Leaf-level light compensation points in shade-tolerant woody seedlings


Photosynthetic traits such as respiration rate and light compensation point (LCP) likely play an important role in determining a plant's tolerance to low light levels, which can dictate long-term partitioning of the light supply among species and successional patterns (Givnish, 1988; Pacala et al., 1994). Walters & Reich (1999) reviewed patterns of leaf-level LCPs and associated leaf parameters in seedlings of tree species. They reported differences in mass-based dark respiration rates (Rd,mass) and specific leaf area (SLA), but no differences in area-based dark respiration rates (Rd,area), quantum yield (QY) or LCP among species that are considered shade-tolerant, shade-intolerant or intermediate in their tolerance of shade. Although other aspects of plant ecophysiology are certainly important for shade tolerance, is it true that leaf LCPs are unimportant?

Shade tolerance: questions

There are, in fact, several reasons to question the truth of this assertion. First, for a given plant, shade leaves are considered to have lower LCPs than sun leaves (Ellsworth & Reich, 1993). If phenotypic and genotypic plasticity operate in the same direction, then shade-tolerant plants would have lower LCPs compared with shade-intolerant species. Second, shade-tolerant species are thought to intercept a higher fraction of the incident light, i.e. species that are tolerant of shade cast deeper shade (Horn, 1971; Canham et al., 1994). For this pattern to exist, the leaves of shade-tolerant plants at the bottom of the canopy would have to have lower LCPs than leaves of shade-intolerant species because mature leaves likely need to be energetically self-sufficient (Lacointe et al., 2004; Sprugel et al., 1991). Third, individual studies that compare shade-intolerant and shade-tolerant species under similar growth conditions often show that shade-tolerant species have lower LCPs than shade-intolerant species (e.g. Lusk, 2002). Finally, data for Rd,area and LCPs were log-normally distributed, and should have been log-transformed to meet the requirements of anova (Sokal & Rohlf, 1994).

Patterns of photosynthetic traits

In order to investigate this, we started with data on Rd,area, QY and LCP from appendix 1 of Walters & Reich (1999), which included the leaves of 67 species of broadleaf evergreen trees and 27 species of temperate deciduous trees that were grown under low light (< 4% full sun) and/or slightly higher light levels (4–12% full sun). Using the same criteria as Walters & Reich, we obtained data on photosynthetic characteristics of seedlings of woody species published since 1998 (see Appendix). These included 22 new species from six publications (Einig et al., 1999; Valladares et al., 2000; Hattenschwiler, 2001; Lusk & Del Pozo, 2002; Lusk, 2002; Muraoka et al., 2003) for a total data set of 115 species. Following Walters & Reich, no needleleaf evergreens were included.

Species means were recalculated for each of the two light levels (< 4% and 4–12% full sun). Also following Walters & Reich, species were categorized with respect to shade tolerance based on published survival data in shade by original authors and/or previously published observations of tolerance. Species with ambiguous classification (e.g. Acer pseudoplatanus of Hattenschwiler, 2001) were not included in this analysis. Data from Einig et al. (1999) were taken from leaves positioned at intermediate height on the seedlings. Hattenschwiler (2001) data were averages of means at 1.3 and 3.4% full sun for ambient CO2. Only shade plants were used from Lusk (2002), because it was uncertain whether the high-light category exceeded 12% full sun. For Muraoka et al. (2003), we averaged data for seedlings in deciduous–coniferous and deciduous forests. We removed one strong outlier from the analysis. Fraxinus americana, an intermediate shade-tolerance species that was reported to have an LCP of 1.1 mol m−2 s−1 (Bazzaz & Carlson, 1982), but a calculated LCP (Rd,area/QY) of 11.3. All analyses were performed in jmp 5.0.1 (SAS, Cary, NC).

When examining the relationships among photosynthetic characteristics, species with lower area-based dark respiration rates and/or higher QY had lower LCPs. Both Rd,area and QY were significant predictors of LCP (log LCP = 1.73 + 0.84 log Rd,area – 11.25 QY; r2 = 0.61, P < 0.001 for each coefficient) with significant bivariate relationships using a Model II regression (Fig. 1).

Figure 1.

Bivariate relationships among Rd,area, QY and LCP. Lines represent Model II regression best fit. LCP and Rd,area were log-transformed before regressions.

Differences in Rd,area, QY and LCP among categorical factors that include species classification (broadleaf evergreen and deciduous), light levels and shade-tolerance were tested using two separate full-factorial anovas. Because the light levels, shade-tolerance categories and leaf habit contrasts represent post hoc categories, nonsignificant predictors (P > 0.05) were removed serially.

LCPs were lower for shade-tolerant species than shade-intolerant species and species intermediate in their tolerance for shade (P < 0.001, Table 1). Shade-tolerant species have positive carbon balance at light levels approximately 3 mol m−2 s−1 less than intermediate or intolerant species (least squares means LSMs = 5.3 vs 8.1, 9.0 mol m−2 s−1, respectively). Combining the intolerant and intermediate species into a single category strengthens the significance of difference between tolerant and other species for Rd,area and LCP (Table 1). The lower LCPs of shade-tolerant species compared with intermediate and intolerant species was associated with lower Rd,area (P = 0.06 for three categories, P = 0.01 with two), but no difference in quantum yield (Table 1).

