Annals of Botany 99: 537–544, 2007 doi:10.1093/aob/mcl283, available online at www.aob.oxfordjournals.org

Trade-offs Between Seedling Growth and Survival in Deciduous Broadleaved Trees in a Temperate Forest K E NJ I S E IWA* Laboratory of Forest Ecology, Graduate School of Agricultural Science, Tohoku University, Naruko-onsen, Osaki, Miyagi 989-6711, Japan Received: 16 August 2006 Revision requested: 23 October 2006 Accepted: 16 November 2006 Published electronically: 22 January 2007

† Background and Aims In spatially heterogeneous environments, a trade-off between seedling survival and relative growth rate may promote the coexistence of plant species. In temperate forests, however, little support for this hypothesis has been found under field conditions, as compared with shade-house experiments. Performance trade-offs were examined over a large resource gradient in a temperate hardwood forest. † Methods The relationship between seedling survival and seedling relative growth rate in mass (RGRM) or height (RGRH) was examined at three levels of canopy cover (forest understorey, FU; small gap, SG; and large gap, LG) and at two microsites within each level of canopy cover ( presence or absence of leaf litter) for five deciduous broad-leaved tree species with different seed sizes. † Key Results Within each species, both RGRM and RGRH usually increased with increasing light levels (in the order FU , SG , LG), whereas little difference was observed based on the presence or absence of litter. Seedling survival in FU was negatively correlated with both RGRM and RGRH in both LG and SG. The trade-off between high-light growth and low-light survival was more evident in the relationship with LG as compared with SG. An intraspecific trade-off between survival and RGR was observed along environmental gradients in Acer mono, whereas seedlings of Betula platyphylla var. japonica survived and grew better in LG. † Conclusions The results presented here strongly support the idea of light gradient partitioning (i.e. species coexistence) in spatially heterogeneous light environments in temperate forests, and that further species diversity would be promoted by increased spatial heterogeneity. The intraspecific trade-off between survival and RGR in Acer suggests that it has broad habitat requirements, whereas Betula has narrow habitat requirements and specializes in high-light environments. Key words: Coexistence, gap, gap size, habitat selection, habitat width, light, niche partitioning, relative growth rate, seed size, successional status.

IN TROD UCT IO N Seedling relative growth rate (RGR) is a key variable reflecting shade tolerance, successional status, and the regeneration niche of plants in their natural habitats (Grime and Hunt, 1975; Tilman, 1988; Kitajima, 1994; Cornelissen et al., 1996; Grubb et al., 1996; Reich et al., 1998a; Poorter, 1999; Walters and Reich, 2000; Baraloto et al., 2005). It is thought that species differences in seedling RGR across environmental resource (i.e. irradiance, water and nutrients) gradients contribute to the maintenance of forest species richness (Grubb et al., 1996; Sack and Grubb, 2001). Seed size, also considered an important component of shade tolerance and successional status, is closely related to RGR. Small-seeded and early successional species tend to have higher potential RGR; thus, their seedlings outgrow seedlings from large-seeded species, particularly when resources are abundant, such as in large canopy gaps (Maranon and Grubb, 1993; Swanborough and Westoby, 1996; Seiwa and Kikuzawa, 1991, 1996; Bloor and Grubb, 2003; Poorter and Rose, 2005). In contrast, large-seeded and late successional species usually have higher survival rates than small-seeded species, even under stressful environmental conditions such as shade, deep leaf litter, or damage by * E-mail [email protected]

pathogens or herbivores (see references in Westoby et al., 1992; Seiwa and Kikuzawa, 1996; Paz and Martinez-Ramos, 2003; Green and Juniper, 2004). Thus, a trade-off between low-light survival and high-light RGR could result in a similar probability of seedling establishment for species with both large and small seeds (Kitajima, 1994; Baraloto et al., 2005). In spatially heterogeneous environments, such a trade-off would promote the coexistence of species with different seed sizes (Kobe, 1999; Chesson, 2000; Walters and Reich, 2000; Dalling and Hubbell, 2002; Baraloto et al., 2005; Sanchez-Gomez et al., 2006). There is increasing evidence for performance trade-offs between survival and growth in woody species of tropical forests at both seedling (Dalling and Hubbell, 2002; Pearson et al., 2003; Wright et al., 2003; Baraloto et al., 2005; Gilbert et al., 2006) and sapling stages (Hubbell and Foster, 1992; Kobe, 1999; Davies, 2001; Poorter and Arets, 2003; Gilbert et al., 2006). In temperate forests, the trade-offs are also well documented for the sapling stage (e.g. Kobe et al., 1995; Pacala et al., 1996), but are reported less frequently for the seedling stage. Furthermore, most of the studies were not conducted under field conditions (but see Schreeg et al., 2005), but under experimental conditions (e.g. Walters and Reich, 1996, 2000; Kaerke et al., 2001), in which seedlings are usually

