H1111111111111111

0 20 40 60 80 100 120 140 160 Rank order of species' abundance

4.8 Bornean butterfly species ranked according to their abundance in primary (open circles) and logged (closed circles) forest. (Redrawn from Willott et al., 2000, with kind permission of Blackwell Publishing.)

It is essential that stenotopic insect species with small geographical ranges are considered in any large-scale assessment, as these species may be confined to particular closed-canopy habitats tolerant of only very minimal disturbance (Spitzer et al., 1997). Conversely, some species are lost locally when there is no disturbance (Brown, 1997). A further point is that different taxa respond differently to disturbance. In Bolivian tropical forests, ants and cockroaches were more abundant in undisturbed areas, whereas grasshoppers and lepidopteran larvae were more abundant in burned and partially logged areas (Fredericksen and Fredericksen, 2002), this being a reflection of life styles and preference for particular amounts of insolation. While much focus has been on moist tropical forests, some dryer forests can also be rich in insect and other arthropod species, with Australian eucalypt forests being much more species-rich than previously thought (Majer et al., 2000).

Saproxylic insects are a rich and varied dominant functional group which depend on dead wood and the old trees that generate it (Grove, 2002). These insects are sensitive to forest management, with managed or secondary forests generally having a depauperate fauna. Forest hygiene, where old and dead wood is removed, coupled with saproxylic insects' often weak powers of dispersal and long-lived larval stages, has severely threatened this group of insects, with many species in Western Europe now regionally extinct.

4.7 Transformation of grasslands, savanna and Mediterranean-type ecosystems

4.7.1 Grasslands

Grasslands in the northern hemisphere alone once covered some 600 million ha, with very few natural remnants left today. Interestingly, across the northern hemisphere there is remarkable similarity in the grasshopper fauna (Lockwood and Sergeev, 2000). Such grasslands are not homogeneous, with the shrub and desert zones deserving highest conservation priority as they have the highest number of rare species. In some respects, this situation is magnified in the southern hemisphere. Having had no extensive Pleistocene ice sheets, yet long periods of relative geological stability coupled with an erosional, topographically complex range of landscapes, the south has an even richer endemic insect fauna (Samways, 1995), with grasshoppers, for example, 47% endemic in South Africa and 90% endemic in Australia. Human-induced fire (Greenslade, 1993; Burchard, 1998), direct habitat destruction such as removal of thick bush (Scholtz and Chown, 1993), afforestation, agricultural development, as well as hunting, pastoralism and livestock rearing have all had major impacts on insects of the southern non-forest ecosystems (Samways, 1995).

Overgrazing has also been extensive in other areas (Samways and Sergeev, 1997), resulting in depression of populations of some species but also providing opportunities for outbreaks of others (Samways and Lockwood, 1998; Lockwood et al., 2000). In North America, tallgrass prairie once covered 68 million ha but has now been reduced to scattered remnants. Besides the effects of loss of area per se, the prairie fragments are subject to invasion of agricultural pests from the surrounding fields. Corn-rootworm beetles Diabrotica spp. damage the flower-heads of adjacent prairie composites, reducing seed set (McKone et al., 2001).

Human impacts on these open, non-forest areas have probably been intense for long periods of time (Flannery, 1994) and what we are seeing today in many areas, from the Russian steppes (Bei-Bienko, 1970) to the Australian wilderness (Greenslade, 1993), is a post-disturbance fauna, and one that has been very under-researched (Scholtz and Chown, 1993). One of the problems with determining the causal factor of human impact on insect diversity is that the various forms of disturbance are synergistic. Fire, overgrazing, impact of invasive aliens all interact, and can aggravate natural drought/flood cycles. Burning of natural grasslands, for example, amplifies the effect of natural cold-air drainage on grasshopper abundance and diversity (Samways, 1990).

Increasing levels of domestic livestock impoverish insect diversity, but at least in Africa, this simulates the effect of indigenous hoofed mammal trampling and grazing (Rivers-Moore and Samways, 1996), which becomes amplified at water-holes that are congregation points (Samways and Kreuzinger, 2001). Even tourist impact can be additive upon natural impacts, with the Greek moth Panaxia quadripunctaria suffering human trampling (Petanidou et al., 1991).

Transformation of grasslands, savanna and Mediterranean-type ecosystems 79

4.9 Density (a) and biomass (b) of insects in each trophic guild, collected at four sites under different grazing regimes and age of exclusion from cattle, at a montane grassland in Central Argentina. (HG) Heavy grazing; (LG) low grazing; (YE) young enclosure (7 years old); (OE) old exclosure (19 years old). Different letters above bars denote significant differences in biomass (total and for (*) marked guilds) among sites. (From Cagnolo et al., 2002, with kind permission of Kluwer Academic Publishers.)

