Plant Ecology

, Volume 214, Issue 5, pp 787–798 | Cite as

Cold tolerance of photosynthesis as a determinant of tree species regeneration patterns in an evergreen temperate forest

  • Sarah J. Richardson
  • Karen I. Bonner
  • Christopher P. Bickford


Niche partitioning of light among seedling species is a key mechanism supporting coexistence in forests. Species sort along light gradients through direct responses to light and through indirect responses mediated by other environmental factors. Canopy gaps in temperate evergreen rainforests experience sub-zero temperatures and thus gap-dependent species are vulnerable to cold photoinhibition from exposure to high light at low temperatures. We used a shadehouse experiment to test two hypotheses: (1) that gap-dependent species are resistant to cold photoinhibition; and (2) that gap-dependence observed in the field may be driven by the interaction between high light and low temperatures. Specifically, we predicted that some species restricted to shade in the field are excluded from gaps because of low resistance to cold photoinhibition. Gap dependence of angiosperm and conifer seedlings was estimated from expert opinion, and from experimental growth and survival responses to light treatments representing a forest understorey and forest gap. Additional seedlings were used to evaluate resistance to cold photoinhibition (sub-zero temperatures at dawn). Gap-dependent species were resistant to cold photoinhibition. Our second hypothesis was supported by Beilschmiedia tawa (Lauraceae), which had low resistance to photoinhibition, a strong positive growth response to the light treatments, and is restricted to shade in the field. Seedling regeneration niches in temperate rainforest are shaped in part by the interaction between light and low temperatures, and this interaction will be crucial for determining seedling responses to climate warming.


Chlorophyll fluorescence Climate change New Zealand Plant functional trait Regeneration niche Shade tolerance 


Light is a limiting resource for tree seedlings in forest ecosystems. Variation among species in their capacity to tolerate low light and respond to high light separates species along gradients of light availability and drives regeneration dynamics and community assembly in forest ecosystems (Canham et al. 1990; Grubb et al. 1996; Valladares and Niinemets 2008). Comparative shade tolerance among species can be inferred from seedling distributions in forest understoreys (e.g. Houter and Pons 2012) and is commonly assessed experimentally using seedling survival and growth in low versus high light (e.g. Kobe et al. 1995). Gap-dependent species are usually defined by low survival in shade and fast growth in high light (Valladares et al. 2000). This trade-off between survival in low light and rapid growth in high light forms the basis for partitioning light gradients into niches among seedling species and for explaining coexistence among woody species (Grubb et al. 1996; Kobe et al. 1995). However, niche partitioning of light is contingent on other limiting resources such as water availability in low-light microsites (Niinemets and Valladares 2006; Sanchez-Gomez et al. 2006; but see Coomes and Grubb 2000), other abiotic stresses associated with canopy gaps and forest margins, and exposure to biotic stresses such as disease and herbivory. Sub-zero temperatures are a significant climate driver of seedling regeneration in temperate forest gaps (Inouye 2000). Forest canopies reduce the incidence of sub-zero temperatures at seedling height, relative to open conditions (Ball et al. 1991), while gaps are exposed to sub-zero temperatures, particularly on gently sloping landforms such as terraces that lack free-air drainage (Fig. 1; Wardle 1985). In cool-temperate, high-latitude and high-altitude forests there is a cost to growing in gap microsites: species that cannot persist in deep shade must possess traits that confer resistance to cold photoinhibition (e.g. Ball et al. 1991; Reyes-Díaz et al. 2009). This interaction between temperature and shade is significant for temperate evergreen species that retain foliage through winter when cold photoinhibition occurs, particularly during spring when new tissues emerge. Temperate evergreen forests are widespread in the Southern Hemisphere especially in Tasmania (Feild and Brodribb 2001), Chile, Argentina and New Zealand, where the deciduous strategy is rare (McGlone et al. 2004). The influence of sub-zero temperatures and cold photoinhibition on shaping the shade tolerance niche among temperate tree species is poorly understood (Robakowski 2005) relative to interactions between shade and water availability (Coomes and Grubb 2000) yet will be critical for determining regeneration dynamics in the future under warmer climates. The frequency of sub-zero temperature events is forecast to decline in temperate regions over the next 50 years (Inouye 2000; MfE 2008) and responses by ecological communities to climate warming will hinge on interactions between low temperatures and other limiting resources.
Fig. 1

Minimum air temperatures during winter daylight hours (June–August inclusive, 8 am to 4 pm) at 15 cm above the ground, inside and outside forest canopies at three topographic positions in evergreen rainforest, New Zealand. a The difference in daily minimum temperature inside and outside a forest canopy, with zero difference shown as a dashed line. (Box plots boxes define the 1st and 3rd quartiles; horizontal lines within boxes are medians; whiskers are the minimum and maximum values). b The number of days with a sub-zero minimum temperature inside (open) and outside (filled) a forest canopy

The damaging influence of sub-zero temperatures on seedlings varies according to whether they coincide with low or high light (Augspurger 2011). Sub-zero temperatures can physically damage leaf tissues, rupturing cells and causing solute leakage (Sakai et al. 1981). High light at sub-zero temperatures imposes extra stress through photoinhibition (Ball et al. 1991; Cunningham and Read 2006). Comparative measurements of chlorophyll fluorescence as an indicator of photoinhibition are appropriate for studying the interaction between light and low temperatures as they explicitly link the damage to photosystems by high light at low temperatures and provide a mechanistic measure of cold tolerance (Ball et al. 1991; Öquist et al. 1987). Repeated damage to photosystems results in cell death and leaf loss (Ball et al. 1991), which reduces seedling growth rates and in severe situations, survival rates (Bader et al. 2007; Coop and Givnish 2008). We test the hypotheses that gap-dependent species are resistant to cold photoinhibition and that niche partitioning of light gradients observed in the field can be driven by the interaction between high light and low temperatures. We predict that some species restricted to shaded microsites in field conditions are minimising exposure to cold photoinhibition rather than preferentially selecting low light. That is, we predict that some species with an apparent preference for shade under field conditions may grow significantly faster under high light in the absence of sub-zero temperatures. Exclusion from large gaps because of cold sensitivity has been proposed for Beilschmiedia tawa (Lauraceae), a tree of subtropical origin from New Zealand that is considered a shade-tolerant or shade-preferring species, and sensitive to low temperatures (Knowles and Beveridge 1982; West 1995).

