In this study we describe the ant assemblage visiting both foliar and bracteal extrafloral nectaries (EFNs) of Passiflora coccinea, investigate the protective role provided by ants against both herbivores and nectar-thief butterflies, and analyze possible fitness benefits for the plant (i.e., fruit and seed production) resulting from the presence of ants. Experiments and observations were performed in two Amazonian terra firme forest sites. Twenty-two ant species were recorded visiting the EFNs of P. coccinea. Camponotus aff. blandus was the most frequent and subordinate among the EFN-visiting ant species, accounting for 20% of the ant attacks on simulated herbivores (termites) and 50% on butterflies. Both the number of ant individuals and the number of species were positively correlated with the combined number of buds and flowers of P. coccinea. Moreover, ant protection behaviour against termites was stronger on flowers than on leaves. Finally, seed set almost doubled among fruits originating from ant-visited flowers as compared to flowers from which ants were artificially excluded. Our results suggest a protective role of ants for flowers of P. coccinea against both herbivores and nectar thieves, improving plant reproductive success.

Ecologists routinely use Bonferroni-based methods to control the alpha inflation associated with multiple hypothesis testing, despite the aggravating loss of power incurred. Some critics call for abandonment of this approach of controlling the familywise error rate (FWER), contending that too many unwary researchers have adopted it in the name of scientific rigour even though it often does more harm than good. We do not recommend rejecting multiplicity correction altogether. Instead, we recommend using an alternative approach. In particular, we advocate the Benjamini–Hochberg and related methods for controlling the false discovery rate (FDR). Unlike the FWER approach, which safeguards against falsely rejecting even a single null hypothesis, the FDR approach controls the rate at which null hypotheses are falsely rejected (i.e., false discoveries are made). The FDR approach represents a compromise between outright refusal to control for multiplicity, which maximizes alpha inflation, and strict adherence to FWER control, which minimizes power. We review the multiplicity problem, illustrate the advantage of the FDR approach, and promote this approach for widespread adoption in ecology.

Factors influencing habitat selection may be scale dependent, leading to different selection patterns at different spatial scales. By limiting habitat-selection studies to a single scale, important selection patterns could be missed. Despite this danger, many studies investigate habitat selection at a single scale, often ignoring macro-habitat selection: the selection of a home range within the study area. We investigated macro- and micro-habitat selection in milksnakes. Because of the importance of thermoregulation to ectotherms, we predicted that snakes would select habitats of high thermal quality at both micro- and macro-habitat scales. In 2003–2004, we located 25 individuals 890 times and characterized the habitat in detail at 279 locations used by milksnakes and at 279 paired random locations. Open habitats (fields, rocky outcrops, marshes) and edges have environmental temperatures that deviate less from the preferred body temperature range of milksnakes and offer characteristics that facilitate thermoregulation compared to forest. At the macro- and micro-habitat scales, milksnakes preferred habitats of high thermal quality: they used fields and rocky outcrops more than forests. Milksnakes also preferred edges at both scales. In addition, milksnakes preferred locations with open canopy and many rocks at the micro-habitat scale. These results support the notion that thermal quality influences habitat use in ectotherms and strengthen the idea that habitat-use studies should be conducted at more than one spatial scale to gain a complete understanding of the factors affecting selection.

We analyzed the long-term canopy dynamics of a 2-ha permanent plot in subalpine old-growth coniferous forest over 43 y using digital surface models (DSMs) and a digital elevation model (DEM). The models contoured canopy and ground surface elevation, respectively. Abies mariesii, A. veitchii, Betula ermanii, Picea jezoensis var. hondoensis, and Tsuga diversifolia were the main species in the forest canopy. Canopy surface DSMs (2.5- × 2.5-m grids) were constructed of the area including the 2-ha plot using aerial photographs from 1959, 1969, 1979, 1989, and 2002, and a DEM was constructed from ground survey data collected in 1991. Canopy height profiles were obtained by calculating the difference between the canopy and the ground surface, and the status of the forest in each grid cell for each year was classified as gap or closed canopy, depending on whether the canopy height was ≤ 25 m or > 25 m, respectively. Tree census data were collected in 2000. The threshold value was decided by comparing the gap from digital elevation data with the result of the field survey. The total gap area in 1959 was greater than 1 ha, indicating that some disturbances had occurred in this plot, probably related to the Isewan Typhoon. A large change occurred during 1969–1989, when mean canopy closure rates were significantly higher than mean gap formation rates. Abies mariesii and B. ermanii tended to occur in the canopy layer in grid cells that contained gaps in 1959 and closed canopy in 2002. The presence of A. veitchii in the canopy layer was also associated with the change from gap to closed canopy, although not significantly so. These results suggest that Abies spp. regenerate more effectively than the other species by establishing seedling or sapling banks before gap formation. Large-scale disturbances, such as the Isewan Typhoon, do not favour the regeneration of spruce over subalpine fir, and species other than spruce are responsible for recovery following such disturbances, according to our analyses of long-term canopy dynamics.

