Coastal forests of eastern Africa: status, endemism patterns and their potential causes

Eastern African coastal forests are located within the Swahili regional centre of endemism and Swahili-Maputaland regional transition zone in eastern Africa, between 1d? North and 25d? South, and 34—41d? East. Approximately 3167 km² coastal forest remains: 2 km² in Somalia, 660 km² in Kenya, 697 km² in Tanzania, 16 km² in Malawi, 3 km² in Zimbabwe and perhaps 1790 km² in Mozambique. Most forests are small (≤ 20 km²), and all but 19 are under 30 km² in area. Over 80% of coastal forest is located on government land, principally Forest Reserves; only 8.3 km² is found in National Parks (6.2 km² in Kenya (Arabuko-Sokoke), 2 km² in Tanzania (Mafia Island) and tiny patches in Zimbabwe). Coastal forests are an important and highly threatened centre of endemism for plants (c 550 endemic species), mammals (6 species), birds (9 species), reptiles (26 species), frogs (2 species), butterflies (79 species), snails (>86 species) and millipedes (>20 species). Endemic species are concentrated in the forests of the Tana River, between Malindi in Kenya to Tanga in northern Tanzania, and in southern Tanzania. Forests with highest numbers of endemics are: lower Tana River, Arabuko-Sokoke, Shimba Hills (Kenya); lowland East Usambara, Pugu Hills, Matumbi Hills, Rondo and Litipo and other plateaux near Lindi (Tanzania); the Tanzanian offshore island of Pemba; Bazaruto archipelago (Mozambique), and tiny forest remnants of southern Malawi, eastern Zimbabwe and Mozambique. Most coastal forest endemics have a narrow distributional range, often exhibiting single-site endemism or with scattered or disjunct distributional patterns. They are best interpreted as relicts and not the result of recent evolution. Relictualization probably started with the separation of the ancient Pan African rainforest into two parts during the Miocene. The coastal forests are interpreted as a ‘vanishing refuge’ with the endemic species gradually becoming more and more relict (and presumably extinct) due historically to climatic desiccation and more recently to human destruction.     

Genetic regulation of colony social organization in fire ants: an integrative overview

Expression of colony social organization in fire ants appears to be under the control of a single Mendelian factor of large effect. Variation in colony queen number in Solenopsis invicta and its relatives is associated with allelic variation at the gene Gp-9, but not with variation at other unlinked genes; workers regulate queen identity and number on the basis of Gp-9 genotypic compatibility. Nongenetic factors, such as prior social experience, queen reproductive status, and local environment, have negligible effects on queen numbers which illustrates the nearly complete penetrance of Gp-9. As predicted, queen number can be manipulated experimentally by altering worker Gp-9 genotype frequencies. The Gp-9 allele lineage associated with polygyny in South American fire ants has been retained across multiple speciation events, which may signal the action of balancing selection to maintain social polymorphism in these species. Moreover, positive selection is implicated in driving the molecular evolution of Gp-9 in association with the origin of polygyny. The identity of the product of Gp-9 as an odorant-binding protein suggests plausible scenarios for its direct involvement in the regulation of queen number via a role in chemical communication. While these and other lines of evidence show that Gp-9 represents a legitimate candidate gene of major effect, studies aimed at determining (i) the biochemical pathways in which GP-9 functions; (ii) the phenotypic effects of molecular variation at Gp-9 and other pathway genes; and (iii) the potential involvement of genes in linkage disequilibrium with Gp-9 are needed to elucidate the genetic architecture underlying social organization in fire ants. Information that reveals the links between molecular variation, individual phenotype, and colony-level behaviors, combined with behavioral models that incorporate details of the chemical communication involved in regulating queen number, will yield a novel integrated view of the evolutionary changes underlying a key social adaptation.     