Table 1.  Backwards elimination anova results for species-averaged Rd,area, QY and LCP
 log Rd,areaQYlog LCP
  1. Note that the table also includes P-values for categorical contrasts (light level and evergreen vs deciduous leaf habit) and least squares means (LSM) for shade-tolerance categories. Superscript letters refer to Tukey's HSD comparisons (P = 0.05) among LSMs (bold text) of shade-tolerance categories for a given factor. Units for Rd,area and LCP are mol m−2 s−1. Neither the interaction between leaf habit and shade tolerance nor the three-way interactions was significant in either model.

Habit 0.08 
Shade tolerance0.06 < 0.001
 Shade tolerant0.25a   5.32a
 Intermediate0.33a   8.10b
 Shade intolerant0.35a   9.04b
Light*Habit 0.04 
Habit 0.08 
Shade tolerance0.01 < 0.001
 Shade tolerant0.25a   5.32a
 Other0.34b   8.61b
Light*Habit 0.04 

Relevance of lower LCPs

There are three major consequences to shade-tolerant species having lower values of Rd,area and LCP. First, net carbon balance in low light should be greater for tolerant than for less-tolerant species, and hence contribute (along with whole-plant characteristics) to the success of tolerant species in shade. Light levels in the understory of forests can be near the LCPs of seedlings for the greater part of the day. Ellsworth & Reich (1992) measured 10-min means of photosynthetic photon flux density over the course of a growing season in the understory of an Acer saccharum forest. Over 90% of the time light levels were below 20 mol m−2 s−1. The low LCP and Rd,area of shade-tolerant species should allow them both to conserve carbon better than less tolerant species and to gain carbon better, having net photosynthesis at lower light levels.

Second, the trade-off between low respiration rates and high maximum photosynthetic rates among species was thought to reflect differential selection for conservation of carbon in shaded habitats vs the ability to photosynthesize at high rates in high-light environments (e.g. Givnish, 1988). With shade-tolerant species having lower LCPs, this trade-off should be extended such that plants can either have leaves that photosynthesize at low light levels or photosynthesize at high rates at high light levels. Hence, there is an additional dimension to the evolutionary trade-off between success in shade and sun habitats. Third, shade-tolerant trees should be able to reduce light levels to lower levels than shade-intolerant species by building canopies with a greater leaf area index that intercept a greater fraction of total incoming radiation. Combined with differences in carbon balance at low light, this could serve as a mechanism by which shade-tolerant species can prevent the establishment of seedlings of shade-intolerant species.


Table 2.  Presented below are data published on photosynthetic characteristics since Walters & Reich (1999). Categorical data include leaf habit (deciduous or evergreen), the shade tolerance of the species and the light levels under which the seedlings were grown (low (< 4% full sun) or medium (4–12% full sun). All species were from temperate habitats. Reported data include Rd,area (mol m−2 s−1), QY (unitless) and LCP (mol m−2 s−1).
Leaf habitShade toleranceLightGenusSpeciesRd,areaQYLCPStudy
DeciduousTolerantMediumAcerdistylum0.190.0685 2.5Muraoka et al. (2003)
DeciduousIntolerantMediumAcerrufinerve0.220.064 2.5Muraoka et al. (2003)
EvergreenTolerantLowAextoxiconpunctatum0.19  3.7Lusk (2002)
EvergreenTolerantMediumAextoxiconpunctatum0.24  7Lusk and Del Pozo (2002)
EvergreenTolerantMediumAmomyrtusluma0.45  8Lusk and Del Pozo (2002)
EvergreenIntermediateLowAristoteliachilensis0.44  6.7Lusk (2002)
EvergreenIntolerantMediumAristoteliachilensis0.73  9Lusk and Del Pozo (2002)
EvergreenIntolerantMediumAuracariaangustifolia1.52 38Einig et al. (1999)
EvergreenIntolerantMediumCaldcluviapaniculata0.61  9Lusk and Del Pozo (2002)
EvergreenTolerantMediumDasyphyllumdiacanthoides0.55  8Lusk and Del Pozo (2002)
EvergreenIntermediateMediumDrimyswinteri0.64  9Lusk and Del Pozo (2002)
EvergreenIntolerantLowEmbothriumcoccineum0.57  6.9Lusk (2002)
EvergreenIntolerantLowEucryphiacordifolia0.42  6.3Lusk (2002)
EvergreenIntolerantMediumEucryphiacordifolia0.68 11Lusk and Del Pozo (2002)
EvergreenTolerantLowGevuinaavellana0.49  6.3Lusk (2002)
EvergreenTolerantLowLaureliaphilippiana0.37  3.8Lusk (2002)
EvergreenTolerantMediumLaureliaphilippiana0.56  7Lusk and Del Pozo (2002)
EvergreenTolerantLowLumaapiculata0.38  5.3Lusk (2002)
EvergreenTolerantLowMyrceugeniaplanipes0.4  3.5Lusk (2002)
EvergreenTolerantMediumMyrceugeniaplanipes0.57  6Lusk and Del Pozo (2002)
EvergreenIntolerantMediumNothofagusdombeyi0.95 15Lusk and Del Pozo (2002)
EvergreenIntolerantMediumNothofagusnitida0.96 12Lusk and Del Pozo (2002)
EvergreenIntolerantLowQuercuscoccifera0.550.0139Valladares et al. (2000)
DeciduousTolerantMediumQuercuscrispula0.310.0745 2.5Muraoka et al. (2003)
EvergreenIntolerantLowQuercusilex0.670.0231Valladares et al. (2000)
DeciduousIntermediateLowQuercusrobur0.28  4.7Hattenschwiler (2001)
EvergreenIntolerantMediumWeinmanniatrichosperma0.63 10Lusk and Del Pozo (2002)