S. Afr.l. Z001. 1990,25(4) 245 Guild composition and seasonal distribution of insects on magnifica and P. laurifolia ()

M.G. Wright * and J.H. Giliomee Department of Entomology & Nematology, University of Stellenbosch, Stellenbosch, 7600 Republic of

Received 12 June 1990; accepted 21 August 1990

Insects were collected over a period of 12 months from and P. laurifO/ia by beating. Analysis of guild composition showed remarkable uniformity in proportion of species in different guilds on the two . Proportions of individuals in different guilds were not uniform, owing to a larger proportion of phyto­ phages being collected on P. magnifica. A distinct seasonal distribution in insect numbers was observed in four out of six study sites, with peak numbers being collected in summer.

Insekte is oar 'n periode van 12 maande vanaf Prot9a magnifica en P. laurifolia deur middel van uitklop versamel. Ontleding van gildestruktuur het noemenswaardige uniformiteit in verhoudings van spesies in verskillende gildes op die twee plante getoan. Verhoudings van individue in verskillende gildes was nie uniform nie, weens 'n groter proporsie fitofage insekte op P. magnifica. 'n Duidelike seisoenale verspreiding in insekgetalle is in vier uit die ses studiepersele waargeneem, met piekgetalle in die somer.

• To whom correspondence should be addressed at: Research Unit, Vegetable and Ornamental Research Institute, Private Bag, Elsenburg, 7607 Republic of South Africa

Very few insect community studies have been undertaken in distribution, and to discuss factors possibly influencing these the fynbos (the Capensis flora, sensu Taylor 1978), and they aspects. have concentrated either on specific taxa (e.g. Formicidae, Donnelly & Giliomee 1985, Koen & Breytenbach 1988), or Materials and Methods a limited range of taxa (e.g. Orthoptera & Hemiptera, Study sites were located in Du Toits Kloof (DTK) (33°

. Schlettwein & Giliomee 1987). The fust study which 1~05'E,

) 38'S/ both plants), the Cedarberg (CB) (32°25'S/

0 attempts to consider all the arthropods associated with

1 l~lO'E, both plants), Groeolandberg (GB) (34°07'S/ 0 selected fynbos plants is that of Coetzee (1989), who 2 1~08'E, P. magnifica) and Mont Rochelle (MR) (33°54'S/

d followed the 'knock-down' approach of Moran & South­

e 1~09"E, P. laurifolia). Population sizes of P. magnifica t a wood (1982). A major problem encountered in this study ranged from 50-250 plants, and P. laurifolia from several d

( was identifying species collected as the insects of the fynbos hundred to thousands of individuals. Populations of the two r e are poorly known (Coetzee 1989). species never overlapped. h s i The present study forms the basis for a comprehensive l Insects were collected on a monthly basis from DTK, MR b investigation of insect-plant interactions (viz. herbivory, u and GB and bimonthly from CB. Collections were made P seed predation and pollination, all reported separately), on from January 1988 to December 1988, except for MR, e h Pro tea magnifica Link and P. laurifolia Thunb., two fynbos t where collecting commenced in February 1988 and ended in

y species which are also used commercially for their flowers.

b January 1989. Collecting was done by beating (Smithers

d The two host plants have similar overall distribution 1981), rather than by the knock-down method, because of e t patterns in the south-, though P. magnifica has n the frequently windy conditions under which collecting was a a discontinuous distribution pattern within its overall range, r done. It was decided to sample 24 branches as a species­ g occurring only on mountain peaks above 1200 m (Rourke e sampling area analysis at DTK indicated that no new species c 1980). Protea laurifolia occurs below 1200 m (Rourke n

e were collected with further effort. Each of 24 randomly c 1980). Also, the architecture of the plants differs in that P. i l selected branches of both species were sampled each month

r magnifica seldom grows higher than 2 m, while P. laurifolia e by beating until insects no longer fell onto the 50 x 50-cm d may reach 8 m (Vogts 1982). This difference should allow a n sheet Insects thus collected were stored in 70% alcohol and

u test of Lawton's (1983) hypothesis that plant height is of

y significance in determining how many insect species utilize later sorted into morphospecies (hereafter referred to as a 'species'). In the case of phytophagous species, specialists w it. Unfortunately, certain factors complicate this test, as P. e t on taxa were consulted to obtain reliable identifications. a magnifica has a relatively disjointed distribution, and is G

subject to a more variable climate than P. laurifolia (Vogts Specimens were allocated to the following guilds: phytopha­ t e 1982). ges [divided into chewers (PC) and suckers (PS)] , predators n i b The of P. magnifica are broader than those of P. + parasitoids (P), ants (A), 'detritivores' (D, all non­ a