Impact of an introduced ant on native rain forest invertebrates: Pheidole megacephala in monsoonal Australia

Pheidole megacephala is an exotic ant species that has severely affected native invertebrate biodiversity throughout the tropics. Its impacts have been documented extensively in relatively depauperate invertebrate communities, but not in species-rich habitats such as tropical rain forests. Here we describe the local distribution of P. megacephala and its impacts on native invertebrate assemblages in and around a rain forest patch at Howard Springs, in Australia's monsoonal tropics. P. megacephala was found to be confined to a single area of approximately 25 ha, with its distribution centered on drainage lines and the rain forest. Significant but weak correlations were found between its abundance and vegetative canopy cover (positive) and distance from the rain forest (negative). In the most heavily infested area within the rain forest, the abundance of P. megacephala was 37–110 times that of total native ant abundance found within uninfested plots, as measured by pitfall traps. The abundance and richness of native ants and other invertebrates were significantly reduced in litter samples, pitfall catches and foliage beats where P. megacephala was present, inversely relative to the abundance of P. megacephala. Only two individuals of a single native ant species were found within the most infested plot, with native ant richness being reduced to about half in the least infested plot. The most persistent functional groups of native ants in infested plots were Cryptic species, which forage primarily within soil and leaf litter, and Opportunists, which exhibit highly generalised foraging behaviour. The highest abundance of P. megacephala corresponded with a 42–85% decrease in the abundance of other native invertebrates. Insect larvae were totally absent from foliage beats collected at the most heavily infested plot. P. megacephala was found overall to be expanding its range, averaging 12 m range expansion in the dry season and contracting 7 m in the wet season. It is able to spread into surrounding savanna habitats by occupying relatively sheltered microsites, such as beneath logs and at the bases of trees. However, it is unlikely to attain high population densities in open savanna habitats because of its relative intolerance of desiccation, and the prevalence of behaviourally dominant native ant species. Howard Springs is currently the only rain forest patch in monsoonal Australia known to be infested by P. megacephala, but clearly this ant is a serious potential threat to the region's rain forest invertebrate fauna. Received: 19 August 1998 / Accepted: 12 May 1999     

Immediate and longer-term effects of fire on seed predation by ants in sclerophyllous vegetation in south-eastern Australia

Following a prescribed spring fire at a woodland site with heathy understorey in south-eastern Australia, the following features changed markedly: the composition of the ant fauna taken in pitfall traps; the abundance of seed-eating species; seedfall from woody-fruited Eucalyptus obliqua and Casuarina pusilla; and rates of seed predation by ants. Broadly adapted species of Monomorium and Rhytidoponera, uncommonly recorded before the fire, became the most abundant ants caught. Seed removal from baits dropped by almost half during the week after fire, despite an increased catch of seed-eating Rhytidoponera tasmaniensis. This coincided with a massive fire-induced release of seeds stored inside woody fruits, suggesting that predator satiation occurred. The fall of Casuarina seeds and Eucalyptus chaff returned to pre-fire (low) levels 1 week after the fire, but a substantial proportion of Eucalyptus seeds fell after several weeks. By this time rates of seed removal were increasing rapidly as the abundance of seed-eating R. tasmaniensis continued to increase. Weekly removal from baits reached 100% 7 weeks after the fire, which was much higher than any level recorded before it. Results from a nearby heathland site, burned by an earlier fire, indicated that unusually high rates of seed removal, paralleling increases in abundance of R. tasmaniensis, can persist for at least 2 years following fire. These results have potentially important implications for post-fire recruitment.     

Forest edges and fire ants alter the seed shadow of an ant-dispersed plant

Exotic species invade fragmented, edge-rich habitats readily, yet the distinct impacts of habitat edges and invaders on native biota are rarely distinguished. Both appear detrimental to ant-dispersed plants such as bloodroot, Sanguinaria canadensis. Working in northeastern Georgia (USA), an area characterized by a rich ant-dispersed flora, fragmented forests, and invasions by the red imported fire ant, Solenopsis invicta , I monitored the interactions between ants and S. canadensis seeds in uninvaded forest interiors, uninvaded forest edges, invaded forest interiors, and invaded forest edges. I observed 95% of the seed dispersal events that occurred within the 60-min observation intervals. Seed collection rates were similar among all four (habitat × invasion) groups. The presence of invasive ants had a strong effect on seed dispersal distance: S. invicta collected most seeds in invaded sites, but was a poorer disperser than four of five native ant taxa. Habitat type (interior versus edge) had no effect on seed dispersal distance, but it had a strong effect on seed dispersal direction. Dispersal towards the edge was disproportionately rare in uninvaded forest edges, and ants in those habitats moved the average dispersed seed approximately 70 cm away from that edge. Dispersal direction was also skewed away from the edge in uninvaded forest interiors and invaded forest edges, albeit non-significantly. This biased dispersal may help explain the rarity of myrmecochorous plants in younger forests and edges, and their poor ability to disperse between fragments. This is the first demonstration that forest edges and S. invicta invasion influence seed dispersal destination and distance, respectively. These forces act independently.     