4.9 Density (a) and biomass (b) of insects in each trophic guild, collected at four sites under different grazing regimes and age of exclusion from cattle, at a montane grassland in Central Argentina. (HG) Heavy grazing; (LG) low grazing; (YE) young enclosure (7 years old); (OE) old exclosure (19 years old). Different letters above bars denote significant differences in biomass (total and for (*) marked guilds) among sites. (From Cagnolo et al., 2002, with kind permission of Kluwer Academic Publishers.)

Grassland deterioration and consequent loss of insect populations and extinction of species have illustrated the vulnerability of the grassland fauna, many species of which have a substantial impact on soil fertility and plant growth (Curry, 1994). Grasshoppers in the arid Karoo, for example, convert plants to nutrient-rich frass, and do so much faster than sheep (Milton and Dean, 1996). Also in the Karoo, monkey beetles (Scarabaeidae, Hopliini), which are pollinators, are influenced by level of grazing, with a shift away from perennial and bulb pollinator guilds towards those favouring weedy annuals in overgrazed areas (Colville et al., 2002).

Cagnolo et al. (2002) showed that in the grasslands of montane Argentina, abundance, richness, diversity and biomass of insect assemblages were minimal in the most intensively cattle-grazed area (Figure 4.9). In addition, besides changes in taxonomic composition, intensively grazed areas had fewer secondary consumers, with chewers replacing suckers as the most abundant herbivore group. Nevertheless, there are also many functionally insignificant species in grassland and some of these are threatened and restricted to 'island' reserves. Especially threatened are some Orthoptera species (Rentz, 1993; Samways, 1997a) and Lepidoptera species (New, 1993). In reality, it may not necessarily be that grasshoppers and butterflies of open habitats are any more threatened than other taxa in the same or different habitats. It may be that we are simply seeing declines and losses in these conspicuous insects.

It is among the huge species richness of the tree canopy (Watt et al., 1997) where the greatest population and species losses are actually taking place (Mawdsley and Stork, 1995). The important point, however, is not which ecosystem is worse off, but are the threats similar, and are there principles that can ameliorate the threats whatever the ecosystem? This point is addressed in Part III.

4.7.2 Mediterranean-type ecosystems

Another low-canopy ecosystem that deserves special mention is the fyn-bos (Cape Floristic Region), where many narrow endemic plant-insect species interactions occur (Wright, 1993). This, and other Mediterranean-type ecosystems (MTEs), are under severe pressure, probably proportionately more than in any other system (Hannah et al., 1995). This inevitably means that many insect species in these systems are threatened (Samways, 1998b). The salient point here is that this anthropogenic pressure must be viewed against the fact that many insects in MTEs are narrow endemics (e.g. Wright and Samways, 1998). As such, they are 'pre-adapted' to surviving in a small area, and their survival becomes an all-or-none affair. If the human impact is intense and squarely overlays their focal population, their demise is highly likely. If, however, the impacts leave them in an undisturbed fragment, their chances of survival are high, so long as there is not too much of an adverse context effect from the surrounding disturbance matrix. It is almost as if we have 'unlucky' versus 'lucky' endemics. The giant flightless cockroach Aptera fusca and the Conspicuous malachite damselfly Chlorolestes conspicuus have a 'lucky' home on top of Table Mountain despite being surrounded by the city of Cape Town (Figure 4.10).

In the MTE of California, the response of arthropods to habitat fragmentation is complex, and depends very much on the taxon in question (Bolger et al., 2000). Nevertheless, fragment area and edge effects were generally so significant, along with the impact of the alien Argentine ant Linepithema humile, that in all likelihood, trophic relationships within the community are changing.

4.8 Deterioration and loss of aquatic systems

4.8.1 Canalization and synergistic impacts

Besides canalization and impoundments (Section 4.5), there are other changes to aquatic systems that threaten insect diversity. A first consideration

4.10 The giant flightless cockroach Apterafusca (a) and the Conspicuous malachite damselfly Chlorolestes conspicuus (b) are narrow-range, Western Cape endemics that find refuge on top of Table Mountain, yet they are surrounded by the city of Cape Town (c) (see next page).

is that stream invertebrate communities are structured to some extent by the type and intensity of disturbance (Power et al., 1988; Resh et al., 1988). But how resilient are stream faunas? The answer depends on the intensity and frequency of the disturbance, and relative sensitivity of the responding taxa and ecological relationships. In contrast, when disturbances are spatially confined and relatively short-lived, recovery of the invertebrate aquatic community can be very rapid, with more complex communities at the small scale (< 1 m2) being the most resilient (Death, 1996). Even at the landscape level, certain aquatic insect populations can recover remarkably quickly, with populations returning to former levels within a year after severe floods (Samways, 1989a).

