Root traits of New Zealand trees: Community-level responses to a soil fertility gradient, and plastic responses to nutrient availability

Abstract

The study of plant traits has great application for understanding plant distribution patterns and community assembly at a variety of scales. Roots are a vital component of plant water and nutrient uptake strategy, and yet root traits are not as well understood as leaf or stem traits. In this thesis I aimed to determine the relationships between seedling fine root, leaf, and stem traits of New Zealand tree species, and investigate whether these traits related to species‟ abundance along a soil fertility gradient. I also investigated how seedling traits compare with adult traits, and whether nutrient availability or method of fertiliser application affected seedling trait expression. To achieve these goals, I measured root, leaf, and stem traits on seedlings of 66 native tree species, and combined this with relative abundance data along a strong soil fertility gradient at Puketī forest. I used principle component analysis and ordination to assess the dimensionality of trait variation across species; and linear regressions to compare community-weighted mean (CWM) traits vs. a summarised axis of soil fertility. I also performed regression analysis between seedling traits and adult traits from the literature. To assess the effects of nutrient availability and application I grew four native tree species under three nutrient treatments: low, pulsed, and slow release. I used two-way ANOVA and co-efficient of variation analyses to determine the strength of responses to nutrient treatments. Dry matter content was positively associated across leaves, roots, and stems, and negatively associated with root nitrogen concentration and relative growth rate, suggesting that at least as seedlings, traits associated with a fast or slow growth strategy are co-ordinated across organs. Root diameter and SRL were independent from this axis of fast-slow growth, suggesting that SRL does not have a direct effect on seedling growth rates and nutrient foraging. A third axis of variation was also identified, strongly influenced by root phosphorus and nitrogen concentrations, but was difficult to interpret. Regression of CWM traits vs. soil fertility showed that traits which comprised the fast-slow PCA axis co-varied strongly with soil fertility. Neither root diameter nor SRL were significantly related to soil fertility, supporting the conclusion that SRL is not adaptive to nutrient foraging ability. Seedling morphology traits are generally well correlated with adults, but tend to be oriented towards a more “acquisitive” growth strategy, suggesting that species may down-regulate their growth over ontogenetic development. The absence of correlation between SRL and soil fertility suggests that alternative root traits may be more applicable for understanding species foraging strategy. Species‟ responses to the nutrient treatments differed for most traits. Root-to-shoot ratio, growth rate, and root nutrient concentrations responded strongly to nutrient availability, and morphological traits did not respond as strongly. Slow release and pulse treatments were typically similar, and both significantly different to low nutrient treatment. These results suggest that it is practical to compare morphological trait data between studies, provided plants are raised in environments conducive to growth, but that nutrient concentrations and biomass allocation traits can be strongly influenced by soil fertility.