Colonisation of experimentally immersed wood in south eastern Australia: responses of feeding groups to changes in riparian vegetation

We investigated macroinvertebrate abundance and functional feeding groups colonising experimentally-positioned woody substrates of different species in streams with three different riparian vegetation types. Native Eucalyptus forest formed a dense closed canopy over our streams; introduced (exotic, alien) pine plantation forest did not fully shade the streams, and grassland streams were completely open, although with woody riparian vegetation well upstream of our sites. Macroinvertebrate assemblages varied taxonomically and functionally with both wood species and riparian vegetation composition. Two specialist feeding groups responded clearly to riparian vegetation: wood gougers were most common in forested streams, and algal grazers in more open streams. Gougers colonised native Eucalyptus wood in preference to alien species. Other feeding groups responses showed complex interactions between vegetation and wood type. Our results indicate the importance of sampling appropriate substrates when assessing questions of this type – if seeking shifts in functional organisation, the substrates on which the feeding groups of interest occur must be sampled. The composition of the riparian strip may influence xylophilous communities as much as the structure (i.e. whether closed or open).     

Biogeography and endemism of ants (Hymenoptera: Formicidae) in Baja California, Mexico: afirst overview

Abstract Aims Ants (Hymenoptera: Formicidae) of the Baja California peninsula are poorly known, with information based largely on scattered museum and literature records. We provide the first comprehensive account of ant species occurring on the peninsula, we examine distribution patterns, and we assess the ‘peninsular effect’ which predicts that species richness declines from the base to the tip of a peninsula. Location Peninsula of Baja California, Mexico. Methods Data collection involved examining, identifying and recording label data from c. 2350 series of ants. These records provide a provisional, if incomplete, species list. We applied the incidence‐based estimator, Chao‐2, to our data base of specimen records to estimate the total number of ant species on the peninsula. We assessed endemism by comparing our peninsular species list to those from adjacent states. The peninsular effect was tested by comparing genus and species level richness between the two states of Baja California, and across five latitudinal blocks. Results We document 170 native ant species in thirty‐three genera, plus six non‐native species, in Baja California. It seems likely that additional species remain to be discovered: the Chao‐2 estimator of species richness, at 206.0 species, is about 20% higher than our observed species richness. About 30% of the species and 20% of the genera are restricted within Baja California to the relatively mesic California Floristic Province of north‐western Baja California. Nearly all of these species also occur in California. Forty‐seven species (27.6%) are peninsula endemics. Using our entire data set, the peninsular effect appears to be strong, with about twice as many species in the northern state of Baja California than are recorded from the southern state of Baja California Sur; the ratio of genera is 33 to 24. However, this effect becomes weak at the species level and absent at the genus level when minimizing habitat effects by omitting species restricted to the California Floristic Province. At a finer scale, across latitudinal blocks of about 1.9°, the number of species declines towards central portions of the peninsula and then increases in the Cape Region. Nine ant species display strongly disjunct distributions, and these occur in two general patterns: peninsula disjuncts and peninsula–mainland disjuncts. Main conclusions The Baja California peninsula supports a diverse and distinctive ant fauna, with the proportion of endemic species similar to that displayed by plants. Patterns of species and genus richness across the five latitudinal blocks provide poor support for the peninsular effect. Moreover, habitat diversity, especially that related to topographic relief, appears to be the most important factor affecting the gradient of ant species richness in Baja California. Additional collections are needed to develop a more complete species list and to determine the boundaries and status of many species. Nevertheless, the present data base provides a useful starting point for understanding the evolution of ant assemblages in Baja California and for comparison with peninsular patterns in other taxa.     

