (2 x 2 nests per nest type = 4 nests per nest type; 4 nests x 5 nest types = totl of 2 nests). Due to this experimentl design, the nts hd the possiility to choose etween the different nest types in every tree (Figure 2). After exposing ll nests for 2 weeks, they were recollected from the trees nd the size of the estlished colonies ws quntified for ll nests y counting the numer of dult nts, lrve, nd pupe. Additionlly, D. thorcicus nts ctive t the tree trunks of m ove ground nd on the leves nd rnches of the lower cnopy were counted or estimted for ll treted trees nd trees nery which were not equipped with rtificil nt nesting Figure 2. Artificil nesting sites plced to lck cco nts: undle of dry coconut leflets (top left), undle of dry nn leves (top center), undle of dry cco leves (top right), plstic g filled with dry cco leves (ottom left), nd moo filled with dry cco leves (ottom right).

3 The potentil of rtificil nesting sites sites ut colonized nturlly y lck cco nts. Therefore, for ech studied plnttion, 2 pirs of trees colonized with lck cco nts ut with nd without rtificil nt nests re ville for comprison. One-wy nlysis of vrince (ANOVA) ws used to nlyse effects of rtificil nest types on coloniztion success nd colony size of the lck cco nts. Scheffé test ws pplied to compre the popultion of lck cco nts on trees with nd without rtificil nests. The progrm Sttistic 5.5 (SttSoft 1999) ws used to perform ll sttisticl nlyses. Mens were given ± stndrd devition. RESULTS AND DISCUSSION Effect of Type of Nesting Sites on Coloniztion Success Blck cco nts colonized ll types of rtificil nesting sites plced in cco trees, ut the five types of nests differed in respect to their coloniztion rte. In Mkmur 1, D. thorcicus colonized ll (1%) Coc nesting sites. In respect to their coloniztion rte, the remining four types of nesting sites were rnked in the following sequence: Bn (83.75% of plced nesting sites colonized), Cc (78.75%), Bm (6.%), nd Bg (36.25%). A similr rnking of nesting sites ws found t Mkmur 2, lthough lower coloniztion rtes were found on Coc (88.5%), Bn (63.75%) nd Cc (58.75%), nd higher percentge of Bm (83.75) nd Bg nesting sites (46.25%) were colonized (Figure 3). Only the rnking of the Bm nesting site showed pronounced difference compred to Mkmur 1. Its coloniztion rte (83.75%) ws 23.75% higher thn tht t Mkmur 1. In generl, Coc ws the rtificil nesting site which proved to e most ttrctive for lck cco nts in oth studied cco plnttions. Effect of Artificil Ant Nests on Colony Size Numer of nt workers differed significntly t oth study sites etween rtificil nt nests colonized y D. thorcicus (one-wy ANOVA; Mkmur 1: F 4,15 = 17.7, p <.1; Mkmur 2: F 4,15 = 12.21, p <.1). In Mkmur 1, Coc nd Bn nests were occupied y significntly lrger colonies thn other types of nesting sites (Cc, Bm, Bg). In Mkmur 2, only the men numer of workers in Coc nesting sites ws significntly higher compred to other types of nesting sites (Figure 4). The numer of nt queens differed significntly t oth study sites etween the rtificil nesting sites colonized y D. thorcicus (one-wy ANOVA; Mkmur 1: F 4,15 = 8.77, Coloriztion rte (%) Figure 3. Mkmur 1 Mkmur 2 Coc Bn Cc Bm Bg Artificil nt nest Coloniztion rtes of Dolichoderus thorcicus on different types of rtificil nesting sites plced in two cco plnttions (Mkmur 1 nd Mkmur 2) in cco trees lredy occupied y D. thorcicus; Coc = undle of dry coconut leflets, Bn = undle of dry nn leves, Cc = undle of dry cco leves, Bm = moo filled with dry cco leves, nd Bg = trnsprent plstic g filled with dry cco leves. p <.1; Mkmur 2: F 4,15 = 9.1, p <.1). In Mkmur 1, significntly more queens were counted in Coc nd Bn nests compred to Bm nd Bg nests. Cc nests did not differ from other rtificil nesting sites. In Mkmur 2, significntly more queens were recorded in Coc nests thn tht in ll other types of rtificil nesting sites (Figure 5). The numer of nt lrve t oth study sites differed significntly etween the plced rtificil nesting sites colonized y D. thorcicus (one-wy ANOVA; Mkmur 1: F 4,15 = 1.2, p <.1; Mkmur 2: F 4,15 = 7.2, p <.1). In Mkmur 1, D. thorcicus colonies in Coc nd Bn nests contined significnt higher numer of lrve compred to Cc nd Bg nests. Only Cc nest did not show significnt difference compred to ny other type of rtificil nesting sites. In Mkmur 2, Coc nest hd significnt higher numer of nt lrve compred to Cc, Bm nd Bg nests, ut not Bn nest which did not differ from ll other types of rtificil nesting sites (Figure 6). The numer of nt pupe in D. thorcicus colonies differed significntly t oth study sites etween the different types of plced rtificil nesting sites (one-wy ANOVA; Mkmur 1: F 4,15 = 5.12, p <.1; Mkmur 2: F 4,15 = 9.25, p <.1). In the cco plnttion of Mkmur 1, Coc nests contined significnt higher numer of pupe thn Bm nd Bg nests. No significnt differences were found etween Bn nd Cc nests nd ll other nests. At Mkmur 2 in Coc nests, significntly more pupe were counted thn tht in ll other nest types (Figure 7). For oth cco plnttions, the comprison etween cco trees with nd without rtificil nesting sites (n = 2 trees per group) showed tht trees with rtificil nesting spce

