Ecological Entomology

The multivoltine non-native box tree moth, Cydalima perspectalis (Lepidoptera: Crambidae), was accidentally introduced in several continents and exhibited a very rapid expansion in its invaded ranges, colonising around 40 countries in less than twenty years. Initially observed in urban areas, it quickly spread to forest stands, causing significant damage to boxwood trees. If the role of the ornamental plant trade in its fast invasion has been investigated, the insects flight capabilities has thus far been overlooked, limiting estimations of its natural dispersal. In this context, we assessed the flight performance of the adults and investigated potential effects of sex, age and generation using computer-monitored flight mills. Under these controled conditions, we estimated the distance flown per night and over the adults lifespan in both sexes for each generation. The adults were able to fly on average 18 km within their lifespan, even though distances flown were highly variable, including several females capable of performing long-distance flights (up to 150 km). The distances covered were partly correlated with the age and body mass of the adults. Mated females flew longer distances than virgin ones, but long dispersal seemed to limit their fecundity. Finally, the overwintering generation presented the highest flight capabilities with individuals able to cover 22 km on average while the late summer generation covered only 10 km. The box tree moth showed good dispersal abilities that likely played a significant role in its rapid local expansion, while human-mediated dispersal favoured its long-distance dispersal.

Resource limitation is a central ecological phenomenon shaping pollinator behavior, reproduction, and community dynamics. Many studies have looked at these interactions, but few have forced competition and explored species-specific responses. Here, we experimentally evaluated behavioral and reproductive responses of honey bees (Apis mellifera), bumble bees (Bombus impatiens), and mason bees (Osmia bruneri) in controlled single and multispecies foraging environments. Across all treatments, each species exhibited distinct forms of behavioral compensation when exposed to increased interspecific competition. Apis mellifera reduced the number of foraging events in mixed species cages yet showed variation in colony growth. B. impatiens increased foraging effort when co-foraging with other species, but this heightened activity did not translate into increased colony growth. O. bruneri maintained consistent foraging effort across treatments but exhibited reduced reproduction and a marked shift in floral host use, switching from preferred species (Phacelia tanacetifolia and Melilotus alba) to less-preferred alternatives (Collinsia grandiflora and Trifolium incarnatum) when competing with social bees. Regardless of compensatory behaviors, all three species demonstrated reduced reproductive success under competitive conditions. Our findings underscore the importance of evaluating interspecific competition when placing managed bees in natural or semi-natural habitats to avoid inadvertently stressing wild or managed bee populations and taking into consideration species-specific responses in addition to community level responses to competition.

Frugivorous insect larvae are dependent on fruiting plants for development, leading to complex host-parasite interactions that may be influenced by many factors at various scales. We compared the relative effects of factors at the individual, neighborhood, and landscape scales in forest patches. Our results suggest that in areas like upstate New York, where agricultural land uses are dominant, individual scale factors are the most influential. Specifically, parasitism increased with host fruit crop size, but was not associated with host species richness or proximity to forest edge. Notably, the most parasitized hosts were non-native species, including Frangula alnus Mill. (Glossy Buckthorn), indicating a potential role of invasive species to shape host-parasite interactions in our system. Our results underscore the importance of host-specific traits in structuring parasitism patterns and suggest management could consider both the ecological context of host traits and the influence of invasive species at multiple scales.

