Arbuscular mycorrhizal fungi favor invasive Echinops sphaerocephalus when grown in competition with native Inula conyzae

Abstract

In a globalized world, plant invasions are common challenges for native ecosystems. Although a considerable number of invasive plants form arbuscular mycorrhizae, interactions between arbuscular mycorrhizal (AM) fungi and invasive and native plants are not well understood. In this study, we conducted a greenhouse experiment examining how AM fungi affect interactions of co-occurring plant species in the family Asteracea, invasive Echinops sphaerocephalus and native forb of central Europe Inula conyzae. The effects of initial soil disturbance, including the effect of intact or disturbed arbuscular mycorrhizal networks (CMNs), were examined. AM fungi supported the success of invasive E. sphaerocephalus in competition with native I. conyzae, regardless of the initial disturbance of CMNs. The presence of invasive E. sphaerocephalus decreased mycorrhizal colonization in I. conyzae, with a concomitant loss in mycorrhizal benefits. Our results confirm AM fungi represent one important mechanism of plant invasion for E. sphaerocephalus in semi-natural European grasslands.

Figure 1

Experimental design. This design was used for both harvests. Both mycorrhizal (M+) and nonmycorrhizal (M−) pots were pre-planted with a nurse plant Festuca pratensis. Soil was disturbed or left intact before target plants were planted, resulting in disturbed or non-disturbed arbuscular mycorrhizal (AM) networks in M+ treatment. Each figured pot contained 5 replicates.

Figure 2

Abundance of the different arbuscular mycorrhizal fungal taxa in the roots of the native plant species Inula conyzae as affected by community assembly i.e., plant combination (invasive-native: invasive and native plant growing in competition; native-native: only native plants of the same species growing together) or community assembly and soil disturbance (disturbed: substrate in the pot disturbed before target plants inserted; non-disturbed: target plants planted into non-disturbed substrate, resulting in non-disturbed CMNs of M+ pots) after the second harvest. Bars represent means accompanied by standard errors (n = 20 or n = 10, respectively). Different letters above individual bars indicate significant differences between means at P < 0.05.

Figure 3

Fraction of shoot dry biomass and fraction of shoot P content of native Inula conyzae (gray box), growing in pairs with invasive Echinops sphaerocephalus (white box) detected in the plant biomass per cultivation pot (i.e., the share of resources diverted to the native plant on a whole cultivation pot basis, with the remaining part of the particular resource being assignable to the invasive plant) as affected by mycorrhizal inoculation (M+: mycorrhizal inoculum added; M−: nonmycorrhizal control), initial soil disturbance (disturbed: substrate in the pot disturbed before target plants inserted; non-disturbed: target plants planted into non-disturbed substrate, resulting in non-disturbed CMNs of M+ pots), and harvest (harvest 1: first harvest; harvest 2: second harvest). Bars represent means accompanied by standard errors (n = 5). Different letters above individual bars indicate significant differences between means at P < 0.05.

Figure 4

Shoot dry biomass and shoot P content of the native plant species Inula conyzae following the second harvest, as affected by mycorrhizal inoculation (M+: mycorrhizal inoculum added; M−: nonmycorrhizal control), community assembly i.e., plant combination (invasive-native: invasive and native plant growing in competition; native-native: only native plants of the same species growing together) and initial soil disturbance (disturbed: substrate in the pot disturbed before target plants inserted; non-disturbed: target plants planted into non-disturbed substrate, resulting in non-disturbed CMNs of M+ pots) after the second harvest. Bars represent means accompanied by standard errors (n = 5). Different letters above individual bars within vertical dashed lines indicate significant differences between means at P < 0.05.