The threats to aquatic systems are often synergistic with loss of habitat (including damming, canalization, water diversion and draining), isolation of source habitats, pollution and threats from alien organisms all contributing, as seen by the threats to British aquatic insects (Shirt, 1987). One of the reasons threats to aquatic insects can be so severe is that water bodies are relatively small, with lakes and marshes generally only small patches in the extensive terrestrial matrix (Angelibert and Giani, 2003). Arrival of new propagules becomes increasingly small as water bodies are lost and populations in source areas decline. Indeed, loss of wetlands worldwide is of considerable concern. In Finland, for example, the decrease in butterfly species has been greatest in those species living in bogs and fens (Saarinen et al., 2003).

An additional factor is pollution which can also be synergistic with other impacts. A pollutant can easily disperse through a relatively small body of water, as well as be carried downstream. Recolonization can be restricted, as the lateral dispersal of adult insects of headwater streams appears to be very limited (Griffith et al., 1998).

4.8.2 Dragonflies as an example

Work on dragonfly conservation in recent years has enabled a focus on what really are the threats rather than supposed threats. For dragonflies at least, perhaps because they are predators, it does not matter very much whether the riverine canopy is composed of alien or indigenous vegetation (Samways, 2003a,b) as long as the right proportions of sunlight versus shade are present (Steytler and Samways, 1995). This does not however, mean that alien plants are not harmful. The problem comes when the trees are invasive and convert grassland banks to alien forested banks. Such dense-canopy aliens include Acacia mearnsii and A. longifolia which are threatening at least 12 highly localized sun-loving odonate species (Samways and Taylor, 2004). Almost perversely, the interpretation behind this is seen when plantation trees, such as pines, are introduced along a grassland stream. What actually happens is that conditions now mimic forest stream conditions so disfavouring the grassland species while favouring forest species (Kinvig and Samways, 2000).

The dragonfly fauna also depends on the type of stream and whether it is constant in flow and level, or not. Large, savanna streams subject to the vagaries of El Nino, for example, generally have an opportunistic habitat-tolerant fauna, components of which, nevertheless, can be washed away when conditions are particularly harsh (Samways, 2003c). In more climatically stable areas, where streams have a more constant, perennial flow, endemic species packing can be high, with threats such as overextraction of water, alien plant invasion and alien fish predation all impacting on the dragonflies (Samways, 1995). However, close-focus on actual cause is essential. Englund and Polhemus (2001) point out that introduced rainbow trout Oncorhynchus mykiss have posed little threat to the endemic Hawaiian damselflies. This contrasts with alien poeciliid fish, which are clearly destructive (Englund, 1999).

4.8.3 Aquatic ecotones and marshland

Aquatic habitats, which also include wetlands, worldwide are of major biodiversity significance. The important point also being that many of these waterbodies are unique, from the tiny and vulnerable clear streams in the caves of Table Mountain, Cape Town, with an unusual endemic fauna (Sharratt et al., 2000), to the enormous Okavango system of Botswana. Yet it is not simply a wetland on the one hand or a dryland on the other, rather a series of aquatic ecotones (Samways and Stewart, 1997). Pullin (1999) focuses on these dynamic, marginal areas, and points out that these habitats have long been threatened by drainage for agricultural use and by loss of water to surrounding land. On top of this, is increasingly severe drying out caused by anthropogenic climate change. But the problems extend beyond simply drying up, to those associated with rapid changes in water levels. Excessively long submergence increases mortality of sawflies (Lejeune et al., 1955) and butterflies (Webb and Pullin, 1998; Joy and Pullin 1999). The extinction of the British Large copper butterfly Lycaena dispar dispar and decline of the Heath fritillary Mellicta athalia are likely to herald changes in insect diversity in wetlands throughout the world, and that localized, habitat specialists need to be watched very carefully as early responders of possible permanent change.