Conifers are often regarded as more cold tolerant than co-occurring evergreen angiosperms because of their physically-resistant needle-like leaf structures and their vasculature. However, a comparison of photosynthetic cold tolerance (fluorescence) between conifers and angiosperms in Tasmanian alpine heath found no difference between the two groups (Feild and Brodribb 2001) because a vessel-less angiosperm species in the Winteraceae had comparably high cold tolerance to co-occurring conifer species. Conifers and angiosperms co-occur widely in New Zealand’s temperate rainforests and many studies have characterised the functional traits that separate these two groups of species (e.g., Holdaway et al. 2011) and how trait differences underpin niche partitioning between conifers and angiosperms along environmental gradients (e.g., Enright and Ogden 1995; Lusk et al. 2009; McGlone et al. 2010). In this study we compare photosynthetic resistance to cold photoinhibition in co-occurring evergreen conifer and angiosperm seedling species ranging in their gap dependence. Gap-dependent evergreen species typically have lower leaf mass per unit area (LMA; Lusk et al. 2008), smaller leaves (Niinemets and Kull 1994) and are more plastic to varying light availability (Valladares et al. 2000), relative to shade-tolerant species. Greater plasticity may be selected for in gap-dependent species because gaps typically close over and plasticity is adaptive in changing light environments (Valladares et al. 2000). We test whether resistance to cold photoinhibition is associated with traits of gap-dependence.

Predicting future forest composition under warmer climates requires a mechanistic understanding of how functional traits underpin seedling performance and niche partitioning at the regeneration phase across multiple, interacting environmental gradients. Little work has addressed the impact of cold-induced photoinhibition in cool-temperate evergreen rainforests where infrequent low temperature events may have large and enduring impacts on seedling establishment (Kelly 1987; Ball et al. 1991). We test two hypotheses to explore how tree seedlings from a cool-temperate rainforest sort according to gap-dependence and cold tolerance: (1) Gap-dependent species will be resistant to cold photoinhibition as this is an essential prerequisite of exploiting gaps in cool-temperate environments. These species will have small leaves and low LMA, as predicted for gap-dependent species (Niinemets and Kull 1994; Lusk et al. 2008); and (2) Niche partitioning of light gradients observed in the field may be driven by the interaction between high light and low temperatures. Specifically, some species restricted to shaded microsites in field conditions are minimising exposure to cold photoinhibition rather than preferentially selecting low light. These species will grow significantly faster in high light than low light under experimental conditions.

We compare cold-tolerance and gap-dependence in co-occurring conifer and angiosperm seedling species from a cool-temperate rainforest community and include a non-native conifer species, Pseudotsuga menziesii (Douglas-fir), that is partially shade tolerant (Bond et al. 1999; Broncano et al. 2005) and predicted to invade forest understoreys in New Zealand (Dehlin et al. 2008). In so doing, we evaluate how cold tolerance and gap dependence define the observed regeneration niches of co-occurring rainforest tree species.

Materials and methods

Study species

We selected seven common, native tree species (4 angiosperms and 3 conifers) from temperate rainforests of central New Zealand (Carswell et al. 2007) and a non-native conifer (P. menziesii). Functional traits defining these species are provided in Table 1. Seedlings for experimental measurements of cold tolerance and gap dependence were either raised from seed (germinated at ~40 % PAR) collected near Ngapūtahi in central New Zealand (38°38′ S, 176°53′ E) or sourced from nurseries close to the seed collection site.
Table 1

Summary of the eight tree species tested for seedling cold tolerance and gap dependence




Adult height (m)

Seed mass (mg)

Leaf size (mm2)

LMA (g m−2)


 Beilschmiedia tawa





1,002 (99.7)

86.4 (4.2)

 Nestegis cunninghamii





1,171 (63.8)

134.9 (2.1)

 Melicytus ramiflorus





1,276 (92.8)

37.6 (2.0)

 Weinmannia racemosa





794 (55.9)

72.1 (4.7)


 Prumnopitys ferruginea





38 (2.4)

69.9 (4.5)

 Dacrycarpus dacrydioides





5 (0.2)

61.0 (2.7)

 Pseudotsuga menziesiia





26 (2.9)

70.5 (3.3)

 Podocarpus totara





55 (3.4)

85.0 (3.8)

Canopy heights of mature adults were taken from McGlone et al. (2010). Seed mass was measured on at least three individuals from the Allan Herbarium, Lincoln, New Zealand. Seedling leaf size and leaf mass per unit area (LMA) were measured on five replicate individuals at the end of the experiment on seedlings grown at 5.7 % PAR (photosynthetically active radiation) following Cornelissen et al. (2003). Data are means with 1 SE

aSpecies not native to New Zealand

Estimating gap dependence

Gap dependence was estimated using (i) expert opinion, and (ii) experimental growth and survival responses by seedlings to two light treatments in a glasshouse. The expert opinion method uses five ordinal categories to describe the typical light environment of seedling occurrence in field situations: (1) mostly under a closed canopy and occasionally in small gaps, (2) mostly in small gaps and under a closed canopy, (3) mostly in gaps of different sizes and occasionally under a closed canopy, (4) exclusively in gaps of different sizes and not under a closed canopy of undisturbed forest, and (5) typically in large gaps and clearings (Houter and Pons 2012). Our study species were scored by one of the authors (SJR) and four ecologists with at least a decade of experience working in similar forests (R.B Allen, P.J. Bellingham, F.E. Carswell and M.C. Smale).

