Infections of one host by multiple parasites are common, and several studies have found that the order of parasite invasion can affect both within-host competition and disease severity. However, it is unclear to what extent coinfection timing might be important to consider when modeling parasite impacts on host populations. Using a model system of two viruses infecting barley, we found that simultaneous infections of the two viruses were significantly more damaging to hosts than sequential coinfections. While priority effects were evident in within-host concentrations of sequential coinfections, priority did not influence any parameters (such as virulence or transmission rate) that affect host population dynamics. We built a susceptible-infected model to examine whether the observed difference in coinfection virulence could impact host population dynamics under a range of scenarios. We found that coinfection timing can have an important but context-dependent effect on projected host population dynamics. Studies that examine only simultaneous coinfections could inflate disease impact predictions.

Herbivores may significantly reduce plant populations by reducing seed set; however, we know little of their impact on seed movement. We show for the first time that the receptacle-feeding weevil Rhinocyllus conicus not only reduces seed production by the invasive thistle Carduus nutans but also inhibits release and subsequent wind dispersal of seeds. These effects generate large, though different, impacts on spatial spread and local abundance in two populations with differing demography, located in the United States and New Zealand. Furthermore, the mechanism is context dependent, with the largest effects through increased terminal velocity in the United States but through reduced seed production in New Zealand. Our results show that the benefit of biocontrol programs may have been underestimated; screenings of potential biocontrol agents should examine effects on pest dispersal and spread, as well as on abundance.

Background: Non-random seed release caused by plant responses to weather conditions is important for seed dispersal. Much is known about the effects of wind speed and turbulence, but our understanding of the effects of water loss on seed release is either qualitative, or indirect and phenomenological. Aims: To quantify the empirical relationship between water loss and seed release. Methods: Capitula of the invasive thistles Carduus acanthoides and C. nutans were collected from the field and treated for either 0, 1, or 2 days in the laboratory at three different vapour pressure deficit levels (3.4, 9.5, and 17.0 hPa) to cause a range of water loss values. Total seed release was quantified before and during wind tunnel trials. Results: Water loss was the only significant predictor of whether or not capitula released any seeds. The number of seeds released was predicted by water loss, capitulum diameter, and herbivore damage, with the same amount of water loss having less effect on larger capitula. Conclusions: These results represent an important step towards using weather data to predict seed release for many xerochastic species. Incorporating the effects of water loss on seed release into mechanistic seed dispersal models will greatly improve predictions of when and how far seeds disperse. While initiation of dispersal in some species is more likely during precipitation events (Pufal and Garnock-Jones 2010), a majority of wind dispersed angiosperms and gymnosperms are xerochastic, meaning that drying enhances seed abscission and release (Greene et al. 2008). For many Asteraceae species, including the invasive thistles Carduus nutans L. and Carduus acanthoides L. (Figure 1), drying causes cohesion tissues located on the outer side of the involucral bracts to lose turgidity and buckle, causing the bracts to be lowered away from the seeds and thus exposing seeds to the wind (Fahn 1990). Drying may also cause contraction of the receptacle away from seeds (Smith and Kok 1984).The ubiquity of xerochastic plant species suggests that seed dispersal models could be made more realistic and accurate for a wide range of species if weather data describing potential evaporation could be used to mechanistically predict seed release. Two important pieces of information need to be re-examined to accomplish this goal. First, we must predict water loss from inflorescences using weather data. A number of models already exist that use weather data to predict water loss from entire plants (de Bruin and Holtslag 1982;Sumner and Jacobs 2005), and even from capitula separately from vegetative structures (Guilioni and Lhomme 2006). The second necessary piece of information, the relationship between water loss and seed release, must still be documented and is the focus of this paper.Currently, our understanding of the effects of water loss and weather on seed release is either qualitative or phenomenological. Physiologically, we understand what effects dry conditions have on capitula (Smith and Kok 1984;Fahn 1990) and seed attachm...

