0707-B2

Insect diversity in forests and pastoral areas in Whangamata New Zealand

John A. McLean ^1,2,3 and Diane C. Jones ²

Abstract

Malaise traps were set out for four weeks to compare the insect populations in Kanuka bush, fern bush, a 25 year old mature radiata pine plantation, a young 6 year old pine plantation as well as in adjacent hayfields, grassed foreshore areas and on coastal sand dunes. Greater numbers of beetle species were found in the forested sites than in the pastoral and urban sites. The mature pine plantation had the greatest number of beetle species. This was probably related to the wide range of shade tolerant native species that had become established on the forest floor among the widely spaced radiata pine trees. The forested locations also had higher degrees of endemism of the fauna. Habitats maintained by constant human activities had more adventive species. High biomasses of Diptera associated with man-maintained habitats.

Introduction

This research was undertaken to provide biological data, as indicated by the abundance of insects, to compare the fauna of various landscape types in the Whangamata area on the Coromandel Peninsula of New Zealand. The study was matched with a behavioural study that evaluated local and visitor input on what they perceived as natural or modified landscapes (Fairweather and Swaffield, 1999). The study area was selected for the comparison of insect biodiversity among sand dunes, developed foreshore areas and pasture, as well as comparisons among a 6 year old young radiata pine, Pinus radiata, plantation, a mature 23 year old radiata pine plantation, fern bush and kanuka bush.

Early settler records from the 18^th century suggest a tall native tree forest once extended to the coastline in this part of New Zealand. The hills were clad by a kauri, Agathis australis, forest while the coastal swampy areas were dominated by kahikatea, Dacrycarpus dacrydioides. After waves of colonization by Polynesians around 600 years BP and Europeans 200 years BP, the land showed the effects of major resource extraction, especially of timber and kauri gum. Adams (1883) rated the eastern side of the Coromandels as "the most ruined and disfigured part of New Zealand."

Against this background, settlers at the beginning of the 20^th century set about establishing an agricultural economy with cattle and sheep farming. The trials and tribulations of this period are well reported by Williamson (1988). Forests of exotic pine species were planted and the mosaic seen today is a reflection of those that established satisfactorily. The area around Whangamata today consists of protected shorelines with extensive sand dunes behind beaches exposed to the Pacific Ocean and an urban centre which serves a small resident population of 7000 that swells to 40,000 during the summer vacation period. Farmers maintain pastoral activities on the coastal plains and in the major valleys. Radiata pine plantations have been planted on the inland hills and areas have been set aside as native bush such as the Wentworth Valley Reserve. The objective of this study was to evaluate the biodiversity across the landscape with special reference to the abundance of beetles.

Methods

Malaise traps (Townes 1972), with 70% isopropyl alcohol in the collecting jars, were set out on December 5^th 1997 at each of eight collecting sites in the Whangamata area (Figure 1). Traps were oriented with the collecting jars towards the north. The sampling protocol of trapping for just the four weeks of December is recommended by Hutcheson and Kimberley (1999) after an earlier intensive season long assessment of malaise trapping (Hutcheson, 1990) showed that collections during this period gave the best discrimination among habitats. Hutcheson emphasises Coleoptera on the basis that it is the largest of the insect orders and that it is well represented across all trophic levels.

Figure 1: Location of the eight collecting sites around Whangamata, New Zealand, where pairs of Malaise traps were set out to collect insects in December 1997. Note the disjunction in parallels 41 and 50 indicating the distance between the Wharekawa Harbour study area (Site 1) and the Whangamata study area (Sites 2-8). The town of Whangamata is located at 37° 13' 40" South, 175° 52' 50" East.

Description of Collecting Sites

Site 1: The Wharekawa Harbour sand dunes. These been planted with radiata pine and maritime pine, Pinus pinaster. The two traps were set in hollows 50 m apart. Each hollow was partially protected from the on-shore winds by maritime pine. The major ground cover was harestail, Lagurus ovata. Other species present were pohuhue Muehlenbeckia complexa, spinifex Spinifex sericeus, knotty clubrush Isolepis nodosa and the sand dune moss Thuidium furfurosum.

