The emission of volatile organic compounds (VOCs) from rotting fruits and wilting flowers.

Abstract

Plants and the microbes that feed on them produce a high diversity of volatile organic compounds (VOCs) that often mediate interspecies communication with other organisms. Two main functions of VOCs emitted by plants are to attract pollinating animals to flowers and seed dispersers to fruits; while volatiles emitted by microbes that feed on plant material can be used as cues by animals searching for food and can also facilitate dispersal of the microbes. The emission of VOCs from flowers and fruits often shows temporal changes that are characteristic for the different ripening stages of the plant organs. In many plants the VOC emission not only increases but also changes in its chemical composition during flower opening and fruit ripening respectively. In addition, VOC emission into the atmosphere carries information about the physiological status and stresses of the plant. However, the vast majority of studies that deal with the function of VOCs, e.g. for attracting pollinating animals or seed dispersers, focus only on those components that are emitted during the time when the plant shows the highest attractiveness for interaction partners. It is reasonable to believe, however, that the decisions made by animals, in terms of host preference and selection, are not only based on the chemical components that are emitted during times of optimal (nutritional) condition. The decision to utilize a flower/fruit is most likely also based on components that indicate to animals that a food source is not worth using, unpalatable, or toxic. For example, early stage flowers and unripe fruits have a low nutritional value, and late stage (wilted) flowers and rotten fruits may, in addition to already depleted resources, also contain toxic chemicals from microbial decomposition of the plant tissue. Compounds that are emitted from flowers and fruits are often difficult to detect with conventional headspace extraction techniques that use solvents. This is a particular problem for researchers interested in fermentation volatiles because many of the emitted compounds (e.g. ethyl acetate, acetic acid, acetoin) overlap in their retention time with solvents that are commonly used for extracting these compounds such as hexane, acetone, or dichloromethane. In this thesis two commonly used extraction techniques were compared: (i) solid phase micro-extraction (SPME) and (ii) micro solid phase extraction (Micro SPE), both the techniques are used to collect fermentation and floral VOCs. For Micro SPE two different chromatography columns (DB5 and Carbowax) were used to determine which is more sensitive in identifying compounds.

For this, a floral and fermentation standard mixture was created by using compounds that represent sweet smelling flowers and rotting/fermenting fruits. Significant differences in the absolute emission of compounds, when using SPME and Micro SPE were found. There were also significant differences with the use of a DB5 and Carbowax column. The selection of the appropriate extraction technique for collection of VOCs should be based upon the type of application and availability of the necessary equipment. From this study, I found that Micro SPE worked better for collecting early to late stage VOCs of flowers and fruits, particularly if samples need to be collected under field conditions. To characterise the typical fermentation volatiles from flowers, the temporal variation in the emission of floral VOCs from the freshly opened to the wilted stages of three plant species namely: Hymenocallis littoralis, Dendrobium chrysotoxum and Camellia reticulata were investigated. The study revealed that there were significant differences in the absolute amounts and in some species also in the number of compounds emitted between early stage flowers and wilted flowers. Hymenocallis littoralis had a higher absolute emission of compounds on day one of flowering and thereafter emission decreased. However, no differences in the number of compounds were detected for this species. The VOCs of H. littoralis on the first day of sampling included: linalool, (1Z)-2-methylbutanal oxime, and 2-methyl-6-methylene-1,7-octadien-3-one, however the composition during the wilted stage included: (3E)-3-hexenyl acetate, heptenal, nonanal, and 2,6-dimethyl-7-octen-2-ol. There was a difference in the absolute emission of compounds for D. chrysotoxum, the emission increased until day six and thereafter decreased. There was also a difference in the number of compounds emitted, with more compounds being emitted from day 1 to 9 of sampling. The VOCs that contributed to the overall scent of the flowers of D. chrysotoxum from day 1 to 11 included: myrcene, linalool, limonene, 3,7-dimethyl-1,3,7-octatriene and α-pinene, however as the flowers began to wilt from day 12 to 13, the VOC composition changed and 4-dimethyl-1-heptene, limonene, terpinen-4-ol and (Z)-verbenol contributed to the late stage scent. For Camellia reticulata significant changes in the absolute emission and the number of compounds during the fresh and wilted flowering stages were detected. It was found that more compounds were emitted during the wilted stage. During the early stage of flowering, ethyl 2-(5-methyl-5-vinyltetrahydrofuran-2-yl) propan-2-yl carbonate, linalool, and 2H-pyran-3-ol, 6-ethenyltetrahydro-2,2,6-trimethyl- contributed to the scent of the flowers. During the wilted

stage, benzaldehyde, benzyl alcohol, (E)-2-Hexen-1-ol, and 2-Heptanol contributed to the scent of the flowers. To investigate typical fermentation volatiles three plant species, Musa acuminata (banana), Pyrus communis (pear) and Rothmannia globosa, were selected. For the three selected species, the temporal variation in the emission of VOCs from fruits during the ripening process (i.e. ripe to overripe stages) were investigated. Significant differences in the absolute amount and number of compounds emitted were detected. For M. acuminata, there was a higher absolute emission and number of compounds emitted from sampling day 1 to 10. Acetoin and 2,3-butanediol contributed to the scent of rotting M. acuminata. For P. communis, more absolute emission took place on day 1 of sampling and thereafter decreased, followed by an increase at day 35 of sampling and thereafter decreasing. Hexyl acetate, n-butyl butanoate, 1-hexanol and n-amyl acetate were found during the ripe stage of the fruit, and the VOC composition changed during rotting with the occurrence of ethyl acetate, acetic acid, 3 methyl-1-butanol, phenylethyl alcohol and benzaldehyde. Rothmannia globosa was sampled during the ripe and overripe stages. Higher absolute emission and number of compounds were emitted during the ripe stage of the fruit. 1-hexanol contributed to the scent of R. globosa during the ripe stages, however during the rotting stage, n-butyl acrylate, benzaldehyde, 1-butanol and 2-ethyl-1-hexyl acetate were found in samples. The findings of this study have relevance for researchers that are interested in the role of temporal VOC changes for the behaviour of pollinators and seed dispersers. Furthermore, it may be beneficial to researchers interested in the chemicals that are emitted by flowers that mimic rotting fruit volatiles. It is likely that microbial VOCs together with VOC changes initiated by the plant play an ecological role for host selection in animals. However, there is still a gap in our knowledge regarding the functional role of these VOCs and further studies are needed to investigate how such changes affect animal behavior and host selection. In addition, some of the findings of this study might be of interest for more applied areas such as horticulture and agriculture where the detection of microbial VOC signatures is important for detecting microbial pathogens, early senescence, and decomposition of plant tissue.