Summary• Selenium (Se) hyperaccumulation has a profound effect on plant-arthropod interactions. Here, we investigated floral Se distribution and speciation in flowers and the effects of floral Se on pollen quality and plant-pollinator interactions.• Floral Se distribution and speciation were compared in Stanleya pinnata, an Se hyperaccumulator, and Brassica juncea, a comparable nonhyperaccumulator. Pollen germination was measured from plants grown with varying concentrations of Se and floral visitation was compared between plants with high and low Se.• Stanleya pinnata preferentially allocated Se to flowers, as nontoxic methylselenocysteine (MeSeCys). Brassica juncea had higher Se concentrations in leaves than flowers, and a lower fraction of MeSeCys. For B. juncea, high floral Se concentration impaired pollen germination; in S. pinnata Se had no effect on pollen germination. Floral visitors collected from Se-rich S. pinnata contained up to 270 lg g , concentrations toxic to many herbivores. Indeed, floral visitors showed no visitation preference between high-and low-Se plants. Honey from seleniferous areas contained 0.4-1 lg Se g)1 , concentrations that could provide human health benefits.• This study is the first to shed light on the possible evolutionary cost, through decreased pollen germination in B. juncea, of Se accumulation and has implications for the management of seleniferous areas.

BackgroundHyperaccumulation, the rare capacity of certain plant species to accumulate toxic trace elements to levels several orders of magnitude higher than other species growing on the same site, is thought to be an elemental defense mechanism against herbivores and pathogens. Previous research has shown that selenium (Se) hyperaccumulation protects plants from a variety of herbivores and pathogens. Selenium hyperaccumulating plants sequester Se in discrete locations in the leaf periphery, making them potentially more susceptible to some herbivore feeding modes than others. In this study we investigate the protective function of Se in the Se hyperaccumulators Stanleya pinnata and Astragalus bisulcatus against two cell disrupting herbivores, the western flower thrips (Frankliniella occidentalis) and the two-spotted spider mite (Tetranychus urticae).ResultsAstragalus bisulcatus and S. pinnata with high Se concentrations (greater than 650 mg Se kg-1) were less subject to thrips herbivory than plants with low Se levels (less than 150 mg Se kg-1). Furthermore, in plants containing elevated Se levels, leaves with higher concentrations of Se suffered less herbivory than leaves with less Se. Spider mites also preferred to feed on low-Se A. bisulcatus and S. pinnata plants rather than high-Se plants. Spider mite populations on A. bisulcatus decreased after plants were given a higher concentration of Se. Interestingly, spider mites could colonize A. bisulcatus plants containing up to 200 mg Se kg-1 dry weight, concentrations which are toxic to many other herbivores. Selenium distribution and speciation studies using micro-focused X-ray fluorescence (μXRF) mapping and Se K-edge X-ray absorption spectroscopy revealed that the spider mites accumulated primarily methylselenocysteine, the relatively non-toxic form of Se that is also the predominant form of Se in hyperaccumulators.ConclusionsThis is the first reported study investigating the protective effect of hyperaccumulated Se against cell-disrupting herbivores. The finding that Se protected the two hyperaccumulator species from both cell disruptors lends further support to the elemental defense hypothesis and increases the number of herbivores and feeding modes against which Se has shown a protective effect. Because western flower thrips and two-spotted spider mites are widespread and economically important herbivores, the results from this study also have potential applications in agriculture or horticulture, and implications for the management of Se-rich crops.

Selenium (Se)-rich plants may be used to provide dietary Se to humans and livestock, and also to clean up Se-polluted soils or waters. This study focused on endophytic bacteria of plants that hyperaccumulate selenium (Se) to 0.5–1% of dry weight. Terminal restriction fragment length polymorphism (T-RFLP) analysis was used to compare the diversity of endophytic bacteria of hyperaccumulators Stanleya pinnata (Brassicaceae) and Astragalus bisulcatus (Fabaceae) with those from related non-accumulators Physaria bellii (Brassicaceae) and Medicago sativa (Fabaceae) collected on the same, seleniferous site. Hyperaccumulators and non-accumulators showed equal T-RF diversity. Parsimony analysis showed that T-RFs from individuals of the same species were more similar to each other than to those from other species, regardless of plant Se content or spatial proximity. Cultivable endophytes from hyperaccumulators S. pinnata and A. bisulcatus were further identified and characterized. The 66 bacterial morphotypes were shown by MS MALDI-TOF Biotyper analysis and 16S rRNA gene sequencing to include strains of Bacillus, Pseudomonas, Pantoea, Staphylococcus, Paenibacillus, Advenella, Arthrobacter, and Variovorax. Most isolates were highly resistant to selenate and selenite (up to 200 mM) and all could reduce selenite to red elemental Se, reduce nitrite and produce siderophores. Seven isolates were selected for plant inoculation and found to have plant growth promoting properties, both in pure culture and when co-cultivated with crop species Brassica juncea (Brassicaceae) or M. sativa. There were no effects on plant Se accumulation. We conclude that Se hyperaccumulators harbor an endophytic bacterial community in their natural seleniferous habitat that is equally diverse to that of comparable non-accumulators. The hyperaccumulator endophytes are characterized by high Se resistance, capacity to produce elemental Se and plant growth promoting properties.

Pseudomonas moraviensis subsp. stanleyae may be useful for stimulating plant growth and for the treatment of Se-laden wastewater.

This is the author manuscript accepted for publication and has undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as