We used a randomised block design to determine the responses of our seedling species to ‘low’ (1.2 % photosynthetically active radiation, PAR) and ‘high’ (5.7 % PAR) light. The two PAR treatments were selected from the range of understorey light conditions measured in rainforests around Ngapūtahi. We used an objective survey of light values in forest understoreys (Carswell et al. 2012) to select our low- and high-light treatments: the minimum value was taken for the low-light treatment (1.2 % PAR) and the median value was taken for the high-light treatment (5.7 % PAR). We used five replicate blocks each with a pair of shade frames. Light treatments were randomly allocated to frames within blocks. Inside each frame we had three individuals of each species, giving us a total of 30 individuals of each species. Species were randomly allocated to numbered 2-L pots within each frame. We used knitted monofilament shade cloth (Egmont Commercial, New Zealand) over wooden frames and adjusted the density of shade cloth to achieve our desired PAR treatments using measurements of PAR inside each frame relativised to measurements from outside. PAR was measured using a Hobo data logger and S-LIA-M003 PAR Smart Sensors (Hobo Data Loggers, OnSet, MA, USA). PAR data from the middle of the day (11 am to 3 pm inclusive) during September 2010 (the austral spring) was used to calculate the proportion of PAR reaching inside the frames. Fully illuminated mean PAR outside during this period was 1,302 and 15.7 μmol m−2 s−1 in the low-light frames (equivalent to 1.2 % PAR) and 70.5 μmol m−2 s−1 in the high-light frames (equivalent to 5.7 % PAR). The experiment was conducted in a glasshouse protected from natural sub-zero temperature events so as to estimate the responses to PAR by species independent of their responses to sub-zero temperatures. Seedlings were grown in a homogenised sandy pumice soil collected near Ngapūtahi (mean total P of 78 mg kg−1; mean total N of 0.21 %; Richardson et al. 2008). Initial seedling biomass and height were estimated from 10 individuals per species in October 2009 when seedlings were approximately 10 months old. Height was measured as pulled seedling height from the root collar to the terminal bud. Seedlings were harvested and separated into root, stem and leaf components. Roots were washed and each component dried at 60 °C until constant mass, then weighed. Final height and biomass data were collected using the same approach in June 2011. The duration of the experiment was 632 days (30th September 2009 to 23rd June 2011) and seedlings experienced nearly two full years starting in spring.

Measuring cold tolerance using chlorophyll fluorescence

Chlorophyll fluorescence measures the amount of light that is successfully dissipated away from photosystems when other resources for photosynthesis are limiting (e.g. at low temperatures). Cold-tolerant plants quickly dissipate excess light and rapidly recover once conditions are favourable for photosynthesis. Cold-intolerant plants dissipate less excess light and take longer to recover pre-exposure photosynthetic rates. We exposed seedlings of our eight species to low temperatures in high-light conditions to measure their capacity to dissipate excess light. Our temperature treatments were selected based on a histogram of daily temperatures in the hour following dawn, measured using TinyTag data loggers (Gemini Data Loggers, Chichester, UK) at 15 cm above the ground (seedling height), inside and outside forest canopies near Ngapūtahi (Fig. 1).

Mature seedlings (2-years-old) were transferred from a shaded glasshouse with low radiation (mean PAR between 1 pm and 2 pm, September 2010 of 398 μmol m−2 s−1 relative to outside, fully sunlit mean PAR of 1,270 μmol m−2 s−1) to a controlled environment growth cabinet 48 h prior to initial measurements. Cabinet conditions were 70 % relative humidity, and PAR was modulated near 77 μmol m−2 s−1—equivalent to the mean PAR in the 5.7 % PAR treatment—by placing shade cloth above the plants, and photoperiod was set to a 14 h day and 10 h night, with 1 h dawn and dusk periods where PAR was reduced by 50 % relative to full PAR. Growth chamber PAR was set to low levels reflective of PAR observed in forest margins and interior at Ngapūtahi. Air temperature was initially 15:9 °C (day:night), and baseline measurements of dark-adapted chlorophyll fluorescence were collected from all plants at this temperature regime, using a portable photosynthesis system (LI-6400XT, LiCor Biosciences, Lincoln, NE, USA) with leaf chamber fluorometer (6400–40). Leaves were covered with dark-adapting clips (LiCor 9964–091) for ≥30 min before measurements. After collecting baseline measurements, seedlings were cold-acclimatised (Krause 1994) by decreasing temperature daily using the following step-wise regime: 14:8 °C, 13:7 °C, 12:6 °C, 11:5 °C, 10:4 °C, 9:3 °C, 8:2 °C, and 7:−2 °C. The night-time freezing temperature excursion was extended into the dawn period and temperature was increased to 0 °C for 1 h at the full PAR intensity before increasing to 7 °C, when dark-adapted chlorophyll fluorescence was again assessed, as described previously. We expressed this second measure of chlorophyll fluorescence (Fv/Fm) relative to the baseline measure and used the proportional change to indicate how well species dissipated excess light at cold temperatures—a measure of cold tolerance (Cunningham and Read 2006).

