Conservation implications of habitat transformation and pesticides on arthropod diversity and abundance in the Elgin district, southwestern Cape Province, South Africa A.B.R. Witt Percy FitzPatrick Institute of African Ornithology University of Cape Town Rondebosch, 7701 Cape Town, South Africa. Supervisors: Prof Tim Crowe, Dr Rob Little Thesis submitted in partial fulfilment of the requirements for the degree of Master of Science in Conservation Biology, University of Cape Town. June 1994
56
Embed
Conservation implications of habitat transformation and ... · Conservation implications of habitat transformation and pesticides on arthropod diversity and abundance in the Elgin
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Conservation implications of habitat transformation
and pesticides on arthropod diversity and abundance
in the Elgin district, southwestern Cape Province,
South Africa
A.B.R. Witt
Percy FitzPatrick Institute of African Ornithology
University of Cape Town
Rondebosch, 7701
Cape Town, South Africa.
Supervisors: Prof Tim Crowe, Dr Rob Little
Thesis submitted in partial fulfilment of the requirements for the degree of Master of
Science in Conservation Biology, University of Cape Town.
June 1994
The copyright of this thesis rests with the University of Cape Town. No
quotation from it or information derived from it is to be published
without full acknowledgement of the source. The thesis is to be used
for private study or non-commercial research purposes only.
Univers
ity of
Cap
e Tow
n
I
Conservation implications of habitat transformation and pesticides on arthropod
diversity and abundance in the Elgin district, southwestern Cape Province,
South Africa
by
A.B.R. Witt
FitzPatrick Institute, University of Cape Town
Rondebosch 7700, South Africa
ABSTRACT
The impact of habitat transformation on arthropod taxa and the effect of
pesticides on non-target arthropods generally has been ignored, especially in
southern Africa. In this study the arthropod diversity of a patch of natural
vegetation (fynbos) is compared to that of two apple orchards, one under
intensive pest management ("sprayedIf) and the other exposed to fungicide
treatments only ("unsprayed"). Samples obtained from pitfall traps and a D
Vac Sampler revealed that 221 insect species or morphospecies were present in
the fynbos compared to 152 and 106 in the unsprayed and sprayed orchards,
respectively. Comparative spider (Araneae) species richness was 38
.morphospecies in fynbos, 24 in unsprayed and 17 in sprayed orchards.4
Hemipterans, hymenopterans, and orthopterans were' the most speciose insect
taYtJI in fynbos. The number of coleopteran species or morphospecies were
sifhilar for all sites, whereas the other insect orders were represented by more
taxa, in the unsprayed compared to the sprayed orchard. The introduced, ,
Argentine ant Linepithema humile was the most abundant species in both
orchards. Diplopodsand especially isopods were more abundant in the
unsprayed compared to the sprayed orchard. Arthropod species richness and
abundance was influenced by the presence of host plants, the structural diversity
of the vegetation, the availability of microhabitats and the intensity of pest
management. Although transformed habitats have lower species richness than
areas of natural vegetation, arthropod diversity in apple orchards can be
enhanced, by..increasingthestructural and plant diversity of the cover crop and
by using selective insecticides.I I
1
INTRODUCTION
The greatest threat to global biodiversity is habit alteration and destruction
brought about by the expansion of human populations and their activities
(perrings etal. 1992; Kim 1993; Samways 1993). If current trends continue it
is estimated that one million animal species will become extinct by the year
2000 (Myers 1980). The majority of extinctions would occur within the
arthropoda since they are the most diverse and abundant taxa in the animal
kingdom (Kim 1993). From an ecological perspective this may have far
reaching ramifications, since insects play an important role in the provision of
ecological services like .pollination, decomposition; seed dispersal, biological
control and as a food source for a myriad of other organisms (Majer 1987;
Wilson 1987; Samways 1993).
Unfortunately, due to taxonomic impediments (Samways 1993), it is not known
how many insect species, with the exception of butterflies, recently have
become extinct in southern Africa. Agricultural development and alien invasive
plant species in the Cape Floristic Region not only pose the largest threat to
indigenous plant species (Rebelo 1992), but also their associated insect fauna.
