Strategies in Ericaceae to acquire phosphorus in phosphorus-impoverished habitats in the southwest Australian biodiversity hotspot.

We hypothesised that some Ericaceae exhibit high leaf manganese (Mn) concentrations [Mn], a proxy for rhizosphere carboxylates, and release root carboxylates. We compared their leaf [Mn] with that of positive and negative reference species, known to release carboxylates or not, respectively. To follow up, we measured the carboxylate-exudation rates of targeted species with high and low leaf [Mn] using seedlings grown in low-P nutrient solutions. Using these complementary approaches, we confirmed that Ericaceae in P-impoverished habitats with high leaf [Mn] exhibit a carboxylate-releasing P-mobilising strategy, like non-mycorrhizal Proteaceae. Surprisingly, some species with low leaf [Mn], which occurred in habitats with high soil pH, also released carboxylates. Therefore, low leaf [Mn] cannot conclusively indicate the absence of carboxylate exudation. These species may release carboxylates along with cations such as potassium or magnesium, which increase the rhizosphere pH, thereby decreasing Mn availability and accumulation in mature leaves. The lack of a significant phylogenetic signal detected for leaf [P] and leaf [Mn] across sampled taxa indicated these nutrient-acquisition traits are not limited to certain clades but likely evolved independently multiple times in Ericaceae. Styphelia sensu lato exhibited the widest trait variation (highest to lowest leaf [P] and [Mn]) of all genera included in this study.

Methods
To test our hypothesis that some Ericaceae exhibit high leaf manganese concentrations [Mn], a proxy for rhizosphere carboxylates, and release root carboxylates, we collected fully expanded mature leaves of 32 Ericaceae in southwest Australia from October 2020 to July 2022. We also collected leaves of positive reference species (Proteaceae) and negative reference species (Xanthorrhoeaceae) at most sample locations. When there was no Xanthorrhoea species nearby, we used young expanding leaves of the target species as a negative reference. Total mature leaves [Mn] (Figure 2) and [P] (Figure 3) were oven-dried leaf samples and ground to a fine powder. The acid was digested and analysed by ICP-OES.

Leaf [Mn] was compared using the Welch t-test to assess the difference in mean concentrations between the negative reference and target species. Differences in leaf [Mn] and [P] across species were analysed using one-way ANOVA, and followed by Tukey-Kramer’s HSD post-hoc tests at P < 0.05.

To measure root carboxylate exudation, we grew seedlings of Ericaceae in a controlled-environment glasshouse at the University of Western Australia at 18/25℃ (night/day) temperatures. After approximately 20 weeks in the nutrient solution, root exudates were collected from excised parts of root systems. Total effective carboxylate is calculated by summing up the concentrations the sum of all di- and tri-carboxylates, with mainly citrate, malate, and oxalate. All carboxylate concentrations are expressed per unit root fresh weight. Our results showed that all the selected plant species, which exhibited either relatively high or low leaf [Mn], all released carboxylates when grown in low-P nutrient solution (Figure 5).