Plant Protection Quarterly Vol.21(1) 2006 39 pods damaged by the two seed feeders, C. ulicetana and E. ulicis, when both were Infestation of gorse pods by Cydia ulicetana and active. Exapion ulicis in the South Island of New Zealand Materials and methods Sample collection Craig R. SixtusA, R. Roderic ScottA and George D. HillB A Sites were chosen at Bainham, Onekaka, Bio-Protection and Ecology Division, PO Box 84, Lincoln University, East Takaka, (collectively later referred to New Zealand. as Golden Bay sites), Hinewai, McLeans B Agriculture and Life Sciences Division, PO Box 84, Lincoln University, Island, Trotters Gorge and Lake Ohau New Zealand. (Figure 1). Fifteen gorse bushes were chosen randomly at each site. The start- ing point was selected without conscious Summary especially where there is abundant au- bias and successive bushes were those at The effectiveness of Cydia ulicetana tumn fl owering (Hill et al. 1991). the distance given by a random number (Haworth) (gorse pod moth) and Exa- To improve control of gorse seed pro- (in metres), from the previous bush in a pion ulicis (Förster) (gorse seed weevil) duction, Cydia ulicetana (Haworth) (previ- straight line. at reducing annual production of gorse ously referred to as C. succedana (Denis & Gorse growth was dense at Hinewai, seed (Ulex europaeus L.) was compared Schiffermüller) (Fowler et al. 2004), was Onekaka and Trotters Gorge and formed at six sites over the South Island, New introduced into New Zealand in 1992 after a complete canopy, whereas at the other Zealand. The highest percentage (45%) completion of research into host-plant spe- sites gorse bushes were scattered in pas- of pods damaged by C. ulicetana was at cifi city in 1991 (Hill and Gourlay 2002). In ture. Sites were inspected monthly and East Takaka in May 2002, and E. ulicis, Europe, C. ulicetana has two generations a samples of 25 mature seed pods were col- was most effective at Lake Ohau (76%) year and larvae attack gorse seed in spring lected, when available. in January 2003. The percentage of dam- and the seed of other Ulex species in the aged pods fluctuated. Although there autumn. The moth is now abundant in Climate data were climatic differences among sites, New Zealand at the original release sites From March 2002 to March 2003 the month- they did not differ signifi cantly in the such as Darfi eld, Canterbury (Hill and ly average daily minimum and maximum proportion of gorse pods damaged by C. Gourlay 2002). temperatures (Figure 2) were recorded at ulicetana. However, there was a signifi - Cydia succedana lays its eggs on Ulex, all sites using thermo-hygrographs. cant seasonal difference (P <0.001) over Genista, Sarothamnus or Lotus plants in Eu- all observations and C. ulicetana was re- rope (Emmet 1988) and it is assumed C. Plant phenology sponsible for damaging approximately ulicetana does the same. In New Zealand, In New Zealand, gorse grows profusely in 20% of gorse pods. C. ulicetana completes both generations in many areas and a mean canopy height of Key words: Gorse, Ulex europaeus, the pods of Ulex europaeus and larvae are 4 m has been recorded with a maximum gorse pod moth, Cydia ulicetana, gorse active from September to June, depend- height of 7 m (Lee et al., 1986). The fl ower seed weevil, Exapion ulicis, effective- ing on the climatic conditions. The second is pea-like, approximately 2 cm long and ness. generation over winters fully-fed from bright yellow with a distinct keel con- October (in Europe) in a cocoon spun in taining the stamens and stigma. A single Introduction leaf-litter or the soil and pupates in the branch 50 cm long can produce 500–1000 Gorse (Ulex europaeus L.) has been a nox- cocoon. fl owers over a two to three month fl ower- ious weed in New Zealand since 1900 The insect is bivoltine in the south of ing period (Markin and Yoshioka 1996). (Moss 1960). To combat gorse spread, vari- the British Isles and is univoltine in Scot- ous insect biological control agents have