S detrimental feeders on the plant, viz., fungivores, detribls laurifolia and occur in a more dense concentration on y b branches. Both plants are evergreen. feeders), flower visitors (FLO) and tourists (1). These guilds d

e The aim of this study was to establish the guild structure were decided on after consulting Moran & Southwood c

u of insects on the two plants in question and their seasonal (1982) and Coetzee (1989). Guild allocations were made by d o r p e R 246 S.-Afr. Tydskr. Dierle. 1990.25(4)

Table 1 Insect taxa, and guild allocations of insects Table 1 Continued collected on Protea magnifica (Pm) and P. laurifolia (PI). Species are grouped by guild, then taxon. (PS = Taxon AcHRP Guild Pm PI phytophagous sucker, PC = phytophagous chewer, Sibillia sp. 1989 PC • FLO flower visitor, P predator/parasite, A ant, = = = SmicroflJ% sp. T = tourist, D = detritivore) 1946 PC • • Gen. et sp. indet. 1945 PC • Taxon AcHRP Guild Pm PI Cryptorhynchinae (Gen. indet.) 1948 PC • Apiooidae (Gen. et sp. indeL) 1954 PC • Hc:miptera Lepidoptera Miridae (Gen. et sp. indet.) 764 PS • • Geometridae (Gen. et sp. indeL) 1947 PC • • Lygaeidae Psychidae (Gen. et sp. indeL) 1942 PC • • Oxycareflus macula/us 624 PS • • Lirnacodidae (Gen. et sp. indeL) 10 PC • • MachituJenuls diplop/erus 208 PS • • Sphingidae (Gen. et sp. indeL) 1960 PC • Caprhobia similis 262 PS • • Pyralidae (Gen. et sp. indet.) 1954 PC • PDe4f11Nu velax 237 PS • • Bos/ra cOflSpicualis 329 PC • Gen. et sp. indeL 1963 PS • • Phyllocnistidae Aradidae 1965 PS • Phyl/ocflis/is sp. 698 PC • • Pentatomidae Family indeL (Larvae) 1937 PC • Afi/es/iopsis IIQriega/a 545 PS • • Family indeL (Larvae) 1939 PC • • Gen. et sp. indeL 680 PS • • Family indeL (Larvae) 1981 PC • Cydnidae Family indet. (Larvae) 1983 PC • Dimegis/us fimbria/us 373 PS • • Coleoptera Membracidae Scarabaeidae Gargaw sp. 529 PS • • Tricos/elha capeflSis 61 FLO • Psyllidae (Gen. et sp. indeL) 1990 PS • • Platycltelus sp. 1132 FLO • • Diaspididae a. rysornelidae LedIJspis dis/iflc/a 926 PS • • ChiTodica ca/coplera 7CYJ FLO • • Gen. et sp. indeL 1937 PS • Staphylinidae Gen. et sp. indeL 926 PS • • Ph/oeoflonlMS sp. 725 FLO • •

. Thysanoptera ) 1953 PS • • Nitidulidae 0 Phurnotodea (Gen. et sp. indet.) 1 1959 PC • Pria ciMrasceflS 713 FLO • • 0

2 Onhoptera Helodidae

d Tettigooidae 1956 PC Helodes sp. 1950 FLO e • • t a Coleoptera Mant.odea (Gen. et sp. indet.) 663 P • d (

Phalacridae Hemiptera r e Olibrus aeroilJ/us Anthocoridae 134 P • h 717 PC • • s i Discolomidae Reduviidae 1994 P l • b No/iophygus sp. Neuroptera u 1952 PC • P Ouysomelidae a.rysopidae 1196 P • e h Prasoidea sericea Coleoptera t 521 PC • •

y Xe1lOOmOrphus sp. Cleridae 258 P • •

b 603 PC • • 720 d RhabdocMoraM sp. 708 PC • • Melyridae P • e t Gen. et sp. indeL 1987 PC • Melyridae 1159 P • • n a 8uprestidae Carabidae r g Spheflop/era sp. A 1938 PC • XefliJefIIU lessalalus 625 P e • • c SpheflOfi/era sp. Curujidae n 8 1951 PC • e c Anthribidae PhycoflOmJIS Iricolor 721 P i • • l

r Holoph/oeus nigel/us 1992 PC • PhycollOmJlS pa/idus 722 P • e d Gen. et sp. indeL 1993 PC • Nitidulidae n u Cybocephalus sp. 825 P Cerambycidae • • y a Gen. et sp. indeL 1744 PC • • Coccinellidae w Rlryzobius javeli 766 P e Cryptophagidae • • t a Micrambe /efluicormis 760 PC • Telsimia lelraslricta 863 P • • G