The spatial arrangement of plants in a landscape influences wind flow, but the extent that differences in the density of conspecifics and the height of surrounding vegetation influence population spread rates of wind dispersed plants is unknown. Wind speeds were measured at the capitulum level in conspecific arrays of different sizes and densities in high and low surrounding vegetation to determine how these factors affect wind speeds and therefore population spread rates of two invasive thistle species of economic importance, Carduus acanthoides and C. nutans. Only the largest and highest density array reduced wind speeds at a central focal thistle plant. The heights of capitula and surrounding vegetation also had significant effects on wind speed. When population spread rates were projected using integrodifference equations coupling previously published demography data with WALD wind dispersal models, large differences in spread rates resulted from differences in average horizontal wind speeds at capitulum caused by conspecific density and surrounding vegetation height. This result highlights the importance of spatial structure for the calculation of accurate spread rates. The management implication is that if a manager has time to remove a limited number of thistle plants, an isolated thistle growing in low surrounding vegetation should be targeted rather than a similar size thistle in a high density population with high surrounding vegetation, if the objective is to reduce spread rates. 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 The influences that differences in wind speed caused by different vegetation elements (e.g. conspecifics or other surrounding vegetation) may have on population spread rates of wind dispersed invasive species are unknown. Here, we examine the effects of differences in conspecific density and surrounding vegetation height together on wind speeds, and simulate population spread of two wind dispersed species from varying source environments. Understanding the effects of these factors on invasive spread has the potential to inform management decisions. For instance, optimal management strategies depend not only on plant demography under different growth conditions, but also on expected spread rates (Menz et al. 1980 Prevention of seed escape. Both study species are invasive weeds, so efforts were taken to prevent seeds from escaping. C. nutans capitula were tightly wrapped in fine pollen bag material. C. acanthoides produces smaller, more numerous capitula, so adhesive spray was used to prevent seed release.Statistical Analysis. Wind speed data were analyzed using linear mixed effects models in R (R Development Core Team 2009). Species, surrounding vegetation height, thistle patch array, and measurement height were used as explanatory variables. Weather station wind speeds were used as a covariate to correct for differences in ambient wind sp...

Nonrandom seed release is an important determinant of how far seeds disperse, but the mechanisms that promote wind-related seed release under varying atmospheric conditions are poorly understood. We explored the use of particle image velocimetry (PIV) to gain a better mechanistic understanding of seed release by visualizing the flow velocities and vorticity in a two-dimensional slice of air around inflorescences. Pilot data taken in a wind tunnel show gradients in air velocity at the top of Carduus nutans capitula that may contribute to lift generation. Additionally, von Kármán vortex streets (vortices of opposite spin that are shed from the wake of an object) were observed shedding from capitula, which cause lateral forces on capitula and increase turbulence downwind at other locations. Avenues for further research include using PIV to examine the mechanisms of seed release and dispersal in wind tunnels and in the field. We found PIV to be a promising method to further explore the mechanisms behind seed release in wind dispersed plants, and a technique rich with opportunities for collaborations between plant dispersal ecologists and fluid dynamics specialists.

Plants and animals host many microbial symbionts, including both pathogens and mutualists. However, most experimental studies include only one symbiont, and few examine interactions of more than two microbes with their host. Here, we examined whether coinfection with two pathogens causes a synergistic reduction in the benefits that hosts receive from a microbial mutualist. We also measured the effects of a microbial mutualist on the within‐ and between‐host competition between coinfecting pathogens. We manipulated the presence of Clover yellow vein virus (ClYVV), Bean common mosaic virus (BCMV), rhizobia bacteria, and nitrogen fertilizer in common beans (Phaseolus vulgaris). We found asymmetric, context‐dependent interactions among the three microbial symbionts and their host. Coinfection with both viruses led to greater than additive negative effects on the amount of nitrogen that plants received from rhizobia. Rhizobia colonization decreased immune signaling in singly infected plants, but not in coinfected plants. Compared to single ClYVV infection, ClYVV reached higher concentrations within hosts coinfected with BCMV, but only in the presence of rhizobia. Coinfection increased BCMV vertical transmission rates for plants without supplemental nitrogen, but overall vertical transmission opportunities were not affected due to reduced seed production. Examining interactions between multiple microbes sharing a host can reveal important insights about nutrient cycling, disease severity, and pathogen epidemiology.

Mechanistic models of seed dispersal by wind include terminal velocity as the main seed characteristic that influences the dispersal process and hence the resulting dispersal kernels and spread rates. Accurate measurement of the terminal velocity of seeds is therefore pivotal. However, compression during shipment through the post or during storage between collection in the field and terminal velocity measurements in the lab may affect these measurements. To evaluate the effects of shipment and storage on terminal velocity measurements, capitula of Carduus nutans, an invasive thistle species from Eurasia, were stored for 1–5 years and subjected to three different packing treatments. Seeds from capitula were then assessed for terminal velocity values, plume area, seed mass, wing loading, number of filaments per pappus, qualitative assessments of pappus damage, and number of intact dispersal units per capitulum. Compression significantly increased seed terminal velocity. However, storage duration for 1–5 years did not cause a significant increase or decrease in any of the response variables. The compression treatment was validated by shipment of seeds from New Zealand to the United States. When capitula that are to be used for terminal velocity measurements are stored or shipped, they should be packaged in incompressible containers to avoid damage to the fragile dispersal structures. Studies using capitula that were originally collected and stored for other purposes, such as transcontinental demographic studies, should rescale observed terminal velocity values to take into account possible damage.

No abstract