Carnivorous Nepenthes plants capture and digest prey in fluid-filled pitchers, whose fluid level is subject to fluctuations caused by rain and evaporation. It was found that N. rafflesiana pitchers can reduce these fluctuations by regulating their fluid volume and composition. Here, we investigate whether and to what extent this ability occurs in other Nepenthes species varying in morphology, trapping mechanisms, habitat, lifespan and nutrient sources. We hypothesised that species more exposed to fluid level fluctuations or relying more on fluid for prey capture would exhibit more intensive regulation. In six Bornean Nepenthes species, we quantified the effect of pitcher fluid level on prey-capture efficiency, the natural fluctuations in fluid volume, and the plants response to manipulations of the fluids volume, concentration and pH. The prey capture of Nepenthes species depended to varying degrees on the pitcher fluid level: in species with waxy inner walls, capture efficiency decreased at high fluid levels, as prey could escape easily. In contrast, species with wax-free walls and steep peristomes retained prey efficiently, regardless of fluid level. Although all Nepenthes species were able to secrete and absorb water and electrolytes, the regulation rate varied considerably. The rates of the various secretion and absorption processes (concentration/volume-dependent rates of electrolyte/water transport) were positively correlated with one another, suggesting that they all depend on the activity of the pitcher glandular epithelium. While differences and similarities in fluid regulation may be partly due to phylogeny, the observed species differences can also be explained by the pitchers trapping mechanisms and ecological factors. We found weaker regulation in Nepenthes species growing in sheltered understorey habitats, and in species utilising alternative nutrient sources, which may reduce their reliance on pitcher fluid. Our study highlights the diversity of ecological strategies in pitcher plants and demonstrates how these are influenced by environmental conditions. Lay SummaryNepenthes pitcher plants have fluid-filled pitchers to capture and digest prey. Can they control the volume and composition of their fluids? We found that all six species studied regulated their fluids, but to varying degrees. Species differences can be explained by the plants habitat, pitcher morphology and nutrient acquisition strategies.

Diapause is a critical adaptive strategy that enables temperate mosquito species to survive adverse environmental conditions and maintain population persistence across seasons. In Culex pipiens, diapause plays a key role in overwintering and influences the seasonal dynamics of arbovirus transmission. However, diapause expression is often assessed using single traits, limiting our understanding of its integrated physiological basis and variation among populations. In this study, we investigated the behavioural, morphological, and reproductive signatures of diapause across three laboratory strains of Culex pipiens (Mogden, Pirbright, and Pirbright Hybrid) reared under diapause-inducing (10 {degrees}C), cold (14 {degrees}C), and control (26-27 {degrees}C) conditions. We quantified blood-feeding behaviour, wing size as a proxy for somatic growth, and spermatheca size as an indicator of reproductive development. Diapause-inducing conditions resulted in a coordinated phenotype characterised by strong suppression of blood-feeding, increased somatic size, and marked inhibition of reproductive development. Mosquitoes reared at 10 {degrees}C exhibited near-complete feeding inhibition and significantly reduced spermatheca size, consistent with reproductive arrest, while those reared at 14 {degrees}C showed intermediate phenotypes. In contrast, control mosquitoes displayed active feeding and fully developed reproductive structures. Wing size increased progressively with decreasing temperature, with the largest individuals observed under diapause-inducing conditions. When analysed together, wing size and spermatheca development exhibited opposing responses across temperature treatments, revealing a strong negative association and indicating a trade-off between somatic growth and reproductive investment. This integrated response supports the interpretation of diapause as a coordinated life-history strategy involving resource reallocation towards survival. Additionally, diapause expression varied among strains, with the Mogden strain showing reduced sensitivity compared with Pirbright and hybrid populations, highlighting the role of genetic background in diapause plasticity. These findings demonstrate that diapause in Culex pipiens is a multi-trait, plastic phenotype with important implications for overwintering success and the seasonal dynamics of arbovirus transmission in temperate regions.

To implement adequate nutritional landscapes to bumble bees and inform conservation efforts, we need a better understanding of how different pollen sources impact bumble reproductive performances and phenology. We assessed the impact of multiple native pollen on bees reproductive success by feeding 40 wild-caught Bombus impatiens and B. griseocollis queens with either commercially wildflower pollen mix commonly used to breed Bombus queens (control group) or 75% red maple pollen (Acer rubrum) and 25% pollen mix (RM group). In a second experiment, we monitored the reproductive success of 45 microcolonies of workers, harvested from B. impatiens colonies. Microcolonies were fed either: 1) pollen mix (N=15), 2) pollen mix and hawthorn (Crataegus canadensis, N=15), or 3) pollen mix and sumac (Rhus typhina, N=15). In experiment 1, RM significantly increased reproductive outputs for both species. RM-fed founding queens were twice more likely to produce workers, and produced them significantly earlier, though total number of workers did not vary between groups. However, RM-fed founding queens produced significantly more brood cells, males and gynes than control colonies, and produced them significantly earlier. In experiment 2, we found an interaction between initial colony diet and current microcolony diet. Overall, workers harvested from RM-fed colonies produced significantly more brood cells and males than workers harvested from control colonies. However, when fed hawthorn pollen during experiment 2, workers from control colonies performed as well as workers from RM-fed colonies. This suggests carry-over effects from early pollen sources provided to colonies on workers, which may be circumstantially offset by pollen sources with specific nutrient profile. Using detailed nutritional information of each pollen (including data on 32 nutrients), we provide nutritional requirements of pollen to support bumble bee reproduction, especially regarding vitamin B3 and vitamin C contents. This study will help inform land managers and captive breeding programs on the ideal pollen sources for bumble bees to ensure reproductively successful colonies.

