The relationship between root system size and length distribution
We showed for the first time that a trade-off between what can be described as a “phosphate scavenging” root phenotype and root system size exists in strawberry. A “phosphate scavenging” phenotype can be defined by a high root length distribution whereby a greater proportion of horizontal roots are seen at the soil surface, as seen in bean (29). Indeed, studies comparing root architectural trends across plant species, found seedlings of plants from nutrient rich regions had larger roots systems (33). Similarly, a trade-off has been observed in Senecto vulgaris roots whereby large root systems have high exploitation but low exploitation efficiency (34). Indeed, a maize root systems can be characterised as small and compact or large and exploratory with trade-offs between exploration and occupation resulting from phenotypic trade-offs rather than biomass restriction (35). As no phenotypic or genotypic correlation was observed between root diameter nor root length distribution and root size traits we hypothesise that these traits are under unique genetic control.
Root system and AMF association
Meta-analysis studies looking into the presence of a universal link between root architecture and mycorrhizal association have found mixed results. A negative relationship has been observed between mycorrhizal dependency (the positive impact of AMF) and total root length across different plant species (36). Furthermore, research looking into the degree to which AMF alter host root systems have found greater root modification potential in plants with the highest mycorrhizal dependency (37). By contrast, another study across multiple species found a coarser root architecture does not lead to a greater mycorrhizal colonisation nor growth benefit from association (38). Here, a higher solidity was genetically correlated with a greater frequency of arbuscule and vesicle AMF structures. A high solidity can be described by “dense roots growing together which thoroughly explore the root network area” (39). Similar studies in rice have found a negative relationship between thoroughness of the root system (solidity) and the extent of the root system (convex area) (39), where by a dense root system can be said to explore a smaller area of soil and thus stand to benefit more from exploiting an AMF extraradicular network. However, such a conclusion cannot be drawn here as we see no relationship between solidity and convex area across our genotypes.
Benefits of AMF colonisation
Greater dry matter and phosphate uptake has been observed in strawberry plants colonised with mycorrhiza under low phosphate conditions (16). Similarly, we saw a positive correlation between above ground dry biomass and AMF vesicle formation amongst genotypes (Figure 6). Further than this, application of AMF to strawberry cultivar ‘Elsanta’ was found to have a greatly improved yield in coir production (13). Unfortunately, no positive relationship was observed between strawberry yield and AMF association across ‘Redgauntlet’ x ‘Hapil’ genotypes grown in terragreen (Figure 6) indicating that AMF strawberry interactions are complex and that any associated benefits may be both substrate and genotype specific.
Parental root architecture and propagation method
We found greater resolution to discriminate differential root architecture in the two parental cultivars when root systems were developed through ‘pinning down’ plantlets as opposed to misted tip propagation. Pinned down plantlets were able to establish a root system whilst attached to the mother plant and thus had access to parental nutrients. It is known that root architecture can be altered in response to soil nutrient status (28,40). Therefore, the misted tip propagation method may have led to an altered root system architecture through the absence of parental nutrients. However, low internal phosphate status was not found to trigger the phosphate starved morphological changes observed in low phosphate external environments (41). Research investigating the performance and nutrient use efficiency of plants developed using different propagation strategies may have downstream implications for best practice nursery propagation. In contrast to the findings above, parental average root diameter was found to vary consistently in both experiments and we also observe that average root diameter appears to be controlled independently from the root traits representing size (Figure 6).
Hormonal control of root architecture
The ethylene and auxin pathway have been identified as key regulators of root branching, root hair elongation and low phosphate detection (42–44). Two of the identified QTL associated with increased root system size and low phosphate tolerance, co-localised with core genes in the auxin pathway of vesca diploid strawberry. The latter is particularly notable as it represents a candidate gene for breeders to enhancing PUE in strawberries. Further research is required to validate the PUE marker across the wider strawberry germplasm.
Heritability of traits and genetic correlations
Heritability scores for all investigated traits were low (5 - 14.5 %; Table 3), indicating a large environmental component controlling variation in the assessed traits. Arbuscules have a heritability of 0%. The low Arbuscule heritability is an artefact of low arbuscule occurrence across 100 transects, with 0, 1 or 2 arbuscules found in each root system. The mean values for arbuscules provided a more quantitative score for QTL assessment. In spite of the low heritability scores, 45 QTL controlling traits were identified with each explaining between 5.9 and 13.4 % of the observed phenotypic variation. Low genetic correlation between root traits can be explained by the highly plastic nature of roots which can alter in response to multiple abiotic factors including nutrient concentration, water availability, oxygen content, soil density and pH (3).
QTL co-localisation
We observed a network of pleiotropic QTL controlling multiple root size traits in strawberry indicating several regions of the genome are important for root architecture. Similar studies have identified regions of the genome associated with multiple root architecture in both rice and maize (35,39).
Co-localisation of arbuscule and vesicle association QTL on chromosome 6B provides strong evidence that this region is involved in AMF association. The locus was associated with 9.1 and 8.1 % effect size, respectively. The QTL on chromosome 7B associated with vesicle formation co-localised with the Verticillium dahliae disease resistance in the ‘Chandler’ strawberry cultivar (45). The focal SNPs representing Verticillium resistance and vesicle formation on chromosome 7B fall at the same genetic position on the ‘Redgauntlet’ x ‘Hapil’ linkage map (0 cM apart) and 19 kb apart on the strawberry consensus physical map. This observation leads us to hypothesise that the same genetic component is involved in controlling both fungal interaction phenotypes. Both SNPs fall within 100 kb of a Receptor Like Kinase gene containing a legume lectin domain based on vesca diploid gene models. Common pathways exist to promote plant entry, indeed symbiosis receptor like kinase have been reported to be important in both AMF and nitrogen fixing recognition (46).
Phosphate tolerance and root architecture
No association between low phosphate tolerance and length distribution was observed in this study. Indeed, no phenotypic correlation was observed between all root architectural traits and low P-tolerance. However, roots have been shown to exhibit phenotypic plasticity in response to low phosphate environments, whereby phosphate starved plants show reduced primary root growth, reduced growth rate due to lower phosphate metabolism and increased lateral branching (30,47). In this experiment root architecture was measured under optimal growth conditions. Further work should determine whether ‘Redgauntlet’ and ‘Hapil’ root architecture is altered under low phosphate conditions and whether specific low-P root traits are associated with enhanced P-tolerance. Furthermore, we can determine whether root plasticity itself is associated with an increase in low P-tolerance in strawberries.