land (Bradley et al. 1979). A similar trend Monthly sampling been introduced (Harman et al. 1996). Exa- has occurred in New Zealand where the Monthly observations were made of the pion ulicis (Förster) was fi rst released at agent is univoltine in the cooler, southern proportion, in 5% steps, of the phenologi- Nelson and Alexandra in February 1931 regions, e.g. Mackenzie Basin, and bivolt- cal stage of the gorse (i.e. the presence of and the insect is now widespread through- ine further north, e.g. Golden Bay. buds, fl owers, green pods and black pods) out New Zealand (Kuschel 1972). The purpose of these studies was two- at the same time as sampling of pods. The Female E. ulicis weevils lay eggs fold: fi rst, to ascertain the percentage of number of damaged pods in each sample through a hole chewed in the side of an gorse pods damaged by C. ulicetana; sec- of 25 pods was recorded as well as the un- immature gorse pod. This hole closes and ond, to determine the percentage of gorse damaged seed. An accurate count of the the hatching larvae proceed to destroy the seed in the pod (Hill et al. 1991). The adults are released from the pod when /NEKAKA it dehisces (Hoddle 1991). Exapion ulicis "AINHAM larvae cannot move between gorse pods and must complete their development in %AST 4AKAKA -C,EANS )SLAND the pod chosen by the adult female. Lar- val survival depends on adult oviposition (INEWAI 2ESERVE (Hoddle 1991). ,AKE /HAU Weevils oviposit only in the spring. In many areas, the low availability of ovipo- 4ROTTERS 'ORGE sition sites, coupled with intense pressure on female weevils to deposit eggs, leads to a very high percentage of weevil-infested pods in the spring and they can infest up to 90% of immature pods. However, seed Figure 1. Map of the South Island, New Zealand showing the location of produced at other times escapes attack, each sample site. 40 Plant Protection Quarterly Vol.21(1) 2006 number of seeds damaged by Cydia uli- cetana or Exapion ulicis was not obtained, however, calculating the average number of seeds in undamaged pods and multi- plying this fi gure by the number of dam- aged pods gave an estimate of the number of seeds that were damaged each month.
"AINHAM /NEKAKA
Data collection and statistical analysis !VERAGE DAILY TEMPERATURE %AST 4AKAKA Data were transformed using an arc sine (INEWAI 4ROTTERS 'ORGE transformation of percentages for analysis ,AKE /HAU of variance with the SYSTAT 9 package. To compare among results at the differ- -AR !PR -AY *UN *UL !UG 3EP /CT .OV $EC *AN &EB -AR ent sites, two analyses were done. The -ONTHS fi rst was among the Golden Bay sites over Figure 2. Monthly average daily temperatures at each experimental site. the whole study because this area differed in that plants bore seed pods for most of the year and Cydia ulicetana larvae were present for most of the year (Figure 3). The A second analysis was a comparison among all the sites over summer when seed pods were commonly produced and both in- IN 'OLDEN "AY sects were present. Results
Insect activity OF GORSE PODS DAMAGED BY %XAPION ULICIS Cydia ulicetana larvae were active at the -AY *UNE *ULY !UG 3EPT /CT .OV $EC *AN &EB Golden Bay sites throughout the winter B "AINHAM until September to early October 2002 /NEKAKA (Figure 3), when the damaged pods were %AST 4AKAKA sampled. There was no winter activity at the southern sites. Larvae began feeding IN 'OLDEN "AY again during mid-November 2002 at the Golden Bay sites. There was no activity at the southern sites during March 2002. Lar- val activity started in December 2002 and ended in January (Lake Ohau) and March OF GORSE PODS DAMAGED BY #YDIA ULICETANA -AY *UNE *ULY !UG 3EPT /CT .OV $EC *AN &EB 2003 at Trotters Gorge and Hinewai (Fig- -ONTH ure 4). Exapion ulicis larvae were active in November-December in Golden Bay and Figure 3. Percentage (±SD) of gorse pods damaged by insects at the Golden in December-January further south (Fig- Bay sites. a) Exapion ulicis, b) Cydia ulicetana. ure 3).