Response of ants and spiders to prescribed fire in oak woodlands of California

Conservation managers of oak woodlands have been reintroducing fire both as an ecological process per se and to assist in restoring native plant communities. To increase our understanding of the impacts of reintroduced fire on ground-dwelling invertebrates we examined the response of ants and spiders to a late season (autumn) prescribed fire conducted in a blue oak (Quercus douglasii) woodland ecosystem in northern California. Twelve 100 m × 100 m plots were established, six plots received a burn treatment and the remaining six plots were unburned controls. Ants and spiders were sampled using pitfall traps left open continuously and collected approximately every 32 days. Sampling was conducted over a year, consisting of four pre-burn and nine post-burn collections. Abundance was analyzed using a repeated measures ANOVA, which showed seed-harvester ants decreased significantly in the two months following the fire. Total spider abundance also showed a significant decrease in two months, although this did not occur immediately after the burn. One spider hunting guild, the ‘diurnal ambush’ group (Thomisidae) remained suppressed for up to nine months. Correspondence analysis measures of ant species abundance with environmental and vegetation variables (percent rock, bare ground, plant species richness and height of herbaceous vegetation) were higher than expected by chance, which assists in explaining some of the responses. Findings from this study revealed that the reintroduction of autumn burns has modest and short-term effect on the invertebrates sampled, suggesting that late season fires are compatible with other conservation goals for oak woodland ecosystems.     

Baseline responses of Alphitobius diaperinus (Coleoptera: Tenebrionidae) to cyfluthrin and detection of strong resistance in field populations in eastern Australia

Resistance to cyfluthrin in broiler farm populations of lesser mealworm, Alphitobius diaperinus (Panzer) (Coleoptera: Tenebrionidae), in eastern Australia was suspected to have contributed to recent control failures. In 2000-2001, beetles from 11 broiler farms were tested for resistance by comparing them to an insecticide-susceptible reference population by using topical application. Resistance was detected in almost all beetle populations (up to 22 times the susceptible at the LC50), especially in southeastern Queensland where more cyfluthrin applications had been made. Two from outside southeastern Queensland were found to be susceptible. Dose-mortality data generated from the reference population over a range of cyfluthrin concentrations showed that 0.0007% cyfluthrin at a LC99.9 level could be used as a convenient dose to discriminate between susceptible and resistant populations. Using this discriminating concentration, from 2001 to 2005, the susceptibilities of 18 field populations were determined. Of these, 11 did not exhibit complete mortality at the discriminating concentration (mortality range 2.8-97.7%), and in general, cyfluthrin resistance was directly related to the numbers of cyfluthrin applications. As in the full study, populations outside of southeastern Queensland were found to have lower levels of resistance or were susceptible. One population from an intensively farmed broiler area in southeastern Queensland exhibited low mortality despite having no known exposure to cyfluthrin. Comparisons of LC50 values of three broiler populations and a susceptible population, collected in 2000 and 2001 and recollected in 2004 and 2005 indicated that values from the three broiler populations had increased over this time for all populations. The continued use of cyfluthrin for control of A. diaperinus in eastern Australia is currently under consideration.     

Occurrence of Myrmicinosporidium durum in red imported fire ant, Solenopsis invicta, and other new host ants in eastern United States

Myrmicinosporidium durum, a parasitic fungus in several ant species, is reported from seven new hosts collected in the eastern United States, including Solenopsis invicta, Solenopsis carolinensis, Paratrechina vividula, Pheidole tysoni, Pheidole bicarinata, Pyramica membranifera, and Pogonomyrmex badius. Spores can be found in most ant body parts, are dark brown when mature, and clear to light brown while immature. Ants infected with mature spores appear darker than normal. Spores from different hosts were 47-57 microm in diameter. Prevalence in host populations varied between 2 and 67% of the ants, and 3-100% of the colonies. Infection was most common in S. carolinensis with prevalence rates between 22 and 67%. Prevalence rates for S. invicta individuals were lower than for other ants, however, prevalence rates within the infected colonies were as high as 31%. Observations of disease were recorded mostly from late summer and fall. Possible implications of this new disease in red imported fire ants are discussed.     