Data analyses

Absolute height and biomass growth rates (AHGR and ABGR, respectively) expressed the total gain in height or mass relative to the duration of the glasshouse experiment in years. Relative height and biomass growth rates (RHGR and RBGR, respectively; George and Bazzaz 1999) were calculated as:
$$ {\text{RHGR }}\left( {{\text{mm mm}}^{ - 1} \,{\text{yr}}^{ - 1} } \right) \, = \left[ {{ \log }_{\text{e}} \left( {{\text{final height}},{\text{ mm}}} \right) \, - \,{ \log }_{\text{e}} \left( {{\text{initial height}},{\text{ mm}}} \right)} \right] / {\text{time in years}}\, $$
$$ {\text{RBGR }}\left( {{\text{mg mm}}^{ - 1}\, {\text{yr}}^{ - 1} } \right) \, = \left[ {{ \log }_{\text{e}} \left( {{\text{final biomass}},{\text{ mg}}} \right) \, - \,{ \log }_{\text{e}} \left( {{\text{initial biomass}},{\text{ mg}}} \right)} \right]/{\text{time in years}}. \, $$
We calculated the Growth plasticity index (Eq. 3, Valladares et al. 2000) for RBGR, RHGR, ABGR and AHGR and took the mean to represent each species.
$$ {\text{Growth plasticity index }} = \, \left[ {{\text{growth at 5}}. 7 \% \; - \;{\text{growth at 1}}. 2 \% } \right]/{\text{growth at 5}}. 7 \% $$

Seedling survival at 1.2 % PAR was calculated as the percentage of individuals for each species that survived the duration of the experiment. We used Spearman ranks to test for correlations between cold tolerance and our three measures of gap dependence (the GDI, mean growth plasticity index and seedling survival at 1.2 % PAR). Differences in plant traits, growth rates, growth plasticity indices and survival between conifer and angiosperm species were assessed using Kruskal–Wallis tests. All analyses were performed using Rv. 13.0.


Gap dependence and seedling responses to varying light

Seedling growth rates varied widely among species, more so at 1.2 % PAR than at 5.7 % PAR (Fig. 2). RHGR varied ninefold among species at 1.2 % PAR but only twofold at 5.7 % PAR (Fig. 2c). ABGR at 5.7 % PAR was marginally greater in angiosperms (mean = 12.8 mg year−1) than conifers (mean = 6.51 mg year−1; Kruskal–Wallis test χ12 = 3.00, P = 0.083; Fig. 2b) but there were no other statistical differences in growth rates or growth plasticity indices between angiosperms and conifers. However, the four conifer species (all small-leaved; Table 1) had high growth plasticity indices (mean = 0.88 ± 0.04 1SE), while the four angiosperm species had larger leaves (Table 1) and widely-ranging growth plasticity indices (mean = 0.67 ± 0.15 1SE; Table 2). Biomass growth rates at 1.2 % PAR were negative for two species (B. tawa and Dacrycarpus dacrydioides; Fig. 2a, b) and these two species had the highest growth plasticity indices (Table 2) suggestive of very low shade tolerance and high gap dependence, which was not apparent from the GDI developed from field observations (Table 2). Only height growth of Melicytus ramiflorus was similar under both PAR treatments (Fig. 2) with one of the lowest growth plasticity indices to varying light (Table 2) indicative of low dependency on gaps for growth.
Fig. 2

Seedling growth responses (means with 1 SE) to low (1.2 % PAR; black bars) and high (5.7 % PAR; grey bars) light. Species are ordered in two groups (angiosperms and conifers) from left to right in order of their mean plasticity of growth to shading (see Table 2). a is relative biomass growth rate, RBGR; b absolute biomass growth rate, ABGR; c relative height growth rate, RHGR; d absolute height growth rate, AHGR. PAR is photosynthetically active radiation

Table 2

Three measures of gap dependence in eight tree species from temperate evergreen rainforests in New Zealand


GDI seedling occurrence

% survival at 1.2 % PAR

Growth plasticity index


 Beilschmiedia tawa

1.2 (0.2)



 Nestegis cunninghamii

2.0 (0.3)



 Melicytus ramiflorus

2.4 (0.2)



 Weinmannia racemosa

2.6 (0.2)




 Prumnopitys ferruginea

1.6 (0.2)



 Pseudotsuga menziesii

2.8 (0.5)



 Dacrycarpus dacrydioides

2.8 (0.4)



 Podocarpus totara

3.0 (0.3)



Gap Dependence Index (GDI) is the mean of ordinal scores (±1 SE) describing the light environment where seedlings naturally occur (following Houter and Pons 2012; see Materials and methods). Seedling survival at 1.2 % PAR (photosynthetically active radiation) measures the persistence of seedlings in low light under experimental conditions (see Materials and methods). Growth plasticity indices (after Valladares et al. 2000) express seedling growth at 1.2 % PAR relative to growth at 5.7 % PAR (see Eq. 3 and Fig. 2). A value close to 0 (either positive or negative) indicates that a species performed equally well in both treatments; values close to 1 indicate that a species performed substantially better at 5.7 % PAR; values greater than 1 indicate that a species lost biomass at 1.2 % PAR. The index is the mean for RHGR (relative height growth rate), AHGR (absolute height growth rate), RBGR (relative biomass growth rate) and ABGR (absolute biomass growth rate)

Survival at 1.2 % PAR was just 20 % for the non-native conifer P. menziesii. In contrast, four native species (B. tawa, M. ramiflorus, Nestegis cunninghamii, Podocarpus totara) had 100 % survival at 1.2 % PAR (Table 2). There was no significant difference between conifer and angiosperm species in their survival at 1.2 % PAR (Kruskal–Wallis test χ12 = 1.17, P = 0.28).

There were no significant correlations between our three measures of gap dependence (Spearman Rank correlations: GDI and survival at 1.2 % PAR r = −0.27, P = 0.52; GDI and growth plasticity index r = 0.06, P = 0.89; survival at 1.2 % PAR and growth plasticity index r = −0.28, P > 0.1; Table 2). The absence of a correlation between growth plasticity and survival at 1.2 % PAR was unexpected and indicates that for these species, for the 2-year duration of our experiment, low or even negative growth at 1.2 % PAR did not result in low survival at 1.2 % PAR. This was pronounced in B. tawa and D. dacrydioides which both had negative biomass growth rates at 1.2 % PAR but high (>90 %) survival (Fig. 2; Table 2). In this regard, the growth plasticity index and survival at 1.2 % PAR provide complementary but independent measures of gap dependence.

