Limits...
Symbiosis limits establishment of legumes outside their native range at a global scale

View Article: PubMed Central - PubMed

ABSTRACT

Microbial symbiosis is integral to plant growth and reproduction, but its contribution to global patterns of plant distribution is unknown. Legumes (Fabaceae) are a diverse and widely distributed plant family largely dependent on symbiosis with nitrogen-fixing rhizobia, which are acquired from soil after germination. This dependency is predicted to limit establishment in new geographic areas, owing to a disruption of compatible host-symbiont associations. Here we compare non-native establishment patterns of symbiotic and non-symbiotic legumes across over 3,500 species, covering multiple independent gains and losses of rhizobial symbiosis. We find that symbiotic legume species have spread to fewer non-native regions compared to non-symbiotic legumes, providing strong support for the hypothesis that lack of suitable symbionts or environmental conditions required for effective nitrogen-fixation are driving these global introduction patterns. These results highlight the importance of mutualisms in predicting non-native species establishment and the potential impacts of microbial biogeography on global plant distributions.

No MeSH data available.


The interaction between symbiosis and human uses.The size of the symbiosis effect on successful introductions at different numbers of human uses, using the full dataset. Dots represent the difference in the predicted number of non-native ranges between non-symbiotic and symbiotic species, standardised by dividing by the predicted number of non-native ranges for non-symbiotic species. The standardisation controls for the differences in the predicted number of non-native ranges between different human use levels (determined by the main effect of number of human uses), and makes the interaction easier to visualize. Error bars represent the 95% confidence interval obtained through parametric bootstrapping. The size of the dots is proportional to the number of species in the dataset with that number of human uses, showing that most species had no uses recorded in the dataset (∼70%). The dotted vertical line represents the mean number of human uses across all species in the dataset (0.77). The x-axis is on a square root scale. Negative values mean the symbiotic species are predicted to have a lower prevalence of non-native ranges relative to non-symbiotic species. Apparent non-linearity is due to logit back-transformation.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC5385628&req=5

f4: The interaction between symbiosis and human uses.The size of the symbiosis effect on successful introductions at different numbers of human uses, using the full dataset. Dots represent the difference in the predicted number of non-native ranges between non-symbiotic and symbiotic species, standardised by dividing by the predicted number of non-native ranges for non-symbiotic species. The standardisation controls for the differences in the predicted number of non-native ranges between different human use levels (determined by the main effect of number of human uses), and makes the interaction easier to visualize. Error bars represent the 95% confidence interval obtained through parametric bootstrapping. The size of the dots is proportional to the number of species in the dataset with that number of human uses, showing that most species had no uses recorded in the dataset (∼70%). The dotted vertical line represents the mean number of human uses across all species in the dataset (0.77). The x-axis is on a square root scale. Negative values mean the symbiotic species are predicted to have a lower prevalence of non-native ranges relative to non-symbiotic species. Apparent non-linearity is due to logit back-transformation.

Mentions: We found that ∼30% of species in our dataset had at least one human use, and that legume species with more uses are much more likely to establish in non-native regions (Table 1). Species with human uses may be more likely to establish due to more frequent intentional introduction attempts (i.e., higher human-mediated propagule pressure), which may mask or confound differential establishment patterns driven by the symbiosis trait itself. Our analysis accounts for this potential bias by including it as a covariate (along with its interaction with symbiosis), and then statistically evaluating the main effect of symbiosis at no (that is, zero) human uses (this is important because of the presence of the interaction16). The main effect of symbiosis (Table 1) thus evaluates any differences in non-native establishment patterns that are least likely to be impacted by human-mediated propagule pressure. After accounting for the number of human uses in this manner, we found that non-symbiotic legumes are still much more likely to establish in non-native regions (Figs 2 and 4; Table 1).


Symbiosis limits establishment of legumes outside their native range at a global scale
The interaction between symbiosis and human uses.The size of the symbiosis effect on successful introductions at different numbers of human uses, using the full dataset. Dots represent the difference in the predicted number of non-native ranges between non-symbiotic and symbiotic species, standardised by dividing by the predicted number of non-native ranges for non-symbiotic species. The standardisation controls for the differences in the predicted number of non-native ranges between different human use levels (determined by the main effect of number of human uses), and makes the interaction easier to visualize. Error bars represent the 95% confidence interval obtained through parametric bootstrapping. The size of the dots is proportional to the number of species in the dataset with that number of human uses, showing that most species had no uses recorded in the dataset (∼70%). The dotted vertical line represents the mean number of human uses across all species in the dataset (0.77). The x-axis is on a square root scale. Negative values mean the symbiotic species are predicted to have a lower prevalence of non-native ranges relative to non-symbiotic species. Apparent non-linearity is due to logit back-transformation.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC5385628&req=5

f4: The interaction between symbiosis and human uses.The size of the symbiosis effect on successful introductions at different numbers of human uses, using the full dataset. Dots represent the difference in the predicted number of non-native ranges between non-symbiotic and symbiotic species, standardised by dividing by the predicted number of non-native ranges for non-symbiotic species. The standardisation controls for the differences in the predicted number of non-native ranges between different human use levels (determined by the main effect of number of human uses), and makes the interaction easier to visualize. Error bars represent the 95% confidence interval obtained through parametric bootstrapping. The size of the dots is proportional to the number of species in the dataset with that number of human uses, showing that most species had no uses recorded in the dataset (∼70%). The dotted vertical line represents the mean number of human uses across all species in the dataset (0.77). The x-axis is on a square root scale. Negative values mean the symbiotic species are predicted to have a lower prevalence of non-native ranges relative to non-symbiotic species. Apparent non-linearity is due to logit back-transformation.
Mentions: We found that ∼30% of species in our dataset had at least one human use, and that legume species with more uses are much more likely to establish in non-native regions (Table 1). Species with human uses may be more likely to establish due to more frequent intentional introduction attempts (i.e., higher human-mediated propagule pressure), which may mask or confound differential establishment patterns driven by the symbiosis trait itself. Our analysis accounts for this potential bias by including it as a covariate (along with its interaction with symbiosis), and then statistically evaluating the main effect of symbiosis at no (that is, zero) human uses (this is important because of the presence of the interaction16). The main effect of symbiosis (Table 1) thus evaluates any differences in non-native establishment patterns that are least likely to be impacted by human-mediated propagule pressure. After accounting for the number of human uses in this manner, we found that non-symbiotic legumes are still much more likely to establish in non-native regions (Figs 2 and 4; Table 1).

View Article: PubMed Central - PubMed

ABSTRACT

Microbial symbiosis is integral to plant growth and reproduction, but its contribution to global patterns of plant distribution is unknown. Legumes (Fabaceae) are a diverse and widely distributed plant family largely dependent on symbiosis with nitrogen-fixing rhizobia, which are acquired from soil after germination. This dependency is predicted to limit establishment in new geographic areas, owing to a disruption of compatible host-symbiont associations. Here we compare non-native establishment patterns of symbiotic and non-symbiotic legumes across over 3,500 species, covering multiple independent gains and losses of rhizobial symbiosis. We find that symbiotic legume species have spread to fewer non-native regions compared to non-symbiotic legumes, providing strong support for the hypothesis that lack of suitable symbionts or environmental conditions required for effective nitrogen-fixation are driving these global introduction patterns. These results highlight the importance of mutualisms in predicting non-native species establishment and the potential impacts of microbial biogeography on global plant distributions.

No MeSH data available.