t Curculionidae Rhyzobius sp. 1995 P • • e n A/rolep/ops coe/uei SCYIMUS morreleli 617 P • • i 664 PC • • b ErelfllUU nr. iJlriJIus Adonia variegiJIa 613 P a 735 PC • • • S Steflotypus sp. 1949 PC • Cheil~fIts 1/UIata 7'12 P • • y b Hipporhillll.f sp.A 1961 PC • Hymenoptera d e Hipporhillus sp.8 Chalcididae c 1941 PC • u ElliMrts liMicoilis 1912 PC Hockeria sp. 1970 P d • • o r p e R S. Afr. J. Z001. 1990,25(4) 247

Table 1 Continued

Taxon AcHRP Guild Pm PI

Dirhillus sp. 972 P .. A Gat. et sp. indeL 280 P .. 24918S Iclmcumonidae P. magnifica P. laurifolia N·1353 N • 821 Gat. et sp. indeL 1304 P .. Vespidae Polisles sp. 1985 P .. Figure 1 Guild composition of insects (number of individuals) on Eupehnidae two Protea species. (A = ants; PC = phytophagous chewers; Gen. et sp. indeL 740 P .. Fill = flower visitors; D = detritivores; P = predators & parasi­ Funily indet. 1969 P .. toids; T = tourists; PS = phytophagous suckers.) Funily indet. 1968 P .. .. Funily indet. 379 P .. Funily indet. 1986 P .. Fonnicidae Camponolus maclllatus 1976 A .. Camponolus fliveoselosus ll50 A .. .. Camponolus folvopiJosus 1979 A .. Camponolus sp.l 1974 A .. .. A 1375" Camponolus sp.2 1977 A .. P. magnifica P. laurifolia N ·85 CampollOlus sp.3 1996 A .. N·89 CampollOlus spA 1980 A .. Cremalogasler perillgueyi 653 A .. .. Figure 2 Guild composition of insects (number of species) on two Cremalogasler lingm.ei 694 A .. .. Protea species. (PC = phytophagous chewers; PS = phytopha­ Cremalogasler sp.l 1972 A .. gous suckers; Fill = flower visitors; P = predators & parasi­ Cremalogasler sp.2 1973 A .. toids; A = ants; T = tourists; D = detritivores.) Alloplolepis cuslodiens 1810 A .. .. Acalllholepis capensis 1809 A .. ..

Table 2 Tests for uniformity of proportion of guilds . Pheidole prob. capensis 1820 A .. .. )

0 1975 A .. .. between Protea magnifica and P. /aurifo/ia using 2X2

1 Plagiolepis sp.

0 contingency tables with Yates correction. Data ana­

2 Myrm.ecaria fligra 1819 A ..

d Iridomyrmu humiJis 656 A .. lysed are for number of species and number of indivi­ e t duals from Figure 1 and Figure 2. (Phy. Phyto­ a Microcoryphia = d

( 1940 D .. phages; P = Predators & parasitoids; All = all guilds

Funily indeL r e Collembola except phy.) h s

i Arthropleona 1970 D .. l Comparison -,( P (df = 1) Conclusion b Thysanura u

P Lepismatidae 1971 D .. .. No. spp.: Phy.: P 0,05 NS Unifonn e Phy.: All h Psocoptera 0,01 NS Unifonn t

y Funily indet. 935 D .. .. No. inds.: Phy.: P 33,48 < 0,001 Non-unifonn b BIattaria : Phy.: All 90,42 < 0,001 Non-unifonn d e t BIauidae 751 o .. .. n D .. .. a BIaberidae 1988 r g

Coleoptera level was possible, so morphospecies were allocated e c Tenebrionidae 1991 D .. accession numbers only (AcHRP = Accession number, n e Lalhridiidae 1610 D .. ..

c Horticultural Research, ). All insects collected were i l Lalhridiidae 899 D .. deposited in the collection of fynbos insects of the VOPRI r e Hemiprera at Elsenburg (near Stellenbosch), or with the National d n Ceroopidae (Gen. indeL) 1966 T .. Collection of Insects (pPRI), Pretoria. u 1967 T .. .. y Fulgoroidea (Var. spp.) An analysis (according to Moran & Southwood 1982), a Cicadellidae (Var. spp.) 1962 T .. .. w was done to establish whether the phytophage guild was e t Aphididae (Gen. indeL) 844 T .. uniformly represented on both plants with respect to a