If one considers that 68% of the more than 8500 indigenous plant species in the
Cape Floristic Region are endemic. (Bond and Goldblatt 1984), it becomes
apparent that the region probably contains a myriad of endemic insect taxa.
-Lycaenids, endemic to the region are threatened, probably because they require.~
the presence of a specific host ant and plant (Hennig and Hennig 1989).
rever, .not all phytophagous insects are restricted to a single host, since
many harbivorous taxa and feed on both indigenous and introduced plants (Liss
letal. 1986). These generalist taxa can colonize and persist in cultivated areas
with .a host of other insects including predators, parasitoids, detritivores,
coprophagous, saprophagous and mycophagous insects provided their life
history strategies are adapted to the conditions and resources available in these
habitats (Liss et al. 1986). Arthropod diversity and abundance within
transformed habitats, especially agricultural areas, is influenced by the
architecture or structural diversity of the crop plant itself (Lawton 1983), the
variety or diversity of crops within an area (Perrin and Phillips 1978; Tonhasca
1993), the abundance, diversity and management of noncrop vegetation (Morris
and Lakhani 1979; Altiera and Schmidt 1985; Liss 1986; Sheehan 1986; GoodI
and Giller 1991), the distance from and the diversity of adjacent vegetation
(Liss et ale 1986; Whalon and Croft 1986; Szentkiralyi and Kozar 1991) and the
intensity of pest management (Mansour et al. 1981; Samways 1981; Basedow et
2
al. 1985; Shires 1985; Cole 1986; Szentkiralyi and Kozar 1991). In the
southwestern Cape the abundance and diversity of apple pests and, to a lesser
extent, their natural enemies is generally known (K.L. Pringle pers. comm.).
However, the impact of various management practices on other arthropods in
apple orchards has not been ascertained. In addition, few studies have
attempted to describe the total insect fauna of apple orchards (Kozar 1987 in
Szentkiralyi and Kozar 1991). The alms of this study are to compare arthropod,
particularly insect, diversity between Mountain fynbos, a vegetation type in the
Cape Floristic Region, with that of apple orchards, under high and low intensity
pest management.
MATERIALS AND METHODS
Study area
The study area was located on the Elgin Experimental Farm, approximately 1
km north of Grabouw (34°05'S;19°05'E), southwestern Cape Province, South
Africa. The study site consisted of two adjacent 31-year old apple orchards
(cultivar Granny Smith), similar in size, and a tract of Mountain Fynbos
approximately 0.75 hectares in extent (Fig.1). One of the apple orchards
(hereafter referred to as unsprayed) received only fungicide treatments whereas
the other orchard received an additional 12 insecticide treatments (hereafter
'referred to as sprayed) between 10 October 1993 and 3 January 1994 (Appendix~
1). Both apple orchards were identical in terms of management of the cover
f'
cr · which included mowing and the use of herbicides. The fynbos patch,
aIt ough small in extent, was relatively undisturbed and was similar to large,
pnfragmented tracts in the general district.
Sampling methods
Pitfall traps and the Dietrick Vacuum (D-Vac) Sampler (Dietrick et al. 1960)
were used to sample arthropods. Pitfall traps are commonly employed to obtain
a rapid census of the epigaeic invertebrate fauna (Majer and Greenslade 1988),
while Dvvac samples are particularly efficient at sampling hemipterans, adult
dipterans and adult hymenopterans (Johnson et al. 1957). To reduce any
possible edge effects.vsamples.iathe .apple orchards were only taken from the
area surrounding 72 trees (72 x 36 m) in the centre of each orchard. An area
was selected for sampling, similar in size to that in the orchards, in the centre of
the fynbos patch (Fig. 1).
3
Eight of the 72 trees in each orchard were selected at random. Four pitfall traps
were placed at 75, 150,225 and 360 ern from the base of each selected tree, at
90° to each other, using the orientation of the tree row as the main reference
line (0°) (pig. 1). However, since insect diversity is influenced by a
combination of factors including shading, aspect and the type of ground cover
the traps were rotated as shown in Table 1. Eight points were selected at
random in the fynbos patch. Trap placement was similar to that used in
orchards. Each pitfall trap consisted of a test tube (25 x 150 mm) within a
plastic pipe which was sunk into the ground so that the lip of the tube was flush
with the soil surface. Each tube was filled with 40 ml of water and
approximately three drops of detergent to break the water tension. Alcohol was
not used since it may act asa repellant to certain arthropod taxa (Southwood
1966). Pitfall traps were set up at least one week before sampling commenced
and sealed with a rubber stopper when not in use. Each trapping period
extended over 10 days, after which all the invertebrates were collected and
placed in vials containing alcohol. Pitfall traps remained sealed for
approximately 20 days before they were reopened for another 10-day period.