Flowering time enter another pod. The highest percent- climates, buds develop and fl ower in au- Gorse fl owers and pods were present at age (45%) of pods damaged by C. ulicetana tumn and winter, while in cooler climates, all Golden Bay sites throughout the year was at East Takaka (Figure 3) in May, (es- peak fl ower production can continue af- presumably due to the milder climatic timated to be 70 seeds per 25 pods) and ter the peak of weevil oviposition activity. conditions (Figure 2). At Hinewai, Trotters 76% by E. ulicis at Lake Ohau in January Gorse reproductive development does not Gorge and Lake Ohau gorse fl owered only (Figure 4), (estimated to be 39 seeds per 25 appear to be affected by day length, but by once and no gorse fl owered in autumn – pods). climate; seed production is seasonal (Hill winter (April – September). At Lake Ohau, Although there were climatic differenc- et al. 1991). The position on the hillside gorse fl owered from September – October es between sites, there was no signifi cant affected plant phenology. The pod abun- with some fl owers still present in Novem- difference between sites in the proportion dance peaked in autumn at the top and in ber. A high percentage of sample plants of gorse pods damaged by C. ulicetana. spring at the bottom (Suckling et al. 1999). bore green pods in November. At Lake However, there was a signifi cant differ- Ripe seed production was continuous at Ohau, mature pods were present in De- ence between collection dates (P <0.001). the Golden Bay sites from May 2002 to Jan- cember and January, although there were The site by time interaction was not signif- uary 2003. Production at the southern sites few pods still unopened in January. Gorse icantly different (P = 0.073). At Hinewai, was limited to December 2002 to March seed production in undamaged pods was Trotters Gorge and Lake Ohau, where the 2003; presumably the difference was due significantly different between Golden insects were active for a shorter period, to the lower temperatures at the latter sites Bay and the other sites, with an average of there was a shorter period of time when (Figure 2). 8.4 ± 1.9 SD seeds per pod, at the Golden the monthly average maximum tempera- Seed production per pod differed. Bay sites compared with an average of 4.1 ture was above 10°C (Figure 2), though The Golden Bay sites had a higher aver- ± 2.5 SD seeds per pod at the other sites ample pods were available (Sixtus 2004). age number of seeds per pod. The lower (P <0.001). number of seeds per pod is attributed to Discussion the cooler and shorter growing season at Pod damage and climate Gorse seasonality the southern sites. The soil fertility could Cydia ulicetana larvae destroy all of the Gorse reproduction was limited by both also contribute to the difference in the seeds within a pod and then move out and altitude and temperature. In warmer number of seeds per pod. Golden Bay sites Plant Protection Quarterly Vol.21(1) 2006 41
larval activity, probably due to the warmer A climate (Figure 2). Damage was very sea- "AINHAM sonal and it appears that the two periods /NEKAKA when the most damage occurred were %AST 4AKAKA (INEWAI May 2002 and December 2002 – January 4ROTTERS 'ORGE 2003. These followed the main periods of ,AKE /HAU gorse pod production of April and No- vember 2002 respectively. Cydia ulicetana has better phenological synchronicity with gorse than E. ulicis, especially in Golden Bay. Cydia ulicetana has two generations a year in Europe (Bradley et al. 1979) and it appears that this is also the case in Golden Bay where the moth generations were syn- chronized with the gorse pod production (Figure 3a). From this work, based on the percentage of infested pods, it appears that