4.9 Pressure on special systems

4.9.1 Oceanic islands

Many oceanic islands have been subject to intense human disturbance (Quammen, 1996). Besides direct loss of indigenous habitat, especially on lowlands, there are many introduced species on islands, from ants to Albizia trees. Such introduced species may be additive upon the island system or they may eliminate or suppress indigenous species through competition, predation or disease. Global climate change is apparently already aggravating the situation (Chown et al., 2002). The overall prognosis for some of these islands thus remains grim, with the current challenge being simply to maintain existing biodiversity (Clout and Lowe, 2000).

Insect diversity on islands is often skewed, with some taxa not being represented. This is the result of the sweepstake effect, where only certain taxa successfully land and colonize an island. Many of those that naturally invaded new islands or were marooned on islands that became separated from the mainland developed into island endemics. This is because the island is separated from neighbouring suitable habitat by a hostile environment, the sea. Counterintuitive as it may seem, many island endemics are not more unusual than their mainland counterparts. Besides adaptive radiation, where species have evolved sympatrically into different niches and where they have acquired evolutionary stability, there is also fugitive radiation. This, according to Adsersen (1995), is the appearance of 'weak' species, which are very local and have to evolve further to avoid extinction. Many insects appear to fall into this latter category and maintain remarkably small populations (Samways, 2003a,b), which presumably are highly susceptible to adverse changes that have not previously been encountered. This emphasizes the risks of synergistic effects of global warming and invasive aliens, which are impacting severely on some island faunas.

4.11 Simplified food-web of lowland Nothofagus forest in the northern South Island of New Zealand, illustrating the impacts of invasive animals (circled). Direction of energy flow is shown by arrows, with solid lines connecting indigenous elements and dotted lines showing predation on indigenous biota by the invasive animals. (From Clout, 1999, with kind permission of Kluwer Academic Publishers.)

4.11 Simplified food-web of lowland Nothofagus forest in the northern South Island of New Zealand, illustrating the impacts of invasive animals (circled). Direction of energy flow is shown by arrows, with solid lines connecting indigenous elements and dotted lines showing predation on indigenous biota by the invasive animals. (From Clout, 1999, with kind permission of Kluwer Academic Publishers.)

These impacts are virtually meteoric, with many of the invertebrate introductions on the relatively unvisited Gough island having occurred in the last 50 years (Jones et al., 2002). On islands that are visited frequently, the invasion frequency of insects is extremely high. Hawaii accumulates 20-30 new insect species per year (Beardsley, 1991) and Guam accumulates 12-15 new species (Schreiner and Nafus, 1986). Evidence is now accumulating that invasive insects, along with other agents of change are affecting certain island food webs, such as Nothofagus woodland in New Zealand (Clout, 1999) (Figure 4.11).

4.9.2 Caves

Special environments inevitably are islands surrounded by less favorable habitat, and need not necessarily be oceanic islands. Many other habitat islands exist, which are pockets of special and/or threatened insect diversity (Stanley and Weinstein, 1996). Among these are caves. Their obligate cavernicolous inhabitants tend to show a high degree of very localized endemism (Barr and Holsinger, 1985). The ancestors of these species may have taken refuge in the humid cave habitats during periods of inclement surface conditions, so as well as being rare and having well-developed troglomorphy, they also have specialized habitat requirements (Howarth, 1987). Inevitably, in the face of anthropogenic pressure, some of these faunal components are threatened (Howarth, 1981; Culver et al., 2000), mostly because a relatively small disturbance, at least in terms of surface standards, can have major repercussions for stenotopic troglobites and stygo-bites (Balleto and Casale, 1991; Sharratt et al., 2000). Some of this disturbance may be allochthonous, with, for example, the cave cricket Speleiacris tabulae and its co-inhabitants being dependent on bat guano, and hence on survival of bats that forage in the surrounding urban environment. This emphasizes that cave arthropod conservation depends largely on an integrated, whole-ecosystem approach (Harrison, 1964; Culver et al., 2000).

There are many other 'special environments' or special localities, and it is one of the aims of prioritization to discover and assess threats to geographically unique or habitat-unique localities. These may then become Sites of Special Scientific Interest or equivalent, and generally require a total protectionist approach. Although the 'island' species have survived genetic bottlenecks, and may not necessarily exhibit metapopulation dynamics, they are likely to be very vulnerable to instantaneous anthropogenic impacts, many of which are synergis-tic. What we do not know yet is how cave insects and other faunal components will survive the new surface changes, especially global warming. Will their insulated home be enough to pull them through?