The traps were surveyed three times over a 50-day period on 24 November, 15
December 1993 and 4 January 1994.
Fifteen Dsvac samples were taken in each of the orchards and in the fynbos
"patch.' Five samples were taken from randomly selected 1m2 quadrats, directly4
under the trees, between the trees within a row and in the work area between
thfows., The ground cover under the trees consisted mainly of leaf litter, the
area between the trees within a row was mainly bare with some leaf litter, while
Ithe work row had a weed and grass cover. Sampling was done between 10hOO
and lShOO' on warm and calm days when insects were most active. Since insect
activity is main~y influenced by temperature and other abiotic factors, five
consecutive samples were taken in a particular site before moving to another
site. The sites and'.quadrats were sampled in random order. All samples
collected, together with the debris, were placed in separate plastic bags. Filter
paper, dipped in ethyl acetate was placed in each bag to kill all arthropods.
Samples were collected on 14 November, 5 December and 24 December 1993.
Identification (All arthropods, excluding collembolans, were sorted and identified to class or
order level. Aranids and insects were identified to species or morphospecies
but, insects were the only taxa identified to family level. Morphospecies were
4
generally based on external morphology only, although genitalia were used in
determining morphospecies in taxa like ground beetles where species are known
to be very similar morphologically. Dimorphic taxa like dermapterans and
polymorphic . taxa like aphids and formicides were split into separate
morphospecies as were taxa which exhibited holometabolous development.
Many insect nymphs could not be associated with any adult taxa collected
during the study and were therefore classified as morphospecies. Although the
use of morphospecies has been criticised (Kim 1993), the use of recognisable
taxonomic units (RTU's) in rapid biodiversity assessments has its merits
especially during preliminary surveys (Oliver and Beattie 1993). In addition,
particular life stages within a single species may be more sensitive to habitat
transformation and pesticides than other instars. If one also considers the
taxonomic impediment (Samways 1993) and the need to rapidly estimate
biodiversity and the effect of habitat destruction (Oliver and Beattie 1993) the
use of morphospecies is warranted. In the remainder of the text morphospecies
will be referred to as species unless stated otherwise. The nomenclature for
insects follows that of Scholtz and Holm (1985), and voucher specimens are
lodged in the Department of Entomology, University of Stellenbosch, South
Africa.
Dataanalysis'The .data obtained from pitfall traps and D-vac samples were combined for all
4
analyses. Correspondence analysis (Greenacre 1986) was used to compare thethi sites-in terms of ~e number of species in each of ~e insect orders
sampled. Psocopterans, trichopterans, neuropterans and phasmids were grouped
together since they were restricted to only one of the three sites.
Correspondence analysis is useful since it graphically illustrates which taxa are
contributing to the separation of sites.
Diversity profiles were used to compare insect diversity in the three study sites.
The method used by De Kock et ale (1992) and devised by Patil and Taille
(1976, 1979) has many desirable properties in that three diversity indices which
are popular amongst ecologists (Dennis et ale 1979) can be graphically
5
represented on the same pair of axes. Dennis et at. (1979) use the following
expression to generate a series of diversity indices:
N86 = L Pi(1 - P,.6)/13, i = 1,2,•... N,
i=l
where N = number of species,
Pi = proportion of individuals in species i, and
13 = measure of the weight attached to eveness.
If 13 is in the range -1 to + 1, then for
13 = -1, .6-1 = N - 1, or the species count, and for
13 approaching 0,.60 = -LP;lnPi , or the Shannon-Weaver Diversity
Index and for
13 = +1,.6+1 = 1 - LP,2, or the Simpson's Species Evenness Index.
The index was modified to standardise it to a scale of between 0 and 1 as