Corrigendum.

Abstract

New PhytologistVolume 224, Issue 4 p. 1698-1701 CorrigendumFree Access Corrigendum This article corrects the following: Logging and soil nutrients independently explain plant trait expression in tropical forests Sabine Both, Terhi Riutta, C. E. Timothy Paine, Dafydd M. O. Elias, R. S. Cruz, Annuar Jain, David Johnson, Ully H. Kritzler, Marianne Kuntz, Noreen Majalap-Lee, Nora Mielke, Milenka X. Montoya Pillco, Nicholas J. Ostle, Yit Arn Teh, Yadvinder Malhi, David F. R. P. Burslem, Volume 221Issue 4New Phytologist pages: 1853-1865 First Published online: September 20, 2018 First published: 20 September 2019 https://doi.org/10.1111/nph.16120 Author for correspondence: Sabine Both Tel: +61 6773 4308 Email: [email protected] AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL New Phytologist 221 (2019), 1853–1865, doi: 10.1111/nph.15444 Since its publication, the authors of Both et al. (2019) have brought to our attention errors in their article. Following reanalysis, the authors have determined that there are errors in some of the values presented in the text, in Figs 2 and 3, in Table 2 and in two items of Supporting Information (Tables S3 and S4). The corrections to the text, figures and tables have been assessed to have no impact on the overall conclusions of the work. The corrected text, figures and tables are shown below. We apologize to our readers for these mistakes. Open Research Data availability Data for Both et al. (2019) is available at doi: 10.5281/zenodo.3247602. [Correction added after online publication 20 September 2019: link to available data inserted.] References Baraloto C, Hérault B, Paine CET, Massot H, Blanc L, Bonal D, Molino J-F, Nicolini EA, Sabatier S. 2012. Contrasting taxonomic and functional responses of a tropical tree community to selective logging. Journal of Applied Ecology 49: 861– 870. Both S, Ruitta T, Paine CET, Elias DMO, Cruz RS, Jain A, Johnson D, Kritzler UH, Kuntz M, Majalap-Lee N et al. 2019. Logging and soil nutrients independently explain plant trait expression in tropical forests. New Phytologist 221: 1853– 1865. Wright IJ, Westoby M, Reich PB, Oleksyn J, Ackerly DD, Baruch Z, Bongers F, Cavender-Bares J, Chapin T, Cornelissen JH et al. 2004. The worldwide leaf economics spectrum. Nature 428: 821– 827. Corrected text, figures and tables Results, subsection ‘Community-weighted mean traits’ (p. 1857) should read: Major gradients in CWM trait expression were visualized by PCA, with the first two axes explaining 75.4% of the variance in functional traits (Fig. 2). There was a clear differentiation of functional composition between OG and SL plots along the first principal component, indicated by a distinct clustering of the study plots. Tree communities in OG plots were characterized by greater investment in defence and tissue density, whereas SL tree communities expressed higher photosynthetic activity and reduced investment into structural components (Table 2). Old-growth forests were characterized by denser wood and tougher leaves. These traits reflect enhanced structural investment, implying longer leaf life span and slower growth rates. Tree communities in SL forest had higher photosynthetic activity represented by higher CWM values of maximum photosynthetic rate, light-saturated photosynthetic rate, and dark respiration (Amax, Asat and Rd, respectively). These higher rates of gas exchange were supported