4.10 Overcollecting

Certain insects have human appeal, as food items, aesthetic collectables or as scientific curiosities. In short, certain insects have utilization value (Morris et al., 1991). But when does consumption outstrip supply? This usually happens as a ramification of fragmentation, or more precisely, in terms of Forman's (1995) model, dissection of the landscape. As paths and roads penetrate natural habitat, this then encourages increasing human traffic and hence exploitation of resources, including insects. This has led, for example, to removal by tourists of 100 000 Panaxia quadripunctaria moths per generation in the Valley of Butterflies, Rhodes, Greece (Petanidou et al., 1991).

Like so many aspects of conservation, overcollecting must be put in perspective and on a rational, non-emotive level. For butterflies at least, which include the most collected of all insects, New (1997) points out that the adverse effects of collecting are probably far less than that of habitat change and that simple bans on collecting play only a minor role, if any, in conservation. However, it is essential that collecting be monitored carefully because certain species with small total populations, and which may also be slow breeders, may be susceptible to overcollecting. Nevertheless, we must be sensitive to the fact that for certain species overcollecting has caused extinction. The British Large copper butterfly Lycaena dispar dispar appears to have been collected out of existence by 1848 (Duffey, 1968). For the 33 species of butterfly listed under the United States Endangered Species Act, 30% are threatened from overcollecting. This human harvesting is a specialized form of predation and as such can result in localized overexploitation and extirpation. This can be extended to harvesting of wild insects. With regards to indigenous African silk moths, it is essential to establish levels of sustainable utilization (Veldtman et al., 2002).

4.11 Summary

Human population growth over the last century has resulted in an increase in consumption of resources by 460%. There is little concrete evidence that industrial pollution (as opposed to the physical footprint of urbanization) has had any major effect on terrestrial insect diversity. Impact on some aquatic systems, however, has been severe. Identifying pollution as a key factor eroding insect diversity is a complex issue and so we may be underestimating its effect. Impact of pollution is a matter of toxicant concentration, with only high levels having a major effect in some cases. Most concern is for the long-term, especially long-lived contaminants, especially in the soil.

Despite the huge consumption of pesticides worldwide, there is no evidence that pesticides have been singularly responsible for any insect extinction. This is because non-persistent pesticides are generally applied only over a limited area over a relatively short time period. Persistent pesticides are more insidious, upsetting predator--prey relationships. These effects are often synergistic with other impacts, including the increased use of fertilizers and herbicides.

Agricultural fragmentation of the landscape has many ramifications, with remnants of natural vegetation in agricultural areas often home to considerable insect diversity. This is especially so for the architecturally more complex agricultural landscapes. Similarly, diverse and complex green nodes in cities can be remarkably rich in insects. The converse situation is perforation of the natural landscape with agricultural or urban patches, as well as road corridors. These impacts have effects that go beyond the patch edge. Some insects benefit from these modified patches, while others do not. The upshot is that the landscape is best viewed as a differential filter, letting some insects through (physically and genetically) but not others.

As the tree canopy is so important for insect diversity, the current rates of deforestation, principally in the tropics, are devastating insect diversity. The remnant reserves are often too small and too vulnerable to disturbance to guarantee long-term survival of many species. The point is that large remnants of intact, virgin areas of forest are often critical for maintaining large-sized, ecologically sensitive, narrow-range endemics. The precautionary principle, of 'keeping all the parts', especially all the important parts of old forests, is pivotal for maintaining current levels of insect diversity. Evidence is pointing to a future dominated by weedy, ecologically tolerant species with tramp invaders among them.

Grasslands and Mediterranean-type ecosystems have also suffered insect loss, principally from the synergistic effects of fire, overgrazing and impact of invasive aliens, which can aggravate natural drought/flood cycles. Of concern is that the decreased arthropod diversity in these systems is reducing soil fertility.

Aquatic communities are remarkably tolerant to natural flood/drought cycles, having been honed over millennia. Human impacts, however, present a new mul-tifaceted force on aquatic insects. Canalization streamlines water flow, while invasive alien plants change the character of banks and water alike. In turn, cattle trample natural riparian vegetation and agricultural run-off contaminates the water. Wetlands, which are the soaks and cleaning agents of the hydrolog-ical landscape, are under siege. Wetlands are 'special environments', like caves and islands, whose insect diversity is under enormous pressure. Added to this, are risks of overcollecting, which is a form of specialized predation by humans of mostly showy species. The bottom line is that these various impacts operate together, and we may not be detecting which is most harmful or able to determine exactly what the long-term impacts will be.

Was this article helpful?

0 0

Post a comment