A common management aim is to keep large herbivore densities at sustainable levels. However, measuring grazing pressure is difficult. We seek combinations of herb species and morphological traits that serve as indicators of grazing pressure in alpine areas of Scandinavia. In a fully replicated, landscape-scale experiment with three replicates of each of three levels of sheep density (zero, low, and high), we measured grazing frequency (numbers of plants eaten, as a proportion), flowering frequency, and plant height for 17 common herb species for two consecutive grazing seasons. For eight species, grazing frequency varied significantly between treatments, and for four of these species there was additional strong between-year variation. The estimate for total grazing frequency of herbs decreased when more species of decreasing preference were included, and it was significantly lower in the second grazing season. There were large between-year differences in plant height probably related to weather patterns. A grazing treatment effect on plant height was found only for two species. Flowering frequency in selected species changed even at low sheep densities and was thus not considered suitable except for indicating ungrazed areas. Grazing frequency of three herbs, Saussurea virgaurea, Rumex acetosa, and Pyrola minor, provided a useful indicator of total grazing pressure, as these species were grazed only at high sheep density. Our study area is likely to be representative of alpine areas in southern Norway, but we advise careful calibration using monitoring procedures before similar management decisions on stocking rates can be fully implemented in other ecosystems.

In boreal forests, wildfire is a dominant ecological process that, among other things, affects the distribution and abundance of terrestrial lichens. Woodland caribou (Rangifer tarandus caribou) occupying peatland complexes of western Canada rely on terrestrial lichens for winter forage. Understanding the relationship between lichen distribution and fire is therefore important in order to understand caribou ecology. We documented abundance of terrestrial lichens in 73 peatland sites in northern Alberta, all within caribou home ranges. Forty-eight of these had been disturbed by forest fire within the past 70 y, while 25 had not been disturbed for at least 70 y. Peatlands that had not been disturbed for at least 70 y had low lichen cover (average 21.7 ±2.1%; range 4.6–54.0%) and lower lichen biomass (660.0 ±63.3 kg ·ha⁻¹) than other studied boreal areas. However, lichen in sites disturbed by fire appeared to have recovered after only 40 y. This rapid recovery seems to have been mediated by a high growth rate: 4.8 ± 0.1 mm·y⁻¹. Controlling for the effect of time since fire, lichen cover was inversely related to cover of Sphagnumspp., while growth rates of lichen were positively related to time since fire. Although the re-growth of lichen after fire was rapid in comparison to other systems, we suggest that fire has a strong effect on lichen distribution and hence on the spatial distribution of foraging habitat for Alberta caribou.

Managers often face the dilemma of planning reserve networks with limited data on species' distributions; “umbrella species,” as surrogates for other co-occurring taxa, were thus proposed. Here, the relative efficiencies of “target species” representation in reserves selected using “single-species umbrellas” and “umbrella species groups” are compared, both relative to each other and to target species representation in randomly selected reserve areas. Distribution data for vertebrates and plants on islands of six Great Lakes basin archipelagos were analyzed. Reserves selected using “umbrella groups” contained more species than did those selected using “single-species umbrellas.” Random selection constrained to the same total area occupied by umbrellas typically performed as well as umbrellas of any type. Reserve systems selected at random but constrained to the same number of islands occupied by umbrellas, however, contained lower proportions of target species than did reserve systems selected using umbrellas. Where data are limited, managers may be consoled by the result that random reserve selection appears to perform at least as well as any of the traditional applications of “umbrella species.”

