Erratum for Blonder et al. 2017. Predicting trait-environment relationships for venation networks along an Andes-Amazon elevation gradient. Ecology 98:1239–1255. https://doi.org/10.1002/ecy.1747

The authors report it has come to their attention that values of vein density (VD) used in this study are overestimated by a constant factor of two. The problem arose due to an error in MATLAB image processing code related to estimating skeleton perimeters. Corrected image analysis code has been provided on Zenodo at: http://doi.org/10.5281/zenodo.3961105.

The original VD values and the revised VD values are scaled by a factor of 2, as visualized in the below figure (blue line, fitted to data; red line, 1:1 expectation).

Thus, any values of VD reported in the main text figures or results should be divided by two. The authors report this error does not affect any of the other vein traits reported in the manuscript or any of the paper's conclusions. None of the paper’s statistical analyses depend on the absolute scale of the vein density measurements. The one analysis which does depend on the absolute scale of variables (the formal test of the process-based model in Figure 8) is not parameterized with VD.

The following denotes how these changes affect the manuscript, with revised text, Figs. 3, 4, 6, and 7, Table 1, and supporting information provided. The only change to these materials is the absolute scale of the VD values.

Main text

In Results: Range of trait variation, the sentence “Vein density (VD) varied from 4.8 mm⁻¹ for Clusia alata (Clusiaceae) to 21.7 mm⁻¹ for Pourouma bicolor (Urticaceae)…” should read “(Species-at-site mean) vein density (VD) varied from 2.5 mm⁻¹ for Clusia alata (Clusiaceae) to 11.3 mm⁻¹ for Pourouma bicolor (Urticaceae)…”. The slight differences in values are due to rounding.

Appendix S1

This file contains supporting figures and tables. It has been updated to reflect corrected VD values for Table S1, Figure S1, Figure S6, and Figure S7.

Data S1

This file contains branch-level vein trait measurements. It has been updated with corrected VD values.

Data S2

This file contains species-at-site vein trait summaries. It has been updated with corrected VD values.

Figure 2 caption

The values of VD should be divided by two, such that the caption that originally read “…from species such as (A) Pourouma bicolor (Urticaceeae) with VD = 22.8 mm⁻¹ and (B) Clusia alata (Clusiaceae), with mean VD = 4.0 mm⁻¹…” instead reads, “…from species such as (A) Pourouma bicolor (Urticaceae) with VD = 11.8 mm⁻¹ and (B) Clusia alata (Clusiaceae), with mean VD = 2.1 mm⁻¹...”. The slight differences in values are due to rounding. Note that these values differ from the main-text corrected VD values described above, as the values in the figure caption are for individual leaves, whereas the values in the text are for species-at-site means (in both the original publication and the erratum).

Figure 3

Figure 4

Figure 6

Figure 7

Table 1

Blonder 等人的勘误表。2017. 预测沿安第斯山脉-亚马逊海拔梯度脉络网络的特征-环境关系。生态学 98：1239-1255。https://doi.org/10.1002/ecy.1747

作者报告说，他们注意到本研究中使用的静脉密度 (VD) 值被高估了一个常数因子 2。该问题是由于与估计骨架周长相关的 MATLAB 图像处理代码中的错误引起的。Zenodo 上提供了更正的图像分析代码：http://doi.org/10.5281/zenodo.3961105。

原始 VD 值和修改后的 VD 值按 2 倍缩放，如下图所示（蓝线，拟合数据；红线，1:1 期望值）。

因此，正文图表或结果中报告的任何 VD 值都应除以 2。作者报告的此错误不会影响手稿中报告的任何其他静脉特征或论文的任何结论。该论文的统计分析均不依赖于静脉密度测量的绝对规模。一种依赖于变量绝对规模的分析（图 8 中基于过程的模型的正式测试）没有使用 VD 进行参数化。

以下表示这些变化如何影响手稿，修改后的文字，图。3、4、6 和 7，表 1，以及提供的支持信息。这些材料的唯一变化是 VD 值的绝对比例。

正文

在结果：范围性状变化，句子“静脉密度（VD）从4.8毫米变化^-1为Clusia翅（藤黄科）至21.7毫米^-1为Pourouma双色（荨麻科）...”应改为“（种-在现场平均值）静脉从2.5毫米变化密度（VD）^-1为Clusia翅（藤黄科）至11.3毫米^-1为Pourouma双色（荨麻科）...”。数值的细微差异是由于四舍五入。

附录 S1

该文件包含支持数字和表格。它已更新以反映表 S1、图 S1、图 S6 和图 S7 的修正 VD 值。

数据 S1

该文件包含分支级静脉特征测量值。它已更新为更正的 VD 值。

数据 S2

该文件包含物种现场静脉特征摘要。它已更新为更正的 VD 值。

图 2 标题

VD 的值应除以 2，使得最初读作“……来自诸如 (A) Pourouma bicolor ( Urticaceeae ) VD = 22.8 mm ^-1和 (B) Clusia alata (Clusiaceae)等物种，平均VD = 4.0 mm ^-1 …”改为“…来自物种，例如 (A) Pourouma bicolor ( Urticaceae )，VD = 11.8 mm ^-1和 (B) Clusia alata (Clusiaceae)，平均 VD = 2.1 mm ^-1……”。数值的细微差异是由于四舍五入。请注意，这些值与上述正文修正的 VD 值不同，因为图标题中的值是针对单个叶子的，而文本中的值是针对现场物种的平均值（在原始出版物和勘误）。

图 3

图 4

图 6

图 7

表格1