1. The root economics spectrum (RES) hypothesis predicts that fast-growing tree species have short-lived roots with high specific root length (SRL) to allow rapid resource uptake, and opposite trait expressions for slow-growing species. Yet, the mixed support for this hypothesis suggests that trees can adopt alternative strategies to increase resource uptake, besides an increase in SRL. 2. We combined a novel mechanistic whole-tree model and empirical fine-root data of ten tree species to test the effects of one of these alternative strategies, notably increasing fine-root mass, on the tree’s net C gain (used here as a proxy for tree performance), and to assess how fine-root lifespan influences the relative importance of SRL and fine-root mass for the C balance of trees. 3. Our results indicate that accounting for the short lifespan of high-SRL roots has important implications for explaining tree performance and the role of roots herein. Without considering their faster turnover, high-SRL roots and low fine-root mass resulted in the highest performance as predicted from the RES. Yet, when their higher turnover rates were accounted for, a high fine-root mass and low SRL lead to the highest performance. Both our model outcomes and field data further show a negative relationship between SRL and fine-root mass through which species aim to realise a similar root length density. This trade-off further indicates how high a SRL and low fine-root mass as well as opposite trait values can both lead to a positive C balance in a similar environment. 4. Our study may explain why high-SRL roots do not necessarily lead to the fastest tree growth as often hypothesised and demonstrates the importance of fine-root mass in combination with fine-root lifespan for explaining interspecific differences in tree performance. More generally, our work demonstrates the value of identifying and investigating different belowground strategies across species from a whole-plant modelling perspective, and identifies the relationship between SRL, fine-root biomass and lifespan as an important functional dimension to variation in species’ performance.

Fine-root data were collected from ten temperate broadleaved tree species (Table 1) growing in a common garden forest at the Hollandse Hout, the Netherlands (52°28’N, 5°26’E), established in the 1960s and 1970s. The forest is subjected to a temperate marine climate with mild winters and summers and an annual rainfall of approximately 800 mm (KNMI, 2009). The soil is described as a calcareous and nutrient-rich clay soil with a homogeneous texture throughout the forest, and the groundwater table ranges between 1.6 and 1.2 m below the soil surface (Weemstra et al., 2013).

Fine-root mass and morphology were measured in three monospecific forest plots per species. Per plot, three soil samples up to 1 m depth and randomly distributed over the plot were collected using a 42 mm-diameter root auger, the roots washed out, separated into fine and coarse roots based on their diameter (fine roots being < 2 mm diameter) and weighed. As roots were sampled with a root auger and cut into smaller fragments, we were unable to accurately sort roots based on their order. Fine roots were scanned, the scans were analysed with WinRhizo (Regent Instruments, Quebec, Canada) to determine total fine-root length, mean diameter and mean root volume, and the fine roots were oven-dried at 60° C for 48h to determine dry mass. Fine-root SRL was calculated by dividing fine-root length by dry weight, fine-root tissue density was calculated by dividing fine-root dry weight by fine-root volume, and fine-root length density (fine-root length per soil volume) was calculated by multiplying SRL and fine-root mass.