by the expression of higher CWM area-based pigment concentrations in SL communities and higher Na and Nm concentrations. Tree communities in SL were enriched in 13C compared with OG communities, indicating greater water-use efficiency. The second axis of functional trait space represented tissue nutrient concentrations and leaf area, but was independent of logging history (Fig. 2). This axis reflects covariation among CWM values of leaf area, leaf Pm, Pa, Mgm and Cam concentrations, and a negative association of these traits with leaf Cm and tannin concentrations. Variability of these traits within both logged and unlogged forests was high, which suggests that the expression of these traits is driven by underlying soil properties rather than by logging history. Results, subsection ‘Variance partitioning’ (pp. 1857–1858) should read: Partitioning the community-level response of traits to logging and the first two principal components of soil properties showed that these factors explained up to 96% of the variation in traits. Overall, the proportion of variance explained was, on average, 73.4% (Fig. 3; Table S4). To present these results, we group the functional traits based on their main association with leaf nutrients, photosynthesis or structure. Variation in mass-based concentrations of leaf Cam, Pm and Km and, to a lesser extent, Mgm were associated with variation in soil properties, both with soil PC1 and 2. Particularly soil PC1, enveloping a gradient from exchangeable Mg to CEC and total P, strongly affected the variance in leaf Cam, Pm, Pa and Cm. For leaf traits related to photosynthesis, 3.1–88.2% of variance was explained by logging and a smaller proportion by soil PC1 (1.2–63.0%) or soil PC2 (0.9–48.8%; Table S4). Structural traits were explained by a combination of both logging history and the independent effects of soil properties. Logging explained, on average, 39.8% of variance in traits reflecting tissue density and structural investment, such as specific force to punch and branch wood density, which had consistently lower values in logged forest plots. Community-weighted mean LDMC was unusual in that it was poorly explained by all the predictor variables. By contrast, leaf size, expressed as CWM leaf area and leaf mass, increased with increasing values of soil PC1, which was linked toplots with higher total N and exchangeable Ca concentrations (Fig. 1). There was an increase in CWM tannin concentrations in logged forest plots and at lower values of soil PC1. Results, subsection ‘Functional diversity’ (pp. 1857–1858, first sentence) should read: Functional diversity, expressed as Rao's Q, did not differ between forest types (Fig. 4; F1,6 = 0.26, P = 0.63), and neither logging nor soil properties explained a significant proportion of its variance (Fig. 3). Fig. 2 and associated legend (p. 1858) should read: Fig. 2 Principal component analysis (PCA) of plot-level community-weighted mean functional traits. Plots cluster by logging history, with increased values of traits that maximize carbon capture and growth in logged forest communities and greater allocation to issue persistence and structural stability in old-growth forests. The highest loadings on the first axis are branch wood density (4.58%), Chlbm (4.58%), Chlba (4.52%), Specific force to punch (4.50%), Amax (4.36%) and Asat (4.35%). The highest loadings on the second axis are Pm (7.57%), Pa (7.44%), Tannins (6.45%), Cm (6.37%) and leaf area (5.88%). Mass-based nutrients are denoted by subscript ‘m’ and area-based values by subscript ‘a’. See Supporting Information Table S3 for all PCA loadings. Rd, dark respiration; LDMC, leaf dry matter content. Discussion, first paragraph, second sentence (p. 1858) should read: Logging was the primary driver of variation in CWM values of functional traits (Fig. 2), and explained more variation than the sum of the two soil PC axes for 18 of 32 traits (Fig. 3). Table 2 (p. 1859) should read: Table 2 Results from linear regression models from which the explained variance was generated; factors are the categorical ‘forest type’ (old growth (OG), selectively logged (SL)), and continuous ‘soil PC1’ and ‘soil PC2’. Functional trait Community weighted mean (CWM) trait value (and 95% CI) Forest type Soil PC1 Soil PC 2 Old-growth Selectively logged F-value P-value F-value P-value F-value P-value Rao's Q 0.794 (0.740–0.847) 0.789 (0.738–0.839) 0.327 0.742 ns 0.154 0.778 ns 3.542 0.306 ns d15N (‰) 1.640 (0.966–2.320) 0.846 (0.204–1.490) 5.032 0.232 ns 0.185 0.766 ns 0.527 0.689 ns Cam (mg g−1) 7.09 (6.12–8.20) 6.60 (5.75–7.59) 6.114 0.206 ns 17.555 0.152 ns 1.578 0.462 ns Mgm (mg g−1) 2.57 (2.06–3.20) 2.41 (1.96–2.97) 0.724 0.609 ns 1.275 0.506 ns 0.490 0.699 ns Km (%) 9.99 (8.4–11.90) 10.80 (9.13–12.7) 0.923 0.569 ns 0.321 0.742 ns 2.758 0.340 ns Nm (%) 1.83 (1.74–1.91) 1.97 (1.89–2.05) 9.191 0.165 ns 7.746 0.174 ns 4.523 0.251 ns Pm (mg g−1) 0.998 (0.912–1.090) 0.99 (0.909–1.08) 1.484 0.463 ns 13.219 0.152 ns 0.444 0.713 ns Cm (%) 44.6 (44.3–44.9) 44.7 (44.4–45) 12.266 0.152 ns 89.717 0.069 ns 0.240 0.748 ns d13C (‰) −32.4 (−32.1 to −32.8) −31.4 (−31.1 to −31.8) 28.775 0.096 ns 8.747 0.165 ns 11.356 0.154 ns Pa (mg mm−2) 8.71 × 10−5 (7.9 × 10−5–9.59 × 10−5) 8.65 × 10−5 (7.9 × 10−5–9.48 × 10−5) 0.842 0.581 ns 9.168 0.165 ns 1.067 0.557 ns Na (mg mm−2) 0.163 (0.156–0.170) 0.173 (0.166–0.180) 9.139 0.165 ns 3.251 0.314 ns 12.433 0.152 ns Rd (μmol CO2 m−2 s−1) −1.03 (−0.87 to −1.19) −1.25 (−1.09 to −1.40) 7.708 0.174 ns 1.546 0.462 ns 5.759 0.210 ns Amax (μmol CO2 m−2 s−1) 11.70 (8.94−14.50) 18.0 (15.4–20.6) 23.237 0.121 ns 0.878 0.577 ns 4.268 0.260 ns Asat (μmol CO2 m−2 s−1) 4.08 (2.66–5.50) 7.03 (5.69–8.38) 20.267 0.134 ns 0.284 0.742 ns 4.495 0.251 ns Carotenoidsa (mg mm−2) 5.96 × 10−5 (5.87 × 10−5–6.06 × 10−5) 5.81 × 10−5 (5.72 × 10−5–5.9 × 10−5) 9.036 0.165 ns 0.196 0.766 ns 5.963 0.207 ns Carotenoidsm (mg g−1) 0.687 (0.667–0.708) 0.667 (0.647–0.687) 6.353 0.206 ns 1.555 0.462 ns 0.264 0.742 ns Chlba (mg mm−2) 9.35 × 10−5 (9.06 × 10−5–9.64 × 10−5) 8.41 × 10−5 (8.14 × 10−5–8.68 × 10−5) 44.515 0.078 ns 0.971 0.563 ns 0.158 0.778 ns Chlbm (mg g−1) 1.09 (1.05–1.13) 0.97 (0.931–1.01) 40.379 0.078 ns 0.005 0.951 ns 2.605 0.346 ns Chlaa (mg mm−2) 0.000227 (0.000224–0.00023) 0.000217 (0.000214–0.00022) 41.759 0.078 ns 0.091 0.827 ns 11.876 0.152 ns Chlam (mg g−1) 2.62 (2.54–2.71) 2.49 (2.41–2.57) 13.853 0.152 ns 0.967 0.563 ns 0.020 0.922 ns Specific leaf area (mm2 mg−1) 12.0 (11.6–12.4) 11.8 (11.5–12.2) 2.822 0.340 ns 3.181 0.314 ns 3.571 0.306 ns Leaf area (mm2) 1.06 × 104 (9.03 × 104–1.24 × 104) 1.24 × 104 (1.07 × 104–1.44 × 104) 1.879 0.428 ns 12.976 0.152 ns 5.323 0.226 ns Leaf dry weight (mg) 922 (784–1.08 × 103) 1.08 × 103 (930–1.27 × 103) 2.077 0.401 ns 10.600 0.163 ns 6.120 0.206 ns Leaf thickness (mm) 0.221 (0.209–0.233) 0.236 (0.224–0.249) 7.196 