Seedling chlorophyll fluorescence

Cold tolerance of photosynthesis, that is the proportional change in leaf chlorophyll fluorescence (Fv/Fm) after the low-temperature high-light treatment, ranged widely among species (Fig. 3). Photosynthesis of three of the four conifer species (P. menziesii, Prumnopitys ferruginea and P. totara) was highly cold tolerant and mean Fv/Fm changed by <10 % between the baseline and post-treatment measurements (Fig 3). In contrast, mean Fv/Fm declined by 30 and 52 % in two large-leaved angiosperm species, M. ramiflorus and B. tawa, respectively (Fig. 3). This difference in the mean proportional change in Fv/Fm between conifers and angiosperms was marginally significant (Kruskal–Wallis test χ12 = 3.00, P = 0.083).
Fig. 3

Seedling leaf chlorophyll fluorescence (Fv/Fm) before and after a high light/sub-zero temperature treatment designed to mimic an early-morning frost event in eight New Zealand tree species under experimental conditions. a Pre-treatment Fv/Fm (circles) and post-treatment Fv/Fm (triangles) for four angiosperms (open symbols) and four conifer (filled symbols) species. b Proportional change in Fv/Fm, a measure of cold tolerance. Data are means (±1 SE) (n = 4 for each species). Species codes are in Table 1

Associations between cold tolerance and measures of gap dependence

Cold tolerance of photosynthesis was positively correlated with seedling survival at 1.2 % PAR (P < 0.10; Fig. 4b) but this result was driven by a single species, P. menziesii, which had low survivorship at 1.2 % PAR and high cold tolerance. Species formed a ‘triangular relationship’, rather than a linear one, between seedling survival and chlorophyll fluorescence. High survival at low light was associated with any degree of cold tolerance, while low survival at low light was only associated with high cold tolerance (Fig. 4b). There was no relationship between cold tolerance and the GDI (Spearman Rank correlation r = −0.36, P > 0.1; Fig. 4a) or seedling growth plasticity to varying light (Spearman Rank correlation r = 0.00, P > 0.1; Fig. 4c).
Fig. 4

Relationships between seedling cold tolerance, defined here as the proportional change in chlorophyll fluorescence after exposure to high light at sub-zero temperatures, and three measures of gap dependence: a the gap dependence index b seedling survival in low light (1.2 % PAR) c plasticity of seedling growth between 1.2 and 5.7 % PAR. Open symbols are angiosperm species and filled symbols are conifers. Species codes are in Table 1

Gap-demanding species based on the glasshouse study were smaller-leaved, as predicted (Spearman Rank correlations: leaf size and survival at 1.2 % PAR r = 0.68, P = 0.07; leaf size and growth plasticity index r = −0.74, P = 0.04) but did not have lower LMA than shade-tolerant species (Spearman Rank correlations: LMA and survival at 1.2 % PAR r = 0.32, P = 0.44; LMA and growth plasticity index r = −0.07, P = 0.87). Cold tolerance of photosynthesis was not significantly associated with either leaf size (Spearman Rank correlation r = 0.55, P = 0.16) or LMA (Spearman Rank correlation r = −0.02, P = 0.96).

We predicted that some species with a low GDI based on field observations would exhibit high plasticity to varying light and low resistance to cold photoinhibition. Such species are confined to shaded microsites in the field to minimise exposure to sub-zero temperatures. Two species had a GDI <2 (B. tawa and P. ferruginea) and had highly plastic growth responses to PAR (Table 2) but only one of them—B. tawa—also had low cold tolerance (Figs. 3 and 4).


Is cold tolerance essential for gap dependent species?

Gap-dependent evergreen species, defined using the gap dependence index (Fig. 1a) or survival at 1.2 % PAR (Fig. 1b), were all resistant to cold photoinhibition, supporting our hypothesis that cold tolerance is a prerequisite for exploiting the gap niche. However, some species typical of shaded understoreys were also cold tolerant (e.g. P. ferruginea) indicating that gap-dependence and cold tolerance are not linearly dependent. This finding expands on previous work contrasting cold tolerance of gap-dependent and shade-tolerant species. Reyes-Díaz et al. (2009) demonstrated that seedlings of a gap-dependent species (Nothofagus dombeyi) were more cold tolerant than a shade-tolerant congeneric (N. nitida)—our study highlights that shade-tolerant species can also be cold tolerant. We suggest that photosynthesis of many evergreen species in temperate forests is likely to be at least partially cold tolerant as their leaves are costly to construct and replacement of damaged tissues is incompatible with their resource-conservative strategy (Öquist and Huner 2003). Chlorophyll fluorescence of six of our eight species changed by less than 25 % after the low temperature/high light treatment (Fig. 3) supporting this hypothesis. Interestingly, the two species that were least cold tolerant (B. tawa and M. ramiflorus) are both vigorous sprouters, a trait that may compensate for periodic loss of aboveground tissues following cold events.

Leaf size and LMA were not correlated across the eight species studied here, and neither trait was correlated with cold tolerance or gap dependence. Species with high growth plasticity to varying light tended to have smaller leaves, as predicted for gap-demanding species (Niinemets and Kull 1994; Valladares and Niinemets 2008), but there was no relationship with LMA. Our data suggest that LMA, which varied 3.5-fold across the eight species, was not critical for determining interspecific differences in responses to either shade or cold. This was unexpected as shade tolerant evergreen species typically have high LMA. Our data suggest that leaf size was a more important trait underpinning species’ responses to shade, and to some extent cold. Although the correlation between cold tolerance and leaf size was non-significant, conifers were more cold tolerant than angiosperms (Table 2), and these species had substantially smaller leaves (Table 1).