Pishing is a term used for the “psshh” noise made by bird watchers to elicit close approaches by small birds. Pishing usually attracts multiple species when used in Holarctic habitats, but it produces limited responses in other regions. We propose that responses to pishing occur most often because the sound mimics predator scold calls of species in family Paridae, whose members are resident primarily in the Holarctic. Using both field playback of recorded alarm calls and pishing and bioacoustic analysis of calls, we tested three hypotheses: (1) a generalized mobbing response to parid scold calls has evolved among forest birds in the Holarctic region; (2) pishing generates overt predator mobbing behaviour in diverse avian taxa; (3) pishing generates mobbing behaviour because of its acoustic similarity to parid scold calls. In playback trials in northern California, scolds of local and exotic parids and pishing elicited more vigorous mobbing responses than did the alarm calls of local non-parid species. Parid scolds shared two frequency metrics distinct from non-parid calls, and pishing shared one frequency metric with parid calls that was distinct from non-parid calls. We provide support for a generalized (mobbing) response elicited in Holarctic bird communities by parid scolds that could explain similar close-approach responses to pishing and provide evidence that scold call structural similarity with pishing may underlie the shared behavioural responses. This is the first test of mechanisms underlying pishing responses that also yields an explanation of the geographic variability in strength of response.

Responding to degradation in their original coastal habitat, increasing numbers of lesser snow geese are rearing their broods farther inland. Goslings collected in this inland, fresh water habitat have substantially lower loads of two species of caecal nematodes than do goslings collected in coastal, salt marsh habitat. This likely reflects differences between the habitats in the levels of infective stages of the parasites that are ingested by goslings during their summer foraging. In the spring, several million northward migrating adult lesser snow geese use the coast of Hudson Bay for staging and feeding rather than using more inland habitat because the latter is usually still snow- and icebound. The spring migrants leave behind copious amounts of feces in the coastal marshes that contain the eggs and larvae of the nematodes. By contrast, the inland habitat receives little fecal deposition until mid-summer and then only by the much smaller resident population of nesting lesser snow geese. There is some evidence that the infectious stages of these parasites survive the winter, but multi-year accumulations would only tend to amplify habitat differences in infective loads related to the spring deposition by migrants. The role of migrants in transmitting these nematodes highlights the important point that local host–parasite dynamics must be considered from a broader spatial scale.

Tundra ecosystems are sensitive to disturbance and slow to recover. To account for environmental costs of development in the North, cumulative impacts of roads and dust deposition must be quantified. After a previous study, we re-examined tundra adjacent to the 577-km-long Dalton Highway in northern Alaska to assess 13 y of additional calcareous road dust deposition. Dust loading continues to alter substrate properties and community composition. Moist, acidic, tussock-sedge tundra typically has a soil pH of 4. At the road margin the pH of the fibric horizon had increased to pH 5.5 by 1989 and to pH 6.0 by 2002. Plots adjacent to the road have significantly higher graminoid and Rubus chamaemorus biomass and less moss, evergreen shrub, lichen, and forb biomass. Graminoid cover ranges from 30% in undisturbed tundra to over 80% within 5 m of the road. We observed an 80 g·m⁻² increase in graminoid biomass and a 130 g·m⁻² decline in moss biomass across the study site between 1989 and 2002. Ordinations indicate a broadened zone of dust disturbance in 2002. This evidence of cumulative impacts of dust will improve our evaluation of the ecological costs of future road development in the North.

We studied the biogeography of vascular plants on 10 islands in Laskeek Bay, Haida Gwaii (Queen Charlotte Islands), British Columbia. The islands varied in size from 4.5 to 395 ha and experienced a range of different browse pressures from introduced black-tailed deer (Odocoileus hemionus). We examined how island size interacted with browse pressure in determining the total species counts for individual islands. Numbers of plant species recorded increased with island area. The regression exponent for the log–log plot of species number on island area was 0.18, at the lower end of the range for such exponents. Many species absent from islands < 25 ha in area were characteristic of forest interiors, and consequently part of the increase in richness on larger islands probably was the result of increased forest interior area. Among the islands < 25 ha in area, the normal species–area and species–isolation relationships were reversed, with smaller, more isolated islands supporting more plant species than larger islands and, for a given area, more isolated islands supporting more species than less isolated ones. This reversal of the normal trend appears to be the result of deer browsing. Small, isolated islands were the only islands without deer and were richer, especially in wildflowers, than the larger, less isolated islands. On large islands, total species complement remained as predicted by area because the effect of deer was mitigated by the presence of deer-free refugia on cliffs and in isolated gullies. We conclude that deer are a major factor structuring the island plant communities and that continued protection of island habitats from introduced deer is essential to maintain the native flora of Haida Gwaii.