G P1ecoptera

predators + parasitoids and all other guilds collectively. t

e Nemouridae (Gen. indeL) 1964 T .. n i Diptera (Var. funs) Various T .. .. b Results a S All taxa collected are listed in Table 1, with their guild y b

allocations and host planL d e consulting literature, or by personal observations of feeding The proportions of individuals and species in different c u habits. In the case of many taxa, no identification to species guilds are given in Figures 1 and 2 respectively. The number d o r p e R 248 S.-Afr. Tydslcr. Dierk. 1990,25(4)

No.insects No. insects .40 r------. ..0 ~------. B '00 '00 80

80

40 40

10 o o JAN fE8 MAR APR MAY JUN .lUI. AUG BEP OCT NOV DEC JAN fEe MAR APA MAY "UN JUL AUG ••P OCT MOV DEC Month Month _ Ants lilt Preds + pars _ Ants _ Preds + pars

~ Phytophages ~ Detritivores ~ Phytophages ~ Detritivores

No. insects No.insects .40,--______---, ..0.------, D ..0 c '00 .00 eo 80 eo eo 40 40

JO 20 o .tAN fEe MAR APR MAY JUN JUL AUG BEP OCT NOV DEC Month _ Ants _ Preds + pars _ Ants _ Preds + pars

~ Phytophages ~ Detritivores ~ Phytophages ~ Detritlvores

Figure 3 Seasonal distribution of insects on two Protea species. A: P. magnifica. DTK & OB; B: P. laurifolia. MR & DTK; C: P. magnifica. CB; D: P.lawifolia. CB.

.

) of species collected in each guild for the two plants studied. inclusion of a number of isolated clumps of P. magnifica, 0

1 showed remarkable uniformity. with possibly dissimilar insect faunas. may have led lO the 0 2

Number of species of phylOphages relative lO other guilds plant having as many species as the more evenly distributed d e P. laurifolia. Even interplant distance has been found lO

t was uniform for the two plants. but not the number of a influence similarity of insect faunas in Bomean forests

d individuals (Table 2). ( (SlOrk 1987b). and a similar phenomenon may occur with P. r A distinct seasonal distribution was apparent for both e magnifica. with its widely separated patches. This aspect h plants at DTIC, MR and GB but appeared lO be different at s i l CB (Figure 3). Proportions of predalOrs + parasitoids : deserves further scrutiny. b u phylOphages appeared lO be lower for the P. magnifica site The similarity (uniformity) of proportions of phytophages P to other guilds is in accordance with fmdings of Moran & e at CB than for the other sites and P.laurifolia (Figure 3). h

t Southwood (1982). Non-uniformity of numbers of individu­

More phytophagous insects were also collected from the CB y als might be attributable to climate or incorrect guild alloca­ b P. magnifica than in other sites (Figure 3).

d tions. For example. the Phalacridae are placed in different e t Discussion guilds by Moran & Southwood (1982); Stork (1987a); Louw n a r The observed similarity in the number of insect species on (1988) and Coetzee (1989). and the possibility thus exists g

e the two plants is to be expected in terms of the similar that they are incorrectly placed in the present study. c n distribution of the two plant species (Lawton & Schroder The number of species and individuals collected (particu­ e c

i 1977). though one might have expected a lower species larly tourists. detritivores and flower visilOrs) were consider­ l

r richness for P. magnifica, in view of the higher altitudes ably lower than the numbers recorded on five Proteaceae by e d where it occurs. as was found by Wolda (1987). The Coetzee (1989). This can probably be attributed to the n u

differences in architecture (e.g. height) of the two plants different collecting method used in the present study. viz. y a also provide grounds to expect a difference in the number of beating. as well as the often inclement weather conditions w e species utilizing each (Moran 1980; Lawton 1983). The (hot/cold, wind) under which collecting was done. Beating t a broad-leaved nature of P. magnifica might cause it to host will tend to collect large. slow-moving insects better than G

t more species than one might expect of a plant with its lower small flying ones. whereas the knock-down method used by e n height and disjointed distribution. considering that Moran & Coetzee (1989) may collect more equable proportions of i b

a Southwood (1982) have shown that broad-leaved trees both types of insects. The cup-shaped form of the inflores­ S harbour more species than narrow-leaved ones. Furthermore, cences of the two study plants probably led lO few hymen­ y b