0.182 ns 0.213 0.761 ns 2.319 0.371 ns Specific force to punch (N mm−2) 1.23 (1.03–1.47) 0.889 (0.751–1.050) 13.209 0.152 ns 0.431 0.713 ns 0.013 0.934 ns Force to punch (N mm−1) 0.266 (0.227–0.311) 0.212 (0.183–0.247) 7.611 0.174 ns 0.300 0.742 ns 0.260 0.742 ns LDMC (mg g−1) 416 (391–440) 410 (387–433) 0.143 0.780 ns 0.332 0.742 ns 0.079 0.827 ns Branch density (g cm−3) 0.564 (0.528–0.599) 0.493 (0.460–0.526) 15.309 0.152 ns 2.442 0.361 ns 1.522 0.462 ns Phenolm (mg g−1) 36.4 (33.7–39.1) 42.7 (40.2–45.3) 29.960 0.096 ns 5.004 0.232 ns 2.762 0.340 ns Tanninm (mg g−1) 8.56 (7.70–9.41) 9.41 (8.6–10.2) 8.888 0.165 ns 8.069 0.174 ns 0.004 0.951 ns Lignin and recalcitrants (%) 19.4 (17.3–21.4) 17.3 (15.3–19.2) 2.707 0.340 ns 3.333 0.312 ns 0.080 0.827 ns Cellulose (%) 22.5 (20.9–24.0) 20.9 (19.4–22.4) 3.343 0.312 ns 0.313 0.742 ns 0.995 0.563 ns Hemicellulose (%) 12.3 (11.4–13.1) 11.8 (110–12.6) 1.728 0.450 ns 0.767 0.600 ns 0.270 0.742 ns a For analyses, values of dark respiration Rd fluxes and δ13C were converted to positive values for ease of interpretation; untransformed values are shown here. For abbreviations and description of the functional traits, see Supporting Information Table S1. ns, not significant; LDMC, leaf dry matter content. Fig. 3 and associated legend (p. 1860) should read: Fig. 3 Proportion of variance in community-weighted mean functional trait values explained by forest type and the first two principal components (PC) of soil properties (Fig. 1). Functional traits are grouped by the ecosystem function to which they most contribute. Statistical significance is derived from linear regression models following false discovery rate correction, asterisks indicate P < 0.05, and ‘+’ and ‘−’ indicate the direction of the relationship. For forest type, ‘+’ indicates that trait values were greater in selectively logged than in old-growth forests (i.e. positive with first PC axis in Fig. 2). For variance explained by soil, ‘+’ indicates a positive relationship with the respective PC axis. See Table 2 and Supporting Information Table S4 for detailed results. Discussion, second paragraph, third sentence (p. 1861) should read: Thus, species in logged forest communities expressed higher CWM values of area-based N, photosynthetic activity and lamina thickness, whereas old-growth forest communities expressed low CWM values of these traits and higher values of traits conferring structural stability and resistance to herbivory, such as branch wood density and leaf toughness. Discussion, third paragraph, first sentence (p. 1861) should read: We observed marginally lower CWM values of SLA in logged forests, in contrast to results from French Guiana (Baraloto et al., 2012), and contrary to the expectation that SLA scales positively with Amax, foliar Na, and foliar Pa concentrations among species (Wright et al., 2004). Discussion, fifth paragraph, third sentence (p. 1861) should read: The second component of soil variation also influenced some leaf traits, particularly δ13C and Na concentrations (Tables 2, S4). Supporting Information Tables S3 & S4: Corrections have been made to Table S3 (Loadings of the community-weighted mean traits in the principal component analysis) and Table S4 (Results from linear regression models underlying the variance partitioning), see Supporting Information in Both et al. (2019). Volume224, Issue4December 2019Pages 1698-1701 ReferencesRelatedInformation