Research Article: Statistical modelling of conidial discharge of entomophthoralean fungi using a newly discovered Pandora species

Date Published: May 22, 2019

Publisher: Public Library of Science

Author(s): Niels Lundtorp Olsen, Pascal Herren, Bo Markussen, Annette Bruun Jensen, Jørgen Eilenberg, Philippe Silar.


Entomophthoralean fungi are insect pathogenic fungi and are characterized by their active discharge of infective conidia that infect insects. Our aim was to study the effects of temperature on the discharge and to characterize the variation in the associated temporal pattern of a newly discovered Pandora species with focus on peak location and shape of the discharge. Mycelia were incubated at various temperatures in darkness, and conidial discharge was measured over time. We used a novel modification of a statistical model (pavpop), that simultaneously estimates phase and amplitude effects, into a setting of generalized linear models. This model is used to test hypotheses of peak location and discharge of conidia. The statistical analysis showed that high temperature leads to an early and fast decreasing peak, whereas there were no significant differences in total number of discharged conidia. Using the proposed model we also quantified the biological variation in the timing of the peak location at a fixed temperature.

Partial Text

Fungi are important as biological control agents, and their effect is due to their infective spores [1]. The mechanisms for spore releases differ among fungal taxons, one mechanism is shooting off spores as found in fungus order Entomophthorales, which are insect and mite pathogens. Conidia are the infective units of entomophthoralean fungi, and for the majority of species they are actively discharged [2]. The large conidia (mostly between 15 and 40 microns in length) in Entomophthorales demands high energy to be discharged. The spore discharge mechanism for entomophthoralean fungi allow fungi to convert elastic energy into kinetic energy, ensuring that spores are discharged at sufficient speeds.

Predicted mean trajectories for u, evaluated at observed time points, along with population means are displayed in Fig 4. We observe a slightly odd behaviour around t = 0. This is an artifice; most observations around t = 0 are zero. When the predicted values of u are exp-transformed, these are mapped into almost-zero values. In concordance with our hypothesis, the three population means are clearly separated and fit well into what we expected: θwarm peaks first and has the highest peak; θmiddle is in-between and θcold peaks latest and has a smaller and more slowly decreasing peak.

With the applied statistical methods, we were able to characterize the temporal patterns of conidial discharge to a much better degree than previous studies, and we characterized the variation between individual fungi at the same temperature (i.e. of the same population). With a 95% prediction interval of roughly 6 hours, the temporal variation is too little for changing the overall shapes, but still large enough to be important for the analysis and to shift the peaks for individual fungi significantly.




Leave a Reply

Your email address will not be published.