Invasive ambrosia beetles and fungal pathogens threaten forest ecosystems worldwide, exemplified in Hawaii by the widespread loss of keystone species [o]hia (Metrosideros polymorpha), due to Rapid [O]hia Death (ROD). A unique occurrence of five ambrosia beetle species (one native, four introduced) that vary in their symbiotic relationships with two introduced fungal pathogens provide an opportunity to test hypotheses of how opportunistic symbioses facilitate disease dynamics involving dominant forest trees. ROD is caused by two novel Ceratocystis fungal pathogens whose spores can spread via association with ambrosia beetles as they bore into [o]hia trees. We examined beetle-pathogen interactions of all five ambrosia beetle species in three ROD-affected regions on Hawaii Island, and used quantitative PCR (qPCR) to provide the first molecular confirmation of the two ROD pathogens associated with the exterior, mycangia, and gut of each beetle species. Results from generalized linear models and correlation networks show that pathogen acquisition and transport, including the potential for consumption and the presence of the pathogens, are determined by beetle invasion status and mycangia morphology. A niche construction framework suggests that both varying symbioses and opportunism facilitate disease spread, with the three invasive Xyleborus species emerging as key disease vectors. Identifying the beetle species that are more likely to contribute to disease spread, and understanding their biology as vectors, can inform targeted conservation strategies for [o]hia and for insect-pathogen threats in forests worldwide, and illustrates the potential ecosystem-level impacts of novel and opportunistic symbioses between globally distributed invasive vectors and pathogens.

Riparian environments are characterised by a high diversity of arthropod species, linking the aquatic with the terrestrial ecosystem. One of the dominant arthropod predators in this ecotone, carabid beetles are of particular interest as they may also act as facultative scavengers, feeding on carrion deposited along the shoreline. To test whether carabids feed on carrion, we examined the feeding preferences of the abundant riparian carabid Bembidion elongatum in a laboratory feeding trial, offering freshly killed and 24 hours post mortem Drosophila melanogaster. We subsequently assessed the detection probability of ribosomal prey RNA and DNA in predator gut contents using Drosophila-specific RT-PCR and PCR assays, and quantified nucleotide abundance by quantitative real-time PCR at 0, 3, 6, and 12 hours post-feeding. In the feeding experiment, B. elongatum showed a significant preference for fresh over carrion prey. Following consumption, the quantities of prey DNA and RNA in the predators gut declined over a 12-hour post-feeding period. However, no differences were detected in prey DNA or RNA quantities between individuals fed fresh prey and those fed carrion. Only immediately after consumption was the DNA:RNA ratio significantly lower in individuals fed fresh prey compared to those fed carrion while this difference was not observed at later time points. Overall, our results indicate that ingested ribosomal prey RNA in predators is present in high quantities, and that the DNA:RNA ratio is not a suitable indicator for distinguishing between consumption of carrion and fresh prey.

Nutrition plays a fundamental role in insect physiology, yet whether growth, immunity and stress responses share a common nutritional optimum remains poorly understood. Using the Geometric Framework for Nutrition, we investigated how dietary protein and carbohydrate balance regulate multiple physiological functions in the specialist herbivore Plutella xylostella. Growth was jointly determined by protein and carbohydrate intake, with maximal biomass accumulation constrained primarily by protein availability. Immune traits showed distinct nutritional optima: total hemocyte counts and phenoloxidase activity were maximised at intermediate protein-carbohydrate (P:C) ratios, whereas survival following Bacillus thuringiensis challenge depended predominantly on protein intake. Antioxidant and detoxification responses likewise varied non-linearly across the nutrient landscape, with individual enzyme systems each exhibiting distinct nutrient-dependent optima. Importantly, no single macronutrient balance simultaneously maximised growth, immune performance and oxidative homeostasis, demonstrating that these functions have divergent nutritional requirements. Together, these findings show that macronutrient balance generates competing physiological demands, offering an integrated view of nutrient-dependent regulation of growth, immunity and stress physiology in a specialist herbivore.