The influence of ants on soil and water losses from an orange orchard in eastern Spain

Herbicide applications have greatly reduced plant cover, and increased soil erosion on a new orange orchard planted on valley slopes in eastern Spain. This has increased the importance of soil fauna, such as ants, in regulating soil erosion processes. Ants increase water infiltration rates by forming soil macropores during nest construction, but new soil brought to the surface by ant activity could increase the sediments available for erosion. Simulated rainfall experiments were conducted on 20 paired plots (20 with ant activity and 20 controls) to study the impact of ants on surface water flow and sediment movement in an intensively managed orange orchard near Valencia, Spain. Simulated rainfall was applied to each plot at a rate of 55 mm/h on a 0.25 m² area for 1 h. We found a reduction of soil bulk density, an increase in soil organic matter, and an increase in macropore flow in ant‐affected soils, as compared to soil without ant activity. These ant‐induced soil changes increased water infiltration rates and runoff discharge. However, the fresh, unconsolidated soil brought to the surface during nest construction resulted in greater soil loss on two plots than their ant‐free controls. Ants can be an important factor in soil erosion processes when surface vegetation is removed by intensive herbicide use.     

Generalised regressions provide good estimates of insect and spider biomass in the monsoonal tropics of Australia

Abstract  The estimation of arthropod biomass is often important in studies of terrestrial ecosystem structure and function, including analyses of the relative importance of different arthropod taxa in the diet of insectivorous animals. In order to estimate arthropod biomass in eucalypt woodlands and rehabilitated mine-land in the monsoonal tropics of northern Australia, insect morpho-species ( n  = 693) and spider morpho-species ( n  = 100) were collected, sorted, then weighed and measured. Body length–weight regressions were determined for spiders, nine insect orders and all insects combined. There was a significant relationship between body length and weight for all taxonomic groups, with the power model being a better predictor than linear or exponential models for all groups except Diptera (which was best predicted by the linear model). Whilst Schoener (1980) found that the length–weight regression slopes of neotropical insects (all orders combined, as well as several individual orders) differed from those of their temperate North American counterparts, our comparisons between monsoon-tropical and temperate Australian arthropods suggested differences among Dipterans and spiders only. We conclude that generalised regressions provide adequate estimates of arthropod biomass across Australia, providing that the body proportions of the dominant taxa do not vary substantively.     

Using ants for rangeland monitoring: Global patterns in the responses of ant communities to grazing

Ants are a prominent invertebrate group used to assess ecological change in response to disturbance. Their application as a bioindicator group has been particularly widespread in Australia, and a recent comprehensive review of their responses to environmental disturbance identified a range of consistent and predictable patterns. Here I conduct a literature review of the responses of ants to grazing globally, and specifically test whether key patterns identified in the review of ant responses to disturbance in Australia apply globally. The patterns tested were (1) soil and vegetation type are primary determinants of ant community composition, and often have a far greater effect on ant community composition than disturbance, (2) disturbance induces species compositional change, but does not necessarily affect overall species richness or abundance, (3) a species’ response is not necessarily consistent across habitats because of variation in inherent habitat suitability, and (4) approximately one quarter to one half of species that are common enough for statistical analysis have significant responses to disturbance. All these patterns were found to hold true for grazing studies worldwide. All but three studies sampling multiple soils/vegetation types found the influence of these variables to override grazing effects. Community composition changed consistently, yet the responses of total ant abundance and species richness were highly inconsistent. All studies that analysed species-level data on multiple soils/vegetation types, showed mixed responses to grazing across habitats. On average, 33% of tested species had statistically significant differences across treatments. This is the first such formulation of global patterns for any terrestrial invertebrate group for their use in bioindication, and provides valuable support to the use of ants as indicators of ecological disturbance. The challenge now is to provide a predictive understanding of this context dependency, as well as to improve the precision of the predictive responses. The confirmation of global patterns to grazing presented here represents a first step in developing the valuable contribution that ants can provide to rangeland monitoring systems.