4 48 Meldy L.A. Hosng et l. Numer of nt workers Mkmur 1 Numer of nt workers Mkmur 2-5 Coc Bn Cc Bm Bg -5 Coc Bn Cc Bm Bg Artificil nt nest Artificil nt nest Figure 4. Men numer of nt workers of Dolichoderus thorcicus per nest in different types of rtificil nesting sites plced in cco plnttions of Mkmur 1 nd Mkmur 2. Significnt differences (Scheffé test) re mrked y different letters. Coc = undle of dry coconut leflets, Bn = undle of dry nn leves, Cc = undle of dry cco leves, Bm = moo filled with dry cco leves, nd Bg = trnsprent plstic g filled with dry cco leves. 2 Mkmur 1 2 Mkmur 2 Numer of nt queens Numer of nt queens Coc Bn Cc Bm Bg -5 Coc Bn Cc Bm Bg Artificil nt nest Artificil nt nest Figure 5. Men numer of Dolichoderus thorcicus queens per nest counted in different types of rtificil nesting sites in the cco plnttions of Mkmur 1 nd Mkmur 2. Significnt differences (Scheffé test) re mrked with different letters. Coc = undle of dry coconut leflets, Bn = undle of dry nn leves, Cc = undle of dry cco leves, Bm = moo filled with dry cco leves, nd Bg = trnsprent plstic g filled with dry cco leves. 2 Mkmur 1 2 Mkmur 2 Numer of nt lrve Numer of nt lrve Coc Bn Cc Bm Bg -5 Coc Bn Cc Bm Bg Artificil nt nest Artificil nt nest Figure 6. Men numer of nt lrve in Dolichoderus thorcicus colonies in different types of rtificil nesting sites in cco plnttions of Mkmur 1 nd Mkmur 2. Significnt differences (Scheffé test) re mrked with different letters. Coc = undle of dry coconut leflets, Bn = undle of dry nn leves, Cc = undle of dry cco leves, Bm = moo filled with dry cco leves, nd Bg = trnsprent plstic g filled with dry cco leves.

5 The potentil of rtificil nesting sites Mkmur 1 Mkmur 2 Numer of pupe Numer of pupe Coc Bn Cc Bm Bg Artificil nt nest -2 Coc Bn Cc Bm Bg Artificil nt nest Figure 7. Men numer of nt pupe counted in Dolichoderus thorcicus colonies from different types of rtificil nesting sites in cco plnttions of Mkmur 1 nd Mkmur 2. Significnt differences (Scheffé test) re mrked with different letters. Coc = undle of dry coconut leflets, Bn = undle of dry nn leves, Cc = undle of dry cco leves, Bm = moo filled with dry cco leves, nd Bg = trnsprent plstic g filled with dry cco leves. hd significntly higher numer of nts forging on tree trunks s well s on rnches nd leves in the lower cnopy (Figure 8). The results showed tht the lck cco nt D. thorcicus ws cple of estlishing colonies in ll of five different rtificil nesting sites plced in the cco trees, i.e. undles of coconut leves (Coc), of nn leves (Bn), of cco leves (Cc), hollow moo filled with dry cco leves (Bm), nd trnsprent plstic gs filled with cco leves (Bg). These colonies were not only stellite colonies ut lso reproductive colonies y themselves. This ws confirmed y their lrge numer of lrve nd pupe. Most of the plced nesting sites were lredy colonized fter 2 weeks. Therefore, it seems tht nesting spce is mjor limiting fctor for the density of D. thorcicus in cco plnttions. When it comes to which type of rtificil nesting sites cn e recommended to frmers who intend to increse the density of lck cco nts in their plnttions, the rnking of rtificil nesting sites is Coc > Bn > Cc = Bm = Bg. Dry coconut leves (such s Bn nd Cc) which re the chepest wy of providing dditionl nesting spce for D. thorcicus were lredy found to e perfect s nesting spce y Khoo (21). He rgued tht the coconut plm ws n excellent source of shelter for the lck cco nts ecuse ggregtions of D. thorcicus could lso nturlly e found under coconut fronds. When the frond ecomes old nd drops from the plm, the drying leflets form long tues, offering nesting site. Cco is commonly grown under shde trees such s Gliricidi, Leucen, Alizi nd coconut. Of these shde trees, coconut plms were clerly fvored y the lck cco nts. Where the nt is well estlished, ctive tril cn e seen running etween the cco trees nd the Numer of nts Figure 8. Mkmur 1 t = 2.28 df = 38 N = 2 p =.28 Mkmur 2 t = 3.73 df = 38 N = 2 p <.1 YES NO YES NO Artificil nesting site Men numer of counted/estimted Dolichoderus thorcicus workers forging in cco trees (plnttions Mkmur 1 nd Mkmur 2) with (YES) nd without (NO) rtificil nesting sites; N = numer of oserved pirs of cco trees. crowns of coconut plms, even those re more thn 18 m tll. A vriety of homopterns, including Ctenococcus hispidus (Morrison), reside on the leves nd fruits of the coconut plms nd lck cco nts ttend them for their honeydew, source of sustennce (Khoo 21). Honeydew from the mely ug C. hispidus is the min source of food for the lck cco nts lthough there re dditionl other sources (Khoo nd Chung 1989). Ho nd Khoo (1997) documented D. thorcicus using C. hispidus lone s source of food over n 8-week period of oservtion without showing ill effects. A successful introduction of the lck cco nts does not only depend on the type of rtificil nesting sites ut lso on the ville food source nd other fctors such s the presence of other dominnt nt species.