Adult mosquitoes rely on plant-derived sugars for survival, reproduction, and flight, yet the plant taxa that mosquitoes encounter in nature and the odors that make those plants attractive remain poorly understood. Most studies of mosquito attraction to plant odors have focused on candidate plants selected a priori, rather than plants linked to field-collected mosquitoes. Here, we combined plant DNA barcoding, semi-field behavioral assays, and volatile profiling to identify field-associated plant resources relevant to Aedes aegypti. Plant DNA recovered from mosquitoes collected across three Florida counties revealed broad plant associations, including 90 genera spanning 37 families, with several taxa recurring across counties or appearing prominently within particular localities. Behavioral experiments in semi-field sticky-trap assays found that five field-associated plant taxa were significantly attractive relative to blank controls, indicating that taxa associated with mosquitoes in nature can also function as attractive cues under semi-field conditions. GC-MS analyses of headspace collections from 42 plant taxa detected 211 volatile compounds and revealed substantial variation in both total emission rate and odor composition among taxa. Although several compounds, including -pinene, limonene, 4-ethylacetophenone, 2-ethyl-1-hexanol, 4-ethylbenzaldehyde, and caryophyllene, were broadly distributed across plant groups, volatile profiles differed significantly among taxa and shared compounds often occurred at markedly different proportional abundances. The five behaviorally tested taxa likewise showed both overlap and divergence, sharing 17 compounds across all five taxa while differing in dominant constituents and total emissions. Together, these results show that Ae. aegypti interacts with a diverse set of plants in the field, and suggests nectar-seeking is shaped not simply by plant identity or total odor abundance, but by the composition and proportional structure of plant odors.

In sharp contrast to biological control that seeks control of pests below economic levels, eradication programs seek total extermination of pest species. However, eradication programs are often undertaken without knowledge of the pests potential geographic distribution and relative abundance (i.e., geographic risk assessment), often leading to eradication efforts conducted across areas much greater than the potential permanent range of the target species. To demonstrate this, we review eradication efforts against the tropical New World screwworm (Cochliomyia hominivorax) in the SE USA and Mexico, the exotic sub-tropical tephritid Mediterranean fruit fly (Ceratitis capitata, medfly) in California, and the exotic tropical pink bollworm (Pectinophora gossypiella, PBW) in the SW USA. We use mechanistic physiologically based demographic models (PBDMs) that are physiologically based time varying life tables to decompose the systems and to assess risk under extant and climate change scenarios.

Assessing the availability of protected habitats for at-risk species is needed to determine how the spatial distribution of land cover and land use shapes conservation outcomes. The Loammi skipper (Atrytonopsis loammi) is a nonmigratory, prairie-associated butterfly that has experienced a significant reduction in its formerly widespread southeastern United States distribution over the past several decades. Observations in recent years have been limited to a small number of isolated Florida populations, and it is unclear how much suitable habitat remains and what proportion of that habitat is protected. Here we used publicly available community science data and collected specimens to identify the predominant land cover types occupied by A. loammi. We then quantified the extent to which each of these land cover types overlap with protected areas to estimate the proportion of the butterflys current distribution that occurs on managed conservation lands. Our findings show that A. loammi relies heavily on protected lands, with over half of the total area of the most strongly associated habitats located within protected areas. Given the continued expansion of human-modified landscapes in Florida and the reliance of A. loammi on protected lands, the long-term persistence of A. loammi will likely depend on the conservation and effective management of its remaining suitable habitat. Implications for insect conservationOur results highlight the critical role of protected areas in sustaining at-risk insect populations, particularly for species with limited dispersal and shrinking ranges. More broadly, they suggest that conserving and restoring protected habitats while maintaining connectivity are essential strategies for effective insect conservation in rapidly developing regions.