Site 2: The Otahu River sand dunes. Several species have been planted throughout this heavily used area in an attempt to stabilize the sand dunes. Trap 2A was set out in a well grassed area with a small pohutukawa Metrosideros excelsa to the south of it. The major grasses present were cocksfoot Dactylis glomerata, tall fescue Festuca arundinacea, knotty clubrush, harestail, Indian doab Cynodon dactylon and kikuyu grass Pennisteum clandestinum. Trap 2B was set out in an area that was dominated by spinifex with catsear Hypocheris radicata, knotty clubrush and lupin Lupinus arboreus as major species. Sand coprosma, pohuehue and shore bindweed trailed through the area around both traps.

Site 3: The grassed walkways that provide public access to the main Whangamata beach from Pohutukawa Drive. These sites were very dry in the middle of summer. Traps 3A and 3B were each anchored to a wooden plank fence on the southern side of each of two of these walkways. Kikuyu Pennisteum clandestinum grass was the dominant species at each site. Both traps had a coast banksia tree Banksia integrifolia close by. There was 30% and 20% bare ground exposed at Traps 3A and 3B respectively. There was some paspalum Paspalum dilatatum close to the fences but it did not spread over the walkway.

Site 4: Pastures locked up for hay production. Trap 4A was set out in a paddock on the south side of Wentworth Valley Road while Trap 4B was erected in the paddock on the north side of the road. The Wentworth River ran through the north paddock. The main grasses around trap 4A included sweet vernal Anthoxanthum odoratum and brown top Agrostis capillaris. Sweet vernal was also the major pasture grass around trap 4B with Yorkshire fog Holcus lanatus, lotus Lotus pedunculatus, red clover Trifolium pratense and paspalum species also present.

Site 5: A 6 year old second growth radiata pine plantation on a east facing slope immediately below the Pa Road Lookout. Pampus grass Cortaderia selloana were highly visible at the edge of the plantation and were present throughout the plantation as was tutu Coriaria arborea. Bare soil was scored at 40% and bracken Pteridium esculentum was widely distributed at low density. Silver fern Cyathea dealbata was moderately abundant. Traps 5A and 5B were located 100m apart along the same contour. .

Site 6: A ridge top mature pine plantation (24 years old) to the south of the Pa Road look out. The wider spacing of this mature stand has allowed light through to the forest floor that supports a wide range of native species. We tabulated 21 species around trap 6A and 28 around trap 6B. Among these were mapou Myrsine australis and bracken as well as karamu Coprosma robusta and mamaku Cyathea medullaris. Rangiora Brachyglottis repanda, silver fern, five finger Pseudopanax arboreus and dwarf ti Cordyline pumilio.

Site 7: Fern bush. Beyond the kanuka Kunzea ericoides and mamaku border of this fern rich site there was an emergent vegetation of rewarewa Knightia excelsa and towai Weinmannia silvicola. Silver fern was the dominant fern. Rangiora, kawakawa Macropiper excelsum, five finger and mahoe Melicytis ramiflorus were present in the Tier 4 (5m-2m) layer. We tabulated 18 plant species around trap 7A and 29 around trap 7B. In addition, several epiphytes were recorded.

Site 8: The kanuka dominated native bush at the beginning of the Wentworth Valley Reserve walking track. Trap 8a was set out in area rich in silver fern, mapou and nikau Rhopalstylis sapida. Wheki Dicksonia squarossa was also common in the understory. We recorded 26 plant species from this location as well as 4 epiphyte species. Trap 8b was set out in an area dominated by towai and mapou while supplejack Rhipogium scandens trailed around the site. Major understory species included hangehange Geniostoma rupestre, kanano Coprosma grandifolia and rangiora. We recorded 32 plant species and the epiphytes supplejack, white climbing rata Metrosideros diffusa and mangemange Lycodium articulatum.

The malaise traps were cleared on December 12^th, 19^th, 26^th 1997 and January 2^nd 1998. Each sample was sorted and the individuals of the orders Diptera, Hymenoptera and Lepidoptera were removed as well as the Coleoptera. Minor orders were returned to a residue jar and stored in 70% ethanol. The sorted material was dried on paper towels, placed in weighed filter paper lined petri dishes and placed in a drying oven for four days at 30°C to reach constant weight.

