Fungal endophyte effects on leaf chemistry alter the invitro growth rates of leaf-cutting ants' fungal mutualist, Leucocoprinus gongylophorus

Fungal symbionts that colonize leaf tissue asymptomatically (endophytes) can alter the foraging behaviour of leaf-cutting ants, and decrease the productivity of this herbivore’s mutualistic fungal cultivar, Leucocoprinus gongylophorus. This negative effect of endophytes on the ant’s cultivar could be the result of direct fungalefungal interaction or indirect reductions in the quality of leaves, the cultivar’s growth substratum. To test for the indirect effects, we measured in vitro growth rates of cultivars in media that contained sterilized leaf extracts from plants with high (Ehigh) and low (Elow) endophyte colonization. We found that, opposite to our expectations, cultivars grew significantly faster in Ehigh leaf extracts compared to Elow extracts. Our results suggest that endophyte-driven changes in leaf chemistry are a less likely explanation for the observed in vivo reduction in the ant’s symbiotic fungal growth and imply that the effect of direct endophyteecultivar interactions inside nests are potentially more important. a 2014 Elsevier Ltd and The British Mycological Society. All rights reserved. Introduction ecological role of most endophytes has not yet been examined Plant-associated microbial communities (plant microbiomes) have a substantial influence on a plant’s ecological interactions and their responses to abiotic factors (Rodriguez et al., 2009; Porras-Alfaro and Bayman, 2011). An important part of this microbiome is a biologically and taxonomically hyperdiverse group of fungal symbionts that live asymptomatically inside plant tissues (hereafter ‘endophytes’ Wilson 1995). The [email protected] (C. E ier Ltd and The British M but studies so far suggest their effect on plant’s ecological interactions is variable and range from negative to positive (Saikkonen et al., 2010). Certain plant-endophyte symbioses have been associated with reductions in infections by pathogenic fungi and with declines in herbivores’ survival, developmental rates, fecundity and host acceptance (Webber, 1981; Clark et al., 1989; McGee, 2002; Jallow et al., 2004; Mej ıa et al., 2008; Van Bael et al., 2009; Crawford et al., 2010; Bittleston strada). ycological Society. All rights reserved. 38 C. Estrada et al. et al., 2011; Gange et al., 2012; Estrada et al., 2013). The protective benefits these fungi may confer on plants have motivated a large body of research focused on describing endophyte communities in a wide range of plants and in searching for species that target particular plant diseases (Suryanarayanan and Johnson, 2005; Arnold and Lutzoni, 2007; Helander et al., 2007; Mejia et al., 2009; Hanada et al., 2010; Higgins et al., 2011). However, in most cases the mechanisms underlying the effects that plant-endophyte interactions have on plant’s natural enemies remain unclear, impeding both our understanding of the importance of endophytes in ecological networks and our evaluation of the potential to use them in biological control programs (Herre et al., 2007; Rodriguez Estrada et al., 2012). Leaf endophytes have been associated with host preferences by leaf-cutting ants (Formicidae: Myrmicinae; Atta), a major herbivore in natural and agricultural ecosystems, primarily in the Neotropics (Van Bael et al., 2011). These ants can harvest more than 2% of leaf biomass in forest and savannas, play a strong role in shaping the species composition of plant communities, and are responsible for losses of up to a billion dollars yr 1 in agriculture (Abreu and Delabie, 1986; H€ olldobler andWilson, 1990; Wirth et al., 1997; Terborgh et al., 2001; Herz et al., 2007; Costa et al., 2008). Previous studies have shown that ants take longer to cut leaves with high relative to low loads of endophytes (Van Bael et al., 2009). In choice experiments when both types of leaves were present, recruitment of workers was higher to leaves with lower endophyte densities and these leaves were typically cut and removed faster than leaves with higher loads of the fungal symbionts (Bittleston et al., 2011; Estrada et al., 2013). The number of ants recruited to a resource indicates the quality of food (Jaffe and Howse, 1979) suggesting that endophyte colonization induces changes in leaves that lower their value to ants. Ant-cut leaf pieces are carried to subterranean nests and used to feed an obligate mutualistic fungal cultivar, Leucocoprinus (1⁄4Leucoagaricus) gongylophorus (Lepiotaceae, Basidiomycota) that ants and larvae use as their main source of nutrition. The fungal cultivar transforms leaf tissue into food for the ants that is delivered in specialized swollen hyphal tips called gongylidia (Cherrett et al., 1989; Aylward et al., 2012). The presence of fungal endophytes in their foragematerial reduces the leaf-cutting ant’s garden productivity in young laboratory colonies (Van Bael et al., 2012a). It is still unclear what mechanisms underlie this reduction or how those relate to the observed ant preferences for leaves with low loads of endophytes. The processing of leaf fragments by the ants greatly reduces the amount of alien microorganisms that enter the garden (Quinlan and Cherrett, 1977; Van Bael et al., 2009; Urriola et al., 2011). However, independently of the potential threat of a direct and detrimental interaction between the persisting endophytes and the ant’s mutualistic fungi, endophyte symbioseswith plants could indirectly affect garden productivity by reducing leaf quality. Endophyte colonization could result in leaf tissues that for the cultivar have lower nutritional value, less favourable pH or a higher content of antifungal metabolites than those from endophyte-free plants. This could be expected because the foliar content of some nutrients and minerals changes as a result of particular endophyte-plant symbioses (Rodriguez Costa Pinto et al., 2000; Gange et al., 2012; Estrada et al., 2013; but see Van Bael et al., 2012a). Furthermore, leaf chemistry also changes in ways detectable to ants (Estrada et al., 2013), which could be due to antifungal compounds expressed by plants in response to endophytes or toxins from these fungi that are constitutive or induced by plant defence responses (Schulz et al., 1999; White and Torres, 2010). We conducted experiments to test the hypothesis that endophyte-driven changes in leaf extracts affect the leafcutting ants’ fungal garden productivity. Our experiments removed the effects of leaf-cutting ant hygiene to observe in vitro growth responses of the fungal cultivar, L. gongylophorus. The ants’ cultivar growth rate was compared in sterilized media containing extracts from leaves harvested from plants hosting low or high densities of endophytes. Since the sterilization process killed the endophytes in the leaves, our experiments focused on indirect rather than direct effects of the plant-endophyte symbiosis. Specifically, we tested for endophyte-mediated changes in leaf quality rather than a direct fungalefungal interaction between endophytes and the ants’ cultivar. We expected that the fungal cultivar would grow better in leaf extracts from plants with lower levels of endophyte colonization.We further assessed the growth rates of two leaf fungal symbionts in the same media to examine the extent to which patterns observed for the ants’ cultivar extend to other fungi. Materials and methods