Understanding how closely related species coexist in highly seasonal and unpredictable environments is central to studies of ecological differentiation and niche partitioning. We investigated the feeding ecology of sympatric populations of Microcebus murinus and M. griseorufus within a contact zone in Andohahela National Park, southeastern Madagascar, across dry and wet seasons. Using a combination of direct behavioral observations (1,611 feeding records), fecal sample analyses (n = 56), and vegetation phenology surveys, we quantified dietary composition, seasonal shifts in resource use, and habitat-related variation. Seasonal changes in diet were pronounced, with dry-season feeding dominated by exudates and wet-season diets incorporating greater proportions of fruit and flowers, closely tracking phenological patterns at both sites. Diets of both species were dominated by plant resources, but consistent interspecific differences in dietary strategy were evident. Although both species consumed comparable proportions of insect prey, M. murinus showed pronounced wet-season increases in the use of high-sugar, carbohydrate-rich floral resources (15.9%) and hemipteran-associated honeydew (30.6%). In contrast, M. griseorufus relied more consistently on predictable exudates throughout the year. Fecal analyses supported observational data but revealed differences in the detectability of dietary components, with increased representation of invertebrates in the wet season and seeds in the dry season. These results indicate substantial dietary overlap but consistent differences in resource use, suggesting that coexistence is facilitated by fine-scale trophic differentiation within a broadly shared omnivorous niche. Such subtle but persistent differences in feeding strategy likely reduce competitive overlap and enable continued sympatry in a climatically variable and resource-limited system.

Devils gardens are among the most striking ant-plant mutualisms in Amazonian forests. In this system, the tree Duroia hirsuta is associated with the ant Myrmelachista schumanni, which actively removes neighboring vegetation and maintains nearly monodominant patches of the host plant. Despite the apparent efficiency of this system, repeated field observations at the Tiputini Biodiversity Station, Yasuni Biosphere Reserve, revealed that individuals of Palicourea alba recurrently occur within active devils gardens. Palicourea alba closely resembles dead plant material, exhibiting leaf morphology and coloration that strongly mimic the surrounding litter layer, and appears to be uncommon outside these gardens. To our knowledge, this "dead-leaf" masquerade has not been previously documented in this intensively studied system, making it a particularly striking and unexpected observation. To evaluate whether this masquerade facilitates persistence within devils gardens, we surveyed 35 gardens and recorded P. alba in 19 (52.8%). When present, P. alba covered on average 27% of plot area, while mean herbivory across sampled leaves remained low (8.6%). Generalized linear mixed models showed that P. alba cover decreased significantly with increasing herbivory (F = 8.09, p = 0.0159), whereas herbivory increased with leaf-litter cover (F = 8.73, p = 0.0120). Field observations further revealed that many individuals are nearly indistinguishable from dry leaf litter, suggesting a role for visual crypsis or masquerade. Together, these results indicate initially, that the persistence of P. alba within devils gardens is mediated by a multi-layered ecological filtering process. First, masquerade likely reduces detection by M. schumanni, allowing seedlings to escape ant-mediated removal. Second, low herbivory suggests either enemy avoidance or reduced apparency to herbivores within the simplified understory. Third, spatial heterogeneity in leaf-litter cover may create microhabitats where both ant activity and herbivore pressure are modulated, reinforcing establishment success. This system thus represents a previously undocumented mechanism in which plant-litter resemblance enables persistence within a highly structured, biotically filtered habitat, highlighting how subtle trait-mediated interactions may modulate outcomes in otherwise strongly deterministic mutualisms.

AimWe develop Atta leafcutter-ants as bioindicators that respond at fine scales to geographic and seasonal climate changes in the Americas, thereby addressing the paucity of versatile insect bioindicator systems capable of monitoring climate in both Southern and Northern Hemispheres. LocationAmerican tropics and sub-tropics from latitudes S33.6{degrees} to N33.2{degrees}, with case studies from Colombia, Mexico, and southern USA. Time Period2012-2024. Taxon StudiedAtta leafcutter-ants. MethodsWe elucidate biogeographic patterns of mating-flight phenology of Atta leafcutter-ants across the entire Atta range from Uruguay/Argentina to the USA, using 2335 records of Atta reproductives from the community database iNaturalist, then ground-truth these patterns by comparison with (i) mating-flight records (n=806) from the Atta literature; and (ii) mating-flight observations (n=836) accumulated by a consortium of experts who have researched Atta for a combined 1000+ work-years. Onset of mating flights can be timed with great precision in Atta populations because mass-mating flights are synchronized and triggered by the first major rainfall of a rainy season. ResultsBiogeographic patterns in climate-dependent mating-flight phenologies recorded at iNaturalist are corroborated by observations accumulated in the literature and by Atta experts. Analyses reveal so-far unknown gradients in mating-flight phenology (e.g., Colombia to USA) that are correlated to geographic climate gradients, and season switches of mating flights from early to late in the year between proximate Atta populations (e.g., 200 kilometers apart), for example in the climatically complex Andean regions of Colombia. Regional differences in Atta mating-flight phenology correspond to temporal differences in rainfall between ecoregions of Colombia. Main ConclusionsAtta ants are tractable insect bioindicators to monitor climate impacts with detailed spatial and temporal resolution across a 9200-kilometer trans-equatorial transect in the Americas. We outline future research directions to explore climate-dependent biology using the continuously growing, and now ground-truthed, information at iNaturalist on Atta mating behavior.