The Coleoptera were identified to species or RTU (recognized taxonomic unit). The current classification of New Zealand Coleoptera families by Klimaszewski and Watt (1997) was followed. Identifications were made by comparing specimens with the Coleoptera collection at Forest Research that had been verified by Hutcheson (1996) against the New Zealand Arthropod Collection held at Landcare Research, Mt. Albert Auckland.

Insect weights were tabulated for each trap/week collection by order and analyzed for variation among weeks of sampling and sites using the general linear model for two-way ANOVA in the computer package SIGMASTAT. Means for levels of significant factors were evaluated with Tukey's test. Box plots of data grouped by site were prepared with SIGMAPLOT.

Vegetation plots were set up around each trapping location following the recce description procedures of Allen and McLennan (1983) as modified with regard to cover scores by Leathwick (1987). A plot radius of 12.6 m was used to survey the required 500 m². Data were tabulated separately for each trap area. A full species list is available on request.

Results

Insect Orders by Weight

All 16 traps were collected each week for 4 weeks to give a total of 64 samples. No samples were obtained from trap 2B in either of the first two weeks as the trap had blown down. In addition, a collecting jar from trap 8A became detached from the malaise trap in week 4 and no sample was collected. This left us with 61 useable samples.

Overall, we sorted 72.3 g dry weight of insects that comprised 27.2% Coleoptera, 18.7% Lepidoptera, 28.7% Diptera and 25.3% Hymenoptera. The distribution of this total catch over time is shown in Table 1. Catches in the fourth week, which was very hot and without any rainfall, were about half those in each of the previous weeks.

Table 1. Weights of four dominant insect orders captured in Malaise traps set out in the Whangamata area of New Zealand in 1997/98. Weekly rainfall and mean weekly temperatures are also shown.

The greatest weight of beetles was recorded in Site 3, significantly more than in any other Site except Site 4. These two grass dominated habitats provided rich resources for the large elaterid Conoderus exsul and the Lucerne weevil Sitona discoideus, which are both introduced species whose larvae thrive in the grass turf in this area.

In order to compare the Coleoptera alpha biodiversity among sites, the numbers of species represented in each trap's weekly catch were determined. The greatest numbers of beetle species were found in the mature pine (Site 6) and in the two native bush sites (Sites 7 and 8). The sand dune sites (1 and 2) had the least number of species represented (Fig. 2).

Figure 2. Box plots showing the variation in numbers of coleopteran species across sites. Letters above box plots indicate significant differences as determined by Tukey's Test.

The dominant species present in the sand dune/grass areas, Sites 1 - 4, were the lucerne weevil and the Conoderus exsul click beetle. Small mildew beetles of the genus Melanophthalma were also consistently present in the samples. The forested sites had their own distinctive beetle fauna. The two radiata pine plantations near the Pa Road lookout were characterized by the presence of the chafer Odontria xanthosticta and the longhorn beetle Oemona hirta. The corylid Arthrolips oblongus and the bark mould beetle Salpingus bilunatus were found in all four forested sites. The anthribiid Liromus pardalis and the small colourful click beetle Panspoeus guttatus were collected from all forested sites except the fern bush (Site 7). All the forest species are rated as detritivores. The additional richness of the mature pine (Site 6) is probably a result of the large number of shade-tolerant native plant species that had become established in the understory as has also been recorded by Ogden et al. (1997). The open canopy of Site 6 is not representative of the majority of New Zealand pine plantations which typically have a closed canopy and much less understory. However, our results support the suggestion by Norton (1998) that with careful management, plantation forests can in themselves contribute to the conservation of indigenous biodiversity by providing habitat for indigenous flora and fauna.

One of the more interesting features of the beetle fauna across was the degree of endemism of the beetle fauna across the eight sites (Table 2). Higher numbers of adventive (=introduced) species were recorded in the grassed sites. These beetles probably originated from discarded soil ballast of early sailing ships and from the deliberate pastoralisation practices carried out in the area with several grasses deliberately introduced over the years. Early settlers were strongly focused on establishing many species of domesticated animals and plants through the activity of acclimatization societies without any consideration of protecting native biodiversity. How times have changed.