it has been shown by Lawton & Strong (1981). that small, opteran flower visilOrs being collected. d e local clumps of plants support less species than larger The seasonal distribution of insects at DTIC, GB and MR c u clumps. but in the present study it is possible that the appeared lO be strongly correlated with winter (cold) and d o r p e R S. Afr. J. Z001. 1990,25(4) 249

summer conditions. This is in conttast with Coetzee's vegetation in the Jonkershoek Valley. J. eN. Soc. sth. Afr. 48: (1989) findings that no seasonal distribution of insects 247-257. (except for ants and flower visitors) occurred on his five KOEN, J.H. & BREYrENBACH, W. 1988. Ant species richness Proteaceae. The sites used in his study were, however, all of fynbos and forest ecosystems in the southern Cape. S. Afr. located at lower altitudes than the sites used in the present J. Zool. 23: 184-188. study, and all have a less harsh winter than the sites in the LA wroN, J.H. 1983. Plant architecture and the diversity of present study. The importance of climate to insects has been phytophagous insects. AM. Rev. Enlomol. 28: 23-39. emphasized by Caugbley & Lawton (1981). LAwroN, J .H. & SCHRODER, D. 19n. Effects of plant type, In conclusion, results obtained in this study show that the size of geographic range and taxonomic isolation on number proportion of phytophagous insects to other guilds is of insect species associated with British plants. Nature 265: 137-140. uniform on P. magnifica and P. laurifolia, as is the case on LAwroN, J.H. & STRONG, D.R. Jr. 1981. Community patterns other plants (Moran & Southwood 1982). Distinct seasonal and competition in folivorous insects. Am.. Nal. 118: 317-338. changes in insect abundance occurred in most study sites, WUW, S. 1988. AIboreal Coleoptera associated with Leucosidea probably owing to climatic variability associated with high sericea (Rosaceae) at the Golden Gate Highlands National altitudes. Park. Koedoe 31: 53-70. MORAN, V.C. 1980. Interactions between phytophagous insects Acknowledgements and their OplUllia hosts. Ecol. E1IUJmol. 5: 153-164. The Chief Director, Chief Directorate of Nature and MORAN, V.C. & SOUTHWOOD, T.R.E.1982. The guild Environmental Conservation (Cape Provincial Administra­ composition of arthropod communities in trees. J. Anim. Ecol. tion) is thanked for permission to work on land under his 51: 289-306. control. R.G. Oberprieler, B. Grobbelaar, N. Verheijen, J.M. ROURKE, J.P. 1980. The proteas of southern Africa. Tafelberg, Millar, H.D. Brown, M. Johnson, V.M. Uys (all of the CapeTown. National Collection of Insects, PPRI), H. Robinson (S.A. SCHLETIWElN,C.H. & GIUOMEE,J.H. 1987. Comparison of Museum) and H. Geertsema (University of Stellenbosch) are insect biomass and community strucbJre between fynbos sites thanked for the identification of insects. This work was done of different ages, with particular reference to ants, leafhoppers as a part of a project of the Vegetable & Ornamental Plant and grasshoppers. AM. Univ. SteU. Series A3, Vol 2: 1-76. Research Institute (GS 2431/20/1(3). SMITHERS, C. 1981. Handbook of insect collecting. Delta Books, Johannesburg.

. STORK, N.E. 1987a. Guild structure of arthropods from Bornean ) References 0 rain forest trees. Ecol. ENomol. 12: 69-80. 1

0 CAUGHLY,G. & LAWfON,J.H. 1981. Plant-herbivore STORK, N.E. 1987b. Arthropod faunal similarity of Bornean rain 2 systems. In: Theoretical ecology: Principles and applications. d forest trees. Ecol. Enlomol. 12: 219-226. e t (Ed.) R.M. May. 2nd edition. Blackwell Scientific Publishers,

a TAYWR, H.C. 1978. Capensis. In: Biogeography and ecology d Oxford. ( of southern Africa. pp. 171-229. (Ed) M.J.A. Werger. Junk r e COETZEE, J.H. 1989. Arthropod communities of Proteaceae Publishers, Holland. h s with special reference to insect-plant interactions. Ph.D. i VOGTS, M. 1982. South Africa's Proteaceae. C. Struik, Cape l b thesis, University of Stellenbosch, Stellenbosch. 117 pp. Town. u

P DONNELLY, D. & GILIOMEE, J.H. 1985. Community structure WOLDA, H. 1987. Altitude, habitat and tropical insect diversity. e

h of epigaeic ants (Hymenoptera: Formicidae) in fynbos Bioi. J. liM. Soc. 30: 313-323. t y b d e t n a r g e c n e c i l r e d n u y a w e t a G t e n i b a S y b d e c u d o r p e R