Table 1 Multiple comparisons of growth, drought indicators and physiological traits in interior lodgepole pine populations

From: Northern forest tree populations are physiologically maladapted to drought

Test

BAI

HI

Resis.

Recov.

Resil.

Rel. Resil.

Hydr. Diam.

Wall. Thick.

δ18O

δ13C

iWUE

LE–CI =0

 < 0.001

 < 0.001

 > 0.999

0.130

0.145

0.033

0.011

0.002

0.025

0.001

0.001

SI–CI = 0

0.481

0.999

0.968

0.999

0.995

 > 0.999

0.927

0.985

0.570

0.674

0.678

TE–CI = 0

0.108

0.104

0.016

0.004

 > 0.999

0.025

0.023

0.811

0.999

0.018

0.018

SI–LE = 0

<0.001

<0.001

0.983

0.176

0.087

0.048

 < 0.001

0.005

 < 0.001

0.038

0.036

TE–LE = 0

0.144

0.013

0.012

 < 0.001

0.156

 < 0.001

0.999

0.033

0.035

0.882

0.876

TE–SI = 0

0.001

0.082

0.004

0.003

0.995

0.019

0.002

0.948

0.522

0.241

0.239

  1. Multiple comparisons among tree populations representing the full north-south range of interior lodgepole pine (Pinus contorta ssp. latifolia). Populations are represented by five provenances grouped into four climatic regions, LE, CI, SI and TE: LE stands for the Leading Edge, i.e. the northern population occupying the area expected to be the leading edge of tree species migrations under climate warming; CI is Central Interior population, located in the central areas of the lodgepole pine range; SI is the Southern Interior population covering the southern range of the central areas of the lodgepole pine range; TE stands for the Trailing Edge, which represents seed sources from the far south of the lodgepole pine range, which is expected to see increased forest maladaptation under climate warming
  2. BAI is basal area increment, HI is height increment, Resis. is drought resistance, Recov. is drought recovery, Resil. is drought resilience, Rel. Resil. is drought relative resilience, Hydr. Diam. is mean hydraulic diameter, Wall. Thick. is mean tracheid wall thickness, and iWUE is intrinsic water-use efficiency derived from tree-ring δ13C. Significance (α = 0.05) is indicated in bold and p-values were adjusted with the Benjamini & Hochberg false discovery rate method