6 5 Meldy L.A. Hosng et l. CONCLUSION All types of rtificil nesting sites were cple of incresing the density of D. thorcicus in cco plnttions nd cn e used to introduce lck cco nts in new loctions. However, we recommend prticulrly nesting sites prepred y using dry leflets of coconut nd nn leves ecuse they proved to e most successful in ttrcting D. thorcicus. Additionlly, they re much cheper thn nesting sites prepred y using moo or plstic gs (filled with dry cco leves). Incresing the popultion density of the lck cco nts in cco trees my effectively contriute to the iologicl control of cco pests. ACKNOWLEDGEMENTS Mny thnks re extended to Prof. Dr. Sjfrid Mnuwoto nd Dr. Purnm Hidyt (Bogor Agriculturl University) for their support of the reserch, their vlule dvice nd mny fruitful discussions. We re grteful to Prof. Dr. Gerhrd Gerold, Prof. Dr. Edi Guhrdj, Dr. Andriyono Kilt Adhi, Dih Rtndewi, Dr. Elim Som, Dr. Silvi Werner, Widi Wti nd Rin Yusuf of the Germn-Indonesin Reserch Project STORMA nd lst ut not lest to ll STORMA drivers for their gret support nd help. Furthermore, we thnk to Pk Drmn, Pk Edy nd Hji Side, who ssisted us in the field nd gve permission to conduct reserch on their lnd, Moh Iql, Iwn Suluk Pdng, Slmet Ridi, Rti Untri, Kent Kumjy, Amr Skti nd Ir Menturuk provided vlule help in the field nd ssisted in sorting the smples. The reserch project ws prtly funded y the Deutsche Forschungsgemeinschft (SFB 552). REFERENCES Flint, M.L. nd R. vn den Bosch Introduction to Integrted Pest Mngement. Plenum Press, New York. Gieserger, G Biologicl control of Helopeltis of coco in Jv. p In H. Toxopeus nd P.C. Wessel. Coco Reserch in Indonesi Americn Coco Reserch Institute, Interntionl Office of Coco nd Chocolte, Wgeningen. Ho, C.T. nd K.C. Khoo Prtners in iologicl control of coco pest: Mutulism etween Dolichoderus thorcicus (Hymenopter: Formicide) nd Ctenococcus hispidus (Hemipter: Pseudococcide). Bull. Entomol. Res. 87: Holldoler, B. nd E.O. Wilson The Ants. Springer-Verlg, Berlin. Hosng, M.L.A. 24. Interction etween Nturl Enemies, Herivores nd Cco in Plolo Vlley, Centrl Sulwesi. PhD Disserttion, Grdute School Bogor Agriculturl University. Khoo, K.C. nd G.F. Chung Use of the lck cco nt to control mirid dmge in coco. The Plnter 65: Khoo, K.C. nd C.T. Ho The influence of Dolichoderus thorcicus (Hymenopter: Formicide) on losses due to Helopeltis theivor (Heteropter: Miride), lck pod disese, nd mmmlin pests in coco in Mlysi. Bull. Entomol. Res. 82: Khoo, K.C. 21. Controlling coco pests in Southest Asi with the lck coco nt. nts.htm. Perfecto, I. nd A. Cstineirs Deployment of the predceous nts nd their conservtion in groecosystems. p In P. Bros (Ed.). Conservtion Biologicl Control. Acdemic Press, Sn Diego. See, Y.A. nd K.C. Khoo Influence of Dolichoderus thorcicus (Hymenopter : Formicide) on coco pod dmge y Conopomorph crmerell (Lepidopter: Grcillriide) in Mlysi. Bull. Entomol. Res. 86: SttSoft Sttistic for Windows, 5.5. SttSoft Inc., Tuls. Wy, M.J. nd K.C. Khoo Reltionships etween Helopeltis theorome dmge nd nts with specil reference to Mlysin coco smllholdings. J. Plnt Protection in the Tropics 6: 1-11.