#YDIA ULICETANA 0ERCENTAGE OF GORSE PODS DAMAGED BY IN THE PERIOD WHEN BOTH GORSE SEED AGENTS ARE ACTIVE the autumn generation of C. ulicetana had a greater impact at the Golden Bay sites. .OVEMBER $ECEMBER *ANUARY &EBRUARY -ARCH Suckling et al. (1999) found that, at Dar- fi eld, Canterbury, there was no synchrony -ONTH between C. ulicetana and gorse during its recorded second burst of seed production in March/April 1997. B Combined agent activity During November 2002 – March 2003, both insects were active in the areas surveyed. The results show that C. ulicetana was more effective for more months in warm regions (Golden Bay) than in the cooler, southern regions. During February and March 2003, there was a high percentage of gorse pods damaged by C. ulicetana at Hinewai Reserve (Figure 4). Exapion ulicis damaged a higher proportion of seed pods than C. ulicetana at the southern sites (Fig- ure 4). This may have been because fewer C. ulicetana moths were present (C.R. Six- %XAPION ULICIS 0ERCENTAGE OF GORSE PODS DAMAGED BY IN THE PERIOD WHEN BOTH GORSE SEED AGENTS ARE ACTIVE tus, personal observation). Either E. ulicis is better adapted to colder temperatures .OVEMBER $ECEMBER *ANUARY &EBRUARY -ARCH than C. ulicetana or it may simply be that C. ulicetana has not yet had suffi cient time -ONTH to build up its population in the south be- cause of the lower temperatures, and the Figure 4. Mean percentage (±SD) of gorse pods damaged at all sites during lack of direct releases at the Trotters Gorge summer (November – March) by a), Cydia ulicetana b) Exapion ulicis. and the Lake Ohau sites. When there were insuffi cient gorse pods, C. ulicetana larvae fed on other legumes were on productive farmland and would releases at Lake Ohau (A.H. Gourlay per- such as Scotch broom (Cytisus scoparius) be receiving fertilizer regularly. The south- sonal communication). There have been and Russell lupin (Lupinus polyphyllus × L. ern sites, especially Hinewai and Trotters no releases at Trotters Gorge but there arboreus) (Paynter et al. 2004). The southern Gorge, would not receive any fertilizer as have been several releases at the nearby sites produced seed only in spring. Thus they were in reserves. Mt Herbert Forest (1992, 1994, 1995) (A.H. seeding plants could have experienced an Gourlay personal communication). Cy- increased attack from both E. ulicis and Establishment of Cydia ulicetana dia ulicetana has now been released suffi - C. ulicetana, providing the insects’ emer- Cydia ulicetana activity depends on tem- ciently long enough in most of the areas to gence was synchronized with gorse seed perature (Sixtus 2004). The cooler tem- become well established. However, it ap- production. If synchronization is absent peratures at the southern sites may have pears that the winter cold in the southern then there is a possibility that the effect of caused C. ulicetana to be univoltine as it is regions (Figure 2) may have slowed the the pod herbivores would be reduced. in Scotland, whereas at warmer sites, like insect’s establishment and spread (C.R. At Lake Ohau, a high percentage (76%) Golden Bay, it is bivoltine as in the south- Sixtus personal observation). of gorse pods was damaged by E. ulicis ern British Isles (Bradley et al. 1979). during the two months that the gorse was Cydia ulicetana was released in Golden Impact of Cydia ulicetana producing mature pods. In the same period Bay in 1993 (R.G. Broadhurst personal There was seasonal variation at all sites in C. ulicetana damaged less than 10% of pods communication). It was initially released both gorse seed production and C. ulice- (Figure 4). The results indicate that >10% on Banks Peninsula in 1992 (Hill and tana activity. Cydia ulicetana larvae dam- of gorse seed escapes predation by both of Gourlay 2002). The Mackenzie Basin saw aged a moderate percentage of pods in the insects even though gorse seed is only releases in 1995–96 but there were no Golden Bay (Figure 3a), because of greater produced once in summer/autumn. This 42 Plant Protection Quarterly Vol.21(1) 