The strongest evidence for an association between gap dependency and cold tolerance came from the non-native tall conifer species P. menziesii that had low survival in low light, high growth plasticity to varying light and the smallest chlorophyll fluorescence response. Photosynthetic responses to light in this species indicate that it is partially shade tolerant relative to many conifers (Bond et al. 1999) and this has been used to predict that it could invade old-growth forest understoreys (Dehlin et al. 2008). Our data suggest that this prediction is unlikely in very low light (<1.5 % PAR) situations, which is supported by two field studies of P. menziesii seedling performance. One study planted seedlings of P.menziesii into forest stands with 0.5, 5.1 and 13.3 % PAR in montane New Zealand forests and reported that only 40 % of P. menziesii seedlings survived at 0.5 % PAR relative to >80 % of seedlings at >5 % PAR (Dehlin et al. 2008). A second examined seedling density of P.menziesii in montane Mediterranean shrublands and revealed that seedlings only persisted at low tree densities under moderate canopy cover (Broncano et al. 2005). Lastly, this species was the most cold-tolerant in our study reflecting strong adaptation to long periods of sub-zero temperatures in much of its home range (western N America), relative to New Zealand tree species that are adapted to predominantly oceanic climates.

Many studies have demonstrated a trade-off between survival in low light and growth in high light (e.g. Kobe et al. 1995; Lusk and Del Pozo 2002) supporting the theoretical expectation that traits advantageous for survival at low light necessarily slow growth responses to high light (Dent and Burslem 2009). This trade-off has been used to explain coexistence in forest understoreys as it accounts for niche partitioning among species across light gradients. We found no evidence for such a relationship—most species had high survival in low light alongside high plasticity of growth to light. This limited the extent to which species could be defined as either gap-dependent or shade-tolerant from experimental data. Mahmoud and Grime (1974) claimed that negative relative growth rates in low light were an unambiguous measure of shade intolerance yet we found that two species—B. tawa and D. dacrydioides—maintained negative RBGR alongside high (>93 %) survival at 1.2 % PAR. Similarly, Baltzer and Thomas (2007) reported negative growth rates in low light, without mortality, in both shade-tolerant and gap-demanding species. One cautionary note is that our survival data are based on 15 seedlings for each species over 2 years, which is statistically sufficient for robustly detecting survival rates of 85 % of less (based on Peltzer et al. 2005). Six of our eight species had higher survival rates than this (four had 100 % survival) and so these rates must be considered indicative of seedling resilience of low light, rather than precise estimates of survival. Even so, the lack of a relationship between negative growth and survival at 1.2 % PAR highlights the resilience to mortality of established forest seedlings in low light regardless of low or even negative growth rates and emphasises the long timescales over which trade-offs shaping shade tolerance are expressed in long-lived seedlings (Antos et al. 2005).

Does cold-induced photoinhibition confine some species to shaded microsites?

We predicted that some species with a low GDI based on field observations would exhibit high plasticity to varying light and low resistance to cold photoinhibition. Such species are confined to suboptimal shaded niches by their low cold tolerance. This prediction was supported by one of our species—B. tawa—which is widely considered to be shade tolerant and is largely confined to shaded microsites where sub-zero temperatures are infrequent relative to adjacent gaps (Knowles and Beveridge 1982). This species regenerates effectively in small gaps with protection from adjacent canopies (West 1995), but not large gaps (Knowles and Beveridge 1982). Further, population stability of this species is compromised by selective logging, because removal of adjacent and emergent canopies exposes individuals to high light and low temperatures (West 1995). Our data clearly reveal that seedling growth of this species increases substantially with greater light under experimental conditions in the absence of sub-zero temperatures. Indeed, this species was one of only two to present negative biomass growth rates in low light. Our chlorophyll fluorescence data suggest that the regeneration niche of this species partly reflects low cold tolerance in gap sites, and a tolerance of shade (i.e. high survival) rather than a preference for shade. Shade tolerance is widely known to be context-dependent varying with drought status and nutrient availability (e.g. Coomes and Grubb 2000; Sanchez-Gomez et al. 2006; Valladares and Niinemets 2008) and our data contribute evidence that shade tolerance of at least one tree species (B. tawa) is shaped by the interaction between cold tolerance and tolerance of low light. Under field conditions, seedling mortality arising from repeated exposure to high light at low temperatures would filter out this species from gaps, thus generating the observed apparent shade-tolerant niche. Our experimental protocol tested the impact of cold-induced photoinhibition in cold-acclimatised trees. Photoinhibition in these same taxa—and the resulting energetic costs (Raven 2011)—may be greater when sub-zero temperatures and high PAR co-occur in non-hardened seedlings (Öquist et al. 1987; Somersalo and Krause 1989).

Niche expansion in warmer climates

Sub-zero temperatures exert a strong filter on seedling regeneration in oceanic forest ecosystems. In contrast to continental ecosystems, the frequency, intensity and timing of cold events varies widely among years. Many temperate woody ecosystems in the Southern Hemisphere are strongly oceanic and have lower cold tolerance than Northern Hemisphere species at similar latitudes (Feild and Brodribb 2001). Climate warming has already reduced frost frequency in temperate regions (Inouye 2000) and frost frequency is predicted to decline throughout New Zealand over the next century (MfE 2008). We anticipate niche expansion of cold-sensitive taxa currently found in shaded microsites, particularly at their cold-temperature-range limits. Niche expansion through the interaction between climate and shade is likely to be widespread. Shade tolerance of Pinus sylvestris varies throughout Europe with climate; the optimum light intensity for seedling emergence declines from the boreal to the Mediterranean because high-light sites in Mediterranean regions are exposed to droughts that increase seedling mortality (Castro et al. 2004). Climate change may generate novel niches, both for indigenous species currently limited by cold-induced photoinhibition or drought and for non-native species. Understanding how climate change shapes the regeneration niche will be critical in island regions such as New Zealand where ecosystems have low resilience to invasions (Allen and Lee 2006) and shade-tolerant invasive trees pose a significant threat to forest composition.



We thank Ellen Cieraad, Tahae Doherty, Chris Morse, Rowan Buxton and Gaye Rattray for assistance; Matt McGlone, Norm Mason and Duane Peltzer for reviews; and the NZ Ministry of Business, Innovation and Employment for funding.