Table 2. Analysis of endemicity of the beetle fauna collected in Malaise traps in the Whangamata area on the Coromandel Peninsula, New Zealand, December 1997.

Details on the Lepidoptera, Hymenoptera and Diptera will be reported in detail in a separate paper.

Discussion

The trap catches from a Malaise trap are very clean and provide a great opportunity to gain data across several orders without the intervention compounding stimuli such as semiochemicals. Biomass data are meaningful in terms of overall biological activity on a particular site. The dry weight determination is a low tech measure of major biological groups in the area. The state of taxonomy of all four groups studied here showed that about half the species of the Coleoptera, Diptera and Hymenoptera were fully described. By comparison, all but 11% of the Lepidoptera species were identified. Even for those insects that were identified to species, it was not always clear exactly what their ecological role might be. We need more support for basic taxonomy including clear keys to the species level if we are to fully understand the biodiversity seen in an insect trap.

In this study we showed some of the major differences in the insect fauna in several habitats in the Whangamata area. The forested sites have the greater alpha biodiversity which is an encouraging finding as the forest guilds re-establish after the extensive extraction of natural resources that occurred in the 19^th century. The more open mature pine plantations are also encouraging the development of native plant and insect species. The grassed areas by contrast provide a rich habitat for fewer insects, a greater proportion of which are adventive species. It is a challenging undertaking to match biological patterns on the landscape with man's perceptions of their naturalness.

Acknowledgments

We thank all those who gave us access across their lands to carry out this trapping study We thank B. Hock and J. Hutcheson for their guidance and comments, Chris Ecroyd for his assistance with the vegetation plots, John Dugdale for identifications of the Lepidoptera and Rod MacFarlane for identification of the Diptera and Hymenoptera.

References:

Adams, J. 1883. On the botany of the Thames goldfields. Trans. Proc. New Zealand Institute. 16:385-393.

Allen, R.B. and M.J. McLennan. 1983. Indigenous Forest Survey Manual: Two Inventory Methods. FRI Bulletin No. 48. 73 pp.

Fairweather, J.R. and S.R. Swaffield. 1999. Public pereptions of natural and modified landscapes of the Coromandel Peninsula, New Zealand. Research Report No. 241. Agribusiness and Economics Unit, Lincoln University, Canterbury, New Zealand. 54 pp.

Hutcheson, J.A. 1990. Characterisation of terrestrial insect communities using Malaise trapped Coleoptera. Ecological Entomology, 15: 143-151.

Hutcheson, J.A. 1996. Characterisation of insect communities of tawa forest in the Onaia ecological area using malaise trapped beetles and indications of influences including 1080 operations, on these communities. M.Sc. Thesis. University of Waikato, New Zealand. 238 pp.

Hutcheson, J.A. and M.O. Kimberley. 1999. A pragmatic approach to insect community characterisation: Malaise trapped Coleoptera. N.Z. J. Ecology 23:69-79

Klimaszewski, J. and J.C. Watt. 1997. Coleoptera: family-group review and keys to identification. Fauna of New Zealand no. 37. Manaaki Whenua Press, Lincoln, New Zealand. 199 pp.

Leathwick, J.R. 1987. Waipapa Ecological Area: a study of vegetation pattern in a scenic reserve. FRI Bulletin No. 130. 85 pp.

Norton, D.A. 1998. Indigenous biodiversity conservation and plantation forestry: options for the future. N.Z. Forestry 42:34-39.

Ogden, J., J. Braggins, K. Stretton and S. Anderson 1997. Plant species richness under Pinus radiata stands in the csntral North Island volcanic plateau, New Zealand. N.Z. Journal of Ecology 21:17-29.

Townes, H. 1972. A light-weight malaise trap. Entomological News 83:239-247.

Williamson, B. 1988. Whangamata - 100 years of change. Goldfields Print Ltd. Paeroa, New Zealand 78 pp.

¹ Department of Forest Sciences, Faculty of Forestry, 2424 Main Mall, Vancouver, B.C., Canada V6T 1Z4
² Research conducted while on administrative leave at Forest Research, Rotorua, New Zealand 1997/98.
³ Study was funded by a New Zealand Government FRST project C04621.
⁴ Forest Research, Private Bag 3020, Rotorua, New Zealand.