2006 maybe suffi cient to maintain gorse plant stage where the number of pods is a limit- against gorse in New Zealand. Proceed- replacement and continue gorse spread. ing resource. The competition between the ings 10th International Symposium on Sixtus et al. (2003a) found 80% of undam- seed-feeders is likely to intensify when the Biological Control of Weeds, ed N.R. aged gorse seed was viable. At Hinewai percentage of infested pods is high. Fur- Spencer, pp 909-17 (Montana State Uni- Reserve and Trotters Gorge, both E. ulicis ther research is required to confi rm if C. versity, Bozeman, USA). and C. ulicetana were active from Janu- ulicetana larvae do consume E. ulicis larvae Hill, R.L., Gourlay, A.H. and Martin, L. ary to March. However, C. ulicetana dam- and, if so, how many E. ulicis larvae are (1991). Seasonal and geographic varia- aged less than 10% of gorse pods while consumed. tion in the predation of gorse seed, Ulex it was active at Trotters Gorge (Figure 4). Knowledge of the amount gorse seed europaeus L., by the seed weevil Apion At Hinewai Reserve, most damage by C. damaged could also be used to help de- ulicis Forst. New Zealand Journal of Zool- ulicetana was in March when 30% of pods cide whether it was necessary to introduce ogy 18, 37-43. were damaged (Figure 3). Exapion ulicis further gorse seed-feeding biological con- Hoddle, M.S. (1991). The reproductive bi- damaged approximately 20% of pods at trol agents into New Zealand. ology of Apion ulicis (Förster) (Coleop- both Hinewai Reserve and Trotters Gorge tera: Apionidae). Unpublished Master (Figure 4), i.e., half the seed production Acknowledgements of Science in Zoology thesis, University escaped damage by the two insects. Ross Patching, Ross and Jenny Haldane, of Auckland, New Zealand, 136 pp. Hill and Gourlay (2002) showed that Les Sixtus, Hugh Wilson, ECan (McLeans Kuschel, G. (1972). The foreign Curculio- C. ulicetana larvae consume any E. ulicis Island), Department of Conservation nidae established in New Zealand (In- larvae they encounter in pods. In a study (Oamaru) and Hamish Weatherall for al- secta: Coleoptera). New Zealand Journal of gorse seed predation in England, less lowing us on their land to sample gorse of Science 15, 273-89. than 20% of U. europaeus pods occupied by bushes. In addition, we would also like Lee, W.G., Allen, R.B. and Johnson, P.N. insects contained both species (Hill 1982). to thank Ross and Jenny Haldane, Lorna (1986). Succession and dynamics of In this study, less than 5% of sampled pods Langford, Felix Langford, Ross Patching, gorse (Ulex europaeus L.) communi- contained both insects. However, several Julie Gill, Hugh Wilson and Elaine Fouhy ties in the Dunedin ecological district, pods had C. ulicetana larvae, or they had of NIWA for providing weather records. South Island, New Zealand. New Zea- been present, but there was no trace of E. This work would not have been possi- land Journal of Botany 24, 279-92. ulicis though E. ulicis larvae were present ble without the monetary assistance of the Markin, G.P. and Yoshioka, E.R. (1996). in other pods. It is possible that E. ulicis Lincoln University Postgraduate Research The phenology and growth rates of the larvae may have already been consumed. Fund, Lincoln Freemasons and Landcare weed gorse (Ulex europaeus) in Hawai’i. For C. ulicetana to have a signifi cant im- Research. Newsletter of the Hawaiian Botanical Soci- pact on gorse seed production, it would ety 35, 45-50. have to reduce seed production to a level References Moss, G.R. (1960). Gorse, a weed problem where seedling recruitment was severely Bradley, J.D., Tremewan, W.G. and Smith, of thousands of acres of farm land. New limited and the lifetime reduction of the A. (1979). British Tortricoid moths; Tor- Zealand Journal of Agriculture 100, 561- seed bank would have to be great. Hill tricidae: Olethreutinae. 