Xylosandrus crassiusculus (Motschulsky), the granulate ambrosia beetle, was one of the first highly-destructive ambrosia beetles introduced into the southern U.S in the 1970s where it was found in South Carolina (Kovach 1986). The Redbay ambrosia beetle, Xyleborus glabratus Eichhoff, was first detected in the U.S. in South Georgia in 2002. This beetle and its associated fungi, the laurel wilt fungus Raffaelea laurelensis and others have caused substantial destruction to native redbay (Persea borbonia) in GA, SC, FL and elsewhere. This beetle-pathogen complex also poses a threat to commercial avocado production in the U.S., Central and South America as well as to valuable other Persea spp. and related plants (Laureacea) that are known hosts. As an addition here, 10 years of the spring appearances (Fig.1) of X. crassiusculus in North Florida is offered for future comparisons. A second unusual appearance is the finding and working with UV mulch and ethanol, as a surprising attraction of X. crassiusculus and other ambrosia beetles including X. glabratus. It was also found that the ambrosia beetles do not respond to yellow and green as expected by most. Also, adding burlap was found to be attractive (increases dead and dying appearing trees) as is silver metallic like UV mulch, while camouflage (camo) was found to work like yellow and green. These occurrences led to the invention and development of UV mulch with new traps to better monitor ambrosia beetles. New traps led to new uses for yellow, green and camo to monitor and decrease damage and losses from ambrosia beetles. The data are presented as evaluated and appear in the figures, discussion and a supplemental section. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=139 SRC="FIGDIR/small/733798v1_fig1.gif" ALT="Figure 1"> View larger version (30K): org.highwire.dtl.DTLVardef@1006101org.highwire.dtl.DTLVardef@1e0a3d6org.highwire.dtl.DTLVardef@1244d1borg.highwire.dtl.DTLVardef@423cb7_HPS_FORMAT_FIGEXP M_FIG O_FLOATNOFigure 1:C_FLOATNO Relative timing of annual emergence of Xylosandrus crassiusculus in north Florida. Collected over 10 years using 5 Baker traps with a 10% ethanol/water solution. Data are from years as marked. Note: data from year 2003 was not collected. C_FIG

Many insect populations are experiencing thermal stress as a result of global change, making it imperative to investigate how their relationship with other organisms will be impacted by heat disturbances. Microbial symbionts, such as bacteria, have the potential to enhance or inhibit an insect's thermal tolerance. Solitary bee larvae host bacteria within their food stores ("pollen provisions"), which have been shown to benefit survival and development; however, it is unclear how heatwaves brought about by climate change will impact their relationships with these bacterial partners. In this study, we subjected blue orchard bee (Osmia lignaria) eggs and larvae to a 4-day heatwave (35 {degrees}C daytime:22 {degrees}C nighttime) or kept them at control temperatures (25 {degrees}C daytime:15 {degrees}C nighttime), then returned all bees to control temperatures for a 5-day recovery period. We assessed bacterial communities within pollen provisions and larval development stage pre-heatwave (Day 0), immediately post-heatwave (Day 4), and following the recovery period (Day 9). Bacterial community composition, diversity, and abundance were resilient to heat stress, but larval bees developed faster when subjected to a heatwave. This finding refutes the hypothesis that bacteria within pollen provisions modulate blue orchard bee responses to heat, suggesting instead that developmental effects could be more largely shaped by bee physiology or interactions with microorganisms other than bacteria.