  1. Allen RB, Lee WG (2006) Biological invasions in New Zealand. Springer, BerlinCrossRefGoogle Scholar
  2. Antos JA, Guest HJ, Parish R (2005) The tree seedling bank in an ancient montane forest: stress tolerators in a productive habitat. J Ecol 93:536–543CrossRefGoogle Scholar
  3. Augspurger CK (2011) Frost damage and its cascading negative effects on Aesculus glabra. Plant Ecol 212:1193–1203CrossRefGoogle Scholar
  4. Bader MY, Geloof I, Rietkerk M (2007) High solar radiation hinders tree regeneration above the alpine treeline in northern Ecuador. Plant Ecol 191:33–45CrossRefGoogle Scholar
  5. Ball MC, Hodges VS, Laughlin GP (1991) Cold-induced photoinhibition limits regeneration of snow gum at tree-line. Funct Ecol 5:663–668CrossRefGoogle Scholar
  6. Baltzer JL, Thomas SC (2007) Determinants of whole-plant light requirements in Bornean rain forest tree saplings. J Ecol 95:1208–1221CrossRefGoogle Scholar
  7. Bond BJ, Farnsworth BT, Coulombe RA, Winner WE (1999) Foliage physiology and biochemistry in response to light gradients in conifers with varying shade tolerance. Oecologia 120:183–192CrossRefGoogle Scholar
  8. Broncano MJ, Vila M, Boada M (2005) Evidence of Pseudotsuga menziesii naturalization in montane Mediterranean forests. For Ecol Manag 211:257–263CrossRefGoogle Scholar
  9. Canham CD, Denslow JS, Platt WJ, Runkle JR, Spies TA, White PS (1990) Light regimes beneath closed canopies and tree-fall gaps in temperate and tropical forests. Can J For Res 20:620–631CrossRefGoogle Scholar
  10. Carswell FE, Richardson SJ, Doherty JE, Allen RB, Wiser SK (2007) Where do conifers regenerate after selective harvest? A case study from a New Zealand conifer–angiosperm forest. For Ecol Manag 253:138–147CrossRefGoogle Scholar
  11. Carswell FE, Doherty JE, Allen RB, Brignall-Theyer ME, Richardson SJ, Wiser SK (2012) Quantification of the effects of aboveground and belowground competition on growth of seedlings in a conifer–angiosperm forest. For Ecol Manag 269:188–196CrossRefGoogle Scholar
  12. Castro J, Zamora R, Hodar JA, Gomez JM (2004) Seedling establishment of a boreal tree species (Pinus sylvestris) at its southernmost distribution limit: consequences of being in a marginal Mediterranean habitat. J Ecol 92:266–277CrossRefGoogle Scholar
  13. Coomes DA, Grubb PJ (2000) Impacts of root competition in forests and woodlands: a theoretical framework and review of experiments. Ecol Monogr 70:171–207CrossRefGoogle Scholar
  14. Coop JD, Givnish TJ (2008) Constraints on tree seedling establishment in montane grasslands of the Valles Caldera, New Mexico. Ecology 89:1101–1111PubMedCrossRefGoogle Scholar
  15. Cornelissen JHC, Lavorel S, Garnier E, Diaz S, Buchmann N, Gurvich DE, Reich PB, ter Steege H, Morgan HD, van der Heijden MGA, Pausas JG, Poorter H (2003) A handbook of protocols for standardised and easy measurement of plant functional traits worldwide. Aust J Bot 51:335–380CrossRefGoogle Scholar
  16. Cunningham SC, Read J (2006) Foliar temperature tolerance of temperate and tropical evergreen rain forest trees of Australia. Tree Physiol 26:1435–1443PubMedCrossRefGoogle Scholar
  17. Dehlin H, Peltzer DA, Allison VJ, Yeates GW, Nilsson MC, Wardle DA (2008) Tree seedling performance and below-ground properties in stands of invasive and native tree species. NZ J Ecol 32:67–79Google Scholar
  18. Dent DH, Burslem D (2009) Performance trade-offs driven by morphological plasticity contribute to habitat specialization of Bornean tree species. Biotropica 41:424–434CrossRefGoogle Scholar
  19. Enright NJ, Ogden J (1995) The southern conifers: a synthesis. In: Enright NJ, Hill RS (eds) Ecology of the southern conifers. Melbourne University Press, Carlton, pp 271–287Google Scholar
  20. Feild TS, Brodribb T (2001) Stem water transport and freeze-thaw xylem embolism in conifers and angiosperms in a Tasmanian treeline heath. Oecologia 127:314–320CrossRefGoogle Scholar
  21. George LO, Bazzaz FA (1999) The fern understory as an ecological filter: growth and survival of canopy-tree seedlings. Ecology 80:846–856CrossRefGoogle Scholar
  22. Grubb PJ, Lee WG, Kollmann J, Wilson JB (1996) Interaction of irradiance and soil nutrient supply on growth of seedlings of ten European tall-shrub species and Fagus sylvatica. J Ecol 84:827–840CrossRefGoogle Scholar
  23. Holdaway RJ, Richardson SJ, Peltzer DA, Coomes DA (2011) Species and community-level patterns in fine root traits along a 120,000 year soil chronosequence. J Ecol 99:954–963CrossRefGoogle Scholar
  24. Houter NC, Pons TL (2012) Ontogenetic changes in leaf traits of tropical rainforest trees differing in juvenile light requirement. Oecologia 169:33–45PubMedCrossRefGoogle Scholar
  25. Inouye DW (2000) The ecological and evolutionary significance of frost in the context of climate change. Ecol Lett 3:457–463CrossRefGoogle Scholar
  26. Kelly D (1987) Slow recovery of Beilschmiedia tawa after severe frosts in inland Taranaki, New Zealand. NZ J Ecol 10:137–140Google Scholar