336 pp. (The 7. and Gourlay (2002) and Sixtus (2004) Royal Society, London). Paynter, Q.E., Fowler, S.V., Gourlay, A.H., found that, after the larvae consumed all Emmett, A.M. (1988). A fi eld guide to the Haines, M.L., Harman, H.M., Hona, of the seed in one pod, they exited through smaller British Lepidoptera. (British S.R., Peterson, P.G., Smith, L.A., Wilson- round holes chewed in the side of the pod Entomological and Natural History So- Davey, J.R.A., Winks, C.J. and Withers, and sought another pod. Approximately ciety, London). T.M. (2004). Safety in New Zealand 80% of larvae transferred to a second gorse Fowler, S.V., Gourlay, A.H., Hill, R.L. and weed biocontrol: a nationwide survey pod, and 70% of those larvae pupated suc- Withers, T. (2004). Safety in New Zea- for impacts on non-target plants. New cessfully (Hill and Gourlay 2002). land weed biocontrol: a retrospective Zealand Plant Protection 57, 102-7. Gorse plants produce many pods each analysis of host specifi city testing and Rees, M. and Hill, R.L. (2001). Large-scale fl owering season and up to 80% of the the predictability of impacts on non- disturbances, biological control and the undamaged seed produced can be viable target plants. Proceedings XIth Inter- dynamics of gorse populations. Journal (Sixtus et al. 2003a). In addition, lightly national Symposium Biological Control of Applied Ecology 38, 364-77. weevil-damaged gorse seed, can still ger- of Weeds, eds. J.M. Cullen, D.T. Briese, Sixtus, C.R. (2004). An investigation of minate (25%) (Sixtus et al. 2003b). With D.J. Kriticos, W.M. Lonsdale, L. Morin the life history of the gorse pod moth such high seed viability, the seed feeders and J.K. Scott, pp. 265-70 (CSIRO, Can- (Cydia succedana) and its effectiveness would have very little impact on gorse berra). at reducing gorse (Ulex europaeus) seed plant abundance. However, if seedling Harman, H.M., Syrett, P., Hill, R.L. and production. Unpublished Master of Ag- survival is low, seed feeders can have a Jessep, C.T. (1996). Arthropod introduc- ricultural Science thesis, Lincoln Uni- dramatic impact on the gorse abundance. tions for biological control of weeds in versity, New Zealand, 136 pp. At highly disturbed sites, young pre-repro- New Zealand, 1929–1995. New Zealand Sixtus, C.R., Hill, G.D. and Scott R.R. ductive gorse plants dominate the gorse Entomologist 19, 71-80. (2003a). The effect of temperature and population. This reduces the seed popula- Hill, R.L. (1982). The phytophagous fauna scarifi cation method on gorse (Ulex eu- tion because very few plants produce seed. of gorse, Ulex europaeus L., and host ropaeus L.) seed germination. New Zea- Under these conditions seed feeders can plant quality. Unpublished doctoral land Plant Protection 56, 201-5. have their greatest impact (Rees and Hill thesis, Imperial College, University of Sixtus, C.R., Hill, G.D. and Scott R.R. 2001). A high C. ulicetana population could London. 268 pp. (2003b). Impact of Exapion ulicis (For- mean that a high percentage of pods were Hill, R.L and Gourlay, A.H. (2002). Host- ster) (Coleoptera: Apionidae) on gorse infested. This would intensify competition range testing, introduction, and estab- seed viability. New Zealand Plant Protec- between C. ulicetana and E. ulicis larvae. lishment of Cydia succedana (Lepidop- tion 56, 206-10. An aspect of competition between the two tera: Tortricidae) for biological control Suckling, D.M., Hill, R.L., Gourlay, A.H. insects is the number of pods available; of gorse, Ulex europaeus L., in New Zea- and Witzgall, P. (1999). Sex attractant- with a relatively low percentage of pods land. Biological Control 25, 173-86. based monitoring of a biological con- infected, there is an abundance of pods Hill, R.L., Gourlay, A.H. and Fowler, S.V. trol agent of gorse. Biocontrol Science and available. Thus we have not reached the (1999). The biological control program Technology 9, 99-104.