Yeasts thriving in floral nectar have a proven ecological engineering capacity, as they modify the internal environment and the rewards of individual flowers in ways that influence how plants and pollinators interact. This floricentric perpective, however, produces only a partial view of the potential role of yeasts in plant-pollinator interactions, and little is known on their possible engineering effects on whole inflorescences. This study assessed yeast ability to shape features of the vertical inflorescences of Gladiolus illyricus (Iridaceae), a Darwins syndrome plant (acropetalous protandrous flowers pollinated by insects foraging from bottom to top). When yeasts were excluded from inflorescences a declining gradient in nectar sugar concentration built up from the oldest, female-stage flowers at bottom up to the youngest, male-stage ones at top. This plant-intrinsic gradient was reversed in inflorescences with yeasts, where sugar concentration increased from the female-stage flowers harboring dense yeasts population to male-stage ones with few yeasts. The two species of yeasts involved (Metschnikowi reukaufii and M. gruessii) differed in their effects. Results add support to the growing consensus that specialized nectar yeasts can obfuscate intrinsic patterns of intra and interspecific variation in flowering features of plants at the level of individual flower, inflorescence, population and plant community.

Tropical Atta leafcutter-ant colonies reach a maximum lifespan of 15-20 years because each colony has only a single queen, queens live for maximally 15-20 years, and colonies do not requeen after queen death. At the northern range limit of leafcutter ants in Louisiana, however, we found that the mortality rate of Atta texana colonies is consistent with a life expectancy of 65-75 years and a maximum colony lifespan exceeding 100 years. Annual mortality of established A. texana colonies in Louisiana is only 1-2% compared to 10-25% in tropical Atta species. Exceptional longevity of A. texana colonies is due to (i) absence of Nomamyrmex army-ant predators specialized to raid Atta colonies; (ii) polygyny of A. texana colonies, creating a unique social organization in A. texana colonies that differs from the monogynous colonies typical for all other Atta species; and (iii) adoption of young queens by old colonies, possibly facilitated by low genetic diversity in range-limit populations of A. texana. Centenarian longevity of A. texana colonies was so far unknown because, from the 1930s to the turn of the century, research prioritized eradication of A. texana as a forestry pest, and landscape-wide pesticide application by aircraft eliminated old A. texana colonies. After the most harmful pesticides were banned 30-50 years ago, A. texana populations recovered, yet A. texana required decades to regain its earlier niche in the forest ecology. In undisturbed habitat, well-established old A. texana colonies with large territories appear to recruit young queens especially well. Old age of a colony therefore does not necessarily increase mortality as in typical lifespan-limiting processes where age drives senescence and mortality (old age [->] mortality); instead, well-rooted residency that is correlated to age of an A. texana colony increases the chance of rejuvenation via queen adoption, reduces or eliminates reproductive senescence of a colony, and consequently prolongs colony lifespan (well-established residency [->] rejuvenation [->] long life). Large size attained in old age imparts potentially unending life.

Across species, changing environmental conditions are altering major phenological milestones. In seasonal environments, this can involve shifts in the timing of activity onset, growth, reproduction, and senescence; however, predicting these responses can be challenging and often requires species-specific information. In this study, we investigated phenological patterns of bumble bee colony development using observations from wild nests of two common species, Bombus griseocollis and B. impatiens. Using observations of nest traffic, we documented dates of nest-searching, peak worker activity, first gyne (new queen) production, and colony senescence over three years of field data collection. We also tested whether colony growth was density dependent, because longstanding life history models have shown that colony growth patterns determine the optimal timing of reproduction in constant environments. Here, we present a novel extension of these models, using them to infer that density-independent colony growth should be associated with extended phenology in warmer years, while density-dependent should have consistent phenology. In total, we found 79 wild nests, including 34 reproductive colonies (19 B. griseocolis and 15 B. impatiens). Bombus impatiens colonies were larger, had a longer activity period, and showed density-independent growth and later reproduction in warmer years, while B. griseocollis colonies were smaller, shorter-lived, and showed density-dependent growth and similar dates of reproduction across years. Both species had higher gyne production in warmer years. These results highlight the role of life history theory for predicting interspecific variation in phenological changes, as well as a research framework for exploring possible evolutionary mismatches of existing life histories in changing environments.