  27. Knowles B, Beveridge AE (1982) Biological flora of New Zealand. 9. Beilschmiedia tawa (A. Cunn) Benth et Hook. f. ex Kirk (Lauraceae) Tawa. NZ J Bot 20:37–54CrossRefGoogle Scholar
  28. Kobe RK, Pacala SW, Silander JA, Canham CD (1995) Juvenile tree survivorship as a component of shade tolerance. Ecol Appl 5:517–532CrossRefGoogle Scholar
  29. Krause GH (1994) Photoinhibition induced by low temperatures. In: Baker NR, Bowyer JR (eds) Photoinhibition of Photosynthesis: from molecular mechanisms to the field. BIOS Scientific, Oxford, pp 331–348Google Scholar
  30. Lusk CH, Del Pozo A (2002) Survival and growth of seedlings of 12 Chilean rainforest trees in two light environments: gas exchange and biomass distribution correlates. Aust Ecol 27:173–182CrossRefGoogle Scholar
  31. Lusk CH, Reich PB, Montgomery RA, Ackerly DD, Cavender-Bares J (2008) Why are evergreen leaves so contrary about shade? Trends Ecol Evol 23:299–303PubMedCrossRefGoogle Scholar
  32. Lusk CH, Duncan RP, Bellingham PJ (2009) Light environments occupied by conifer and angiosperm seedlings in a New Zealand podocarp–broadleaved forest. NZ J Ecol 33:83–89Google Scholar
  33. Mahmoud A, Grime JP (1974) Comparison of negative relative growth-rates in shaded seedlings. New Phytol 73:1215–1219CrossRefGoogle Scholar
  34. McGlone MS, Dungan RJ, Hall GMJ, Allen RB (2004) Winter leaf loss in the New Zealand woody flora. NZ J Bot 42:1–19CrossRefGoogle Scholar
  35. McGlone MS, Richardson SJ, Jordan GJ (2010) Comparative biogeography of New Zealand trees: species richness, height, leaf traits and range sizes. NZ J Ecol 34:137–151Google Scholar
  36. MfE (2008) Climate change effects and impacts assessment: a guidance manual for local government in New Zealand. Ministry for the Environment, WellingtonGoogle Scholar
  37. Niinemets U, Kull K (1994) Leaf weight per area and leaf size of 85 Estonian woody species in relation to shade tolerance and light availability. For Ecol Man 70:1–10CrossRefGoogle Scholar
  38. Niinemets U, Valladares F (2006) Tolerance to shade, drought, and waterlogging of temperate Northern Hemisphere trees and shrubs. Ecol Monogr 76:521–547CrossRefGoogle Scholar
  39. Öquist G, Huner NPA (2003) Photosynthesis of overwintering evergreen plants. Ann Rev Plant Biol 54:329–355CrossRefGoogle Scholar
  40. Öquist G, Greer DH, Ögren E (1987) Light stress at low temperature. In: Kyle DJ, Osmond CB, Arntzen CJ (eds) Photoinhibition. Elsevier Science, Amsterdam, pp 67–87Google Scholar
  41. Peltzer DA, Allen RB, Rogers GM (2005) Dieback and recruitment of the forest dominants Nothofagus fusca and Libocedrus bidwillii, central North Island, New Zealand, vol 255. Department of Conservation, Science for Conservation, Wellington, p 33Google Scholar
  42. Raven JA (2011) The cost of photoinhibition. Physiol Plant 142:87–104PubMedCrossRefGoogle Scholar
  43. Reyes-Díaz M, Ivanov AG, Huner NPA, Alberdi M, Corcuera LJ, Bravo LA (2009) Thermal energy dissipation and its components in two developmental stages of a shade-tolerant species, Nothofagus nitida, and a shade-intolerant species, Nothofagus dombeyi. Tree Phys 29:651–662CrossRefGoogle Scholar
  44. Richardson SJ, Allen RB, Doherty JE (2008) Shifts in leaf N:P ratio during resorption reflect soil P in temperate rainforest. Funct Ecol 22:738–745CrossRefGoogle Scholar
  45. Robakowski P (2005) Susceptibility to low-temperature photoinhibition in three conifers differing in successional status. Tree Phys 25:1151–1160CrossRefGoogle Scholar
  46. Sakai A, Paton DM, Wardle P (1981) Freezing resistance of trees of the south temperate zone, especially subalpine species of Australasia. Ecology 62:563–570CrossRefGoogle Scholar
  47. Sanchez-Gomez D, Zavala MA, Valladares F (2006) Seedling survival responses to irradiance are differentially influenced by low-water availability in four tree species of the Iberian cool temperate-Mediterranean ecotone. Acta Oecol Int J Ecol 30:322–332CrossRefGoogle Scholar
  48. Somersalo S, Krause GH (1989) Photoinhibition at chilling temperature–fluorescence characteristics of unhardened and cold-acclimated spinach leaves. Planta 177:409–416CrossRefGoogle Scholar
  49. Valladares F, Niinemets U (2008) Shade tolerance, a key plant feature of complex nature and consequences. Annu Rev Ecol Evol Syst 2008:237–257CrossRefGoogle Scholar
  50. Valladares F, Wright SJ, Lasso E, Kitajima K, Pearcy RW (2000) Plastic phenotypic response to light of 16 congeneric shrubs from a Panamanian rainforest. Ecology 81:1925–1936CrossRefGoogle Scholar
  51. Wardle P (1985) New Zealand timberlines. 3. A synthesis. NZ J Bot 23:263–271CrossRefGoogle Scholar
  52. West CJ (1995) Sustainability of Beilschmiedia tawa-dominated forest in New Zealand: population predictions based on transition matrix model analysis. Aust J Bot 43:51–71CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2013

Authors and Affiliations

  • Sarah J. Richardson
    • 1
  • Karen I. Bonner
    • 1
  • Christopher P. Bickford
    • 1
    • 2
  1. 1.Landcare ResearchLincolnNew Zealand
  2. 2.Department of BiologyKenyon CollegeGambierUSA

Personalised recommendations