In this paper we investigate to what extent the occurrence of foliar diseases is affected by plant relative abundance in the understory community of the Los Tuxtlas tropical rain forest, how this changes with season of the year, and how plant species with different life histories present at the understory are affected by disease. To provide a context to these analyses we also include a general description of the floristic composition of the understory community. Using the eight most common species, we found that their relative abundance in each sampling location significantly explained the proportion of diseased plants. Accordingly, using the relative abundance of all plant species, we found that the probability of a host plant species being free of infection showed a significant decrement with abundance. At the plant level, we found that relative abundance had the greatest effect on the variation in leaf area/plant affected by pathogens, although the proportion of explained deviance was only 12%. Seasonality did not affect disease incidence and disease levels per plant. Throughout the year, plant relative abundance was very much lower in lianas, tree seedlings, and palms than in perennial herbs and ferns, and disease incidence was very much higher in the latter two, the most abundant life forms. These results collectively suggest that both intraspecific and interspecific variation in plant relative abundance explain variation in leaf damage by pathogenic fungi in tropical forest understories.

Habitat destruction and fragmentation caused by highways can negatively affect animal populations, but a better understanding of the effects of highways on population genetic structure is still needed to improve conservation plans in urbanized landscapes. We investigated the degree of genetic variability and differentiation within and among seven Rana dalmatina populations located far from highly trafficked roads (non-fragmented populations) and four populations sampled on both sides of a major highway (fragmented populations). The degree of population subdivision was significantly higher among fragmented (F_ST = 0.238) as compared to non-fragmented populations (F_ST = 0.022). Furthermore, in the four fragmented populations, significantly lower allelic richness as compared to non-fragmented populations was observed. Together with potential high levels of road mortality leading to smaller population size, these results suggest that separation by highways not only has reduced the genetic diversity and polymorphism in local populations over two decades, but also has resulted in a higher degree of population differentiation, most likely due to genetic drift.

Joshua tree (Yucca brevifolia) is a distinctive and charismatic plant of the Mojave Desert. Although floral biology and seed production of Joshua tree and other yuccas are well understood, the fate of Joshua tree seeds has never been studied. We tested the hypothesis that Joshua tree seeds are dispersed by seed-caching rodents. We radioactively labelled Joshua tree seeds and followed their fates at five source plants in Potosi Wash, Clark County, Nevada, USA. Rodents made a mean of 30.6 caches, usually within 30 m of the base of source plants. Caches contained a mean of 5.2 seeds buried 3–30 mm deep. A variety of rodent species appears to have prepared the caches. Three of the 836 Joshua tree seeds (0.4%) cached germinated the following spring. Seed germination using rodent exclosures was nearly 15%. More than 82% of seeds in open plots were removed by granivores, and neither microsite nor supplemental water significantly affected germination. Joshua tree produces seeds in indehiscent pods or capsules, which rodents dismantle to harvest seeds. Because there is no other known means of seed dispersal, it is possible that the Joshua tree–rodent seed dispersal interaction is an obligate mutualism for the plant.

We studied spatial and temporal patterns of nitrogen pools and fluxes in soils at treeline and forested sites within three mountain ranges across a 785-km transect in Alaska during 2001–2002. We measured pools of soil mineral (ammonium and nitrate) and organic (amino acid and microbial biomass) nitrogen, in situ rates of net mineralization, net nitrification, net amino acid production, and decomposition, as well as soil carbon turnover in a laboratory incubation experiment. Soils at treeline were mostly colder than forested soils, particularly during fall and over winter, and had reduced rates of nitrogen cycling and litter decomposition relative to forested stands. Treeline soils also had lower rates of potential respiration per unit carbon, suggesting reduced soil organic matter quality relative to forest soils. Therefore, effects of both colder temperatures and poorer substrate quality appeared to suppress rates of nitrogen turnover at treeline. Seasonal patterns of nitrogen turnover were similar across latitudes (i.e., mountain ranges). On average, 70% of total annual net nitrogen mineralization occurred from August through May, suggesting that fall and winter are critical periods for soil nitrogen transformations in both forested and treeline ecosystems. Among mountain ranges, pool sizes and fluxes of nitrogen were similar despite significant variation in growing season length and mean annual temperatures. Soil moisture and soil organic matter quality may have stronger effects on variation in nitrogen cycling than temperature at our sites.