# The Author 2007. Published by Oxford University Press on behalf of the Annals of Botany Company. All rights reserved. For Permissions, please email: [email protected]

538

Seiwa — Seedling Growth and Survival in Temperate Trees

protected from both biotic and abiotic stressors such as herbivores, pathogens, temperature extremes and shading by neighbouring plants. Findings from such studies assume that seedling mortality is mainly caused by the collapse of the whole-plant carbon balance. Under field conditions, however, most tree seedlings are killed directly by one or several stressors. Most tree seedlings invest resources in defence against stressors and in energy storage. Resource allocation to defence and storage can influence survivorship by providing a buffer against temporary resource deprivation (e.g. restoring tissue lost to herbivores; Coley et al., 1985; Canham et al., 1999; Dalling and Hubbell, 2002; Stamp, 2003). Under field conditions, severe competition for light with neighbouring plants also usually occurs, particularly in high-light environments such as canopy gaps. In such high-light environments, earlier vertical growth is advantageous for seedling establishment (Ross and Harper, 1972; Seiwa, 2000) because the light conditions experienced by the seedlings are dramatically improved with increasing seedling height (e.g. Givnish, 1982). If seedling vertical growth is critical for seedling survival, a performance trade-off between seedling survival in low light and RGR in high light could be evaluated by measuring the relative growth rate in height (RGRH) and relative growth rate in mass (RGRM) under field conditions. For temperate trees, there has been little research to examine intraspecific trade-offs between survival and RGR along resource gradients, particularly at the seedling stage, although sapling-based studies have been conducted (e.g. Kobe et al., 1995; Pacala et al., 1996). Although interspecific differences in seedling performance could reveal a number of strategic trade-offs restricting a species to optimal performance at a narrow range of environmental gradients, in reality, all of the species co-occurring within a forest community are not always restricted to a narrow range of the habitat. Tree seedlings usually show a plastic response to spatially heterogeneous environments, and the extent of this plasticity may be related to a species’ habitat selection ability (Bazzaz, 1991; Seiwa, 1998). For example, if seedling survival is negatively correlated with seedling RGR along the environmental gradient, the performance trade-off suggests a broad niche for the species, whereas a positive relationship shows a narrow niche and specialized adaptation to a particular habitat. Thus, it is important to recognize that contrasts observed within species along two extremes of the resource continuum would be useful for evaluating the habitat requirements or status of shade tolerance (i.e. resource requirements) within a forest community, as well as the contrast among species at the two extremes. Litter has a major effect on plant community structure through seedling establishment (Facelli and Pickett, 1991; Seiwa and Kikuzawa, 1996). For example, seedling emergence by penetration through litter was positively associated with seed size, whereas the absence of litter inhibited the emergence of large-seeded species because it allows the soil and seeds to dry out (Shaw, 1968; Seiwa and Kikuzawa, 1996). Because litter accumulation usually varies with the disturbance regime among, and even

within, sites (e.g. Facelli and Pickett, 1991), litter accumulation would affect not only the emergence stage, but also subsequent seedling growth and mortality. Thus, it is important to evaluate the effects of litter on the seedling performance trade-off between RGR and survival, and the consequent community structure. Therefore, in this study, hypotheses were tested concerning the adaptation and coexistence of tree species along resource gradients by examining the among- and within-species trade-offs between seedling survival and RGRM and RGRH at a wide range of resource levels. The relationship between seedling survival and seedling RGR was examined at three levels of canopy cover (forest understorey, small gap and large gap) and at two microsites within each level of canopy cover ( presence or absence of leaf litter) for five deciduous broad-leaved tree species with different seed sizes in a temperate forest. Specifically, the following questions were asked. To what extent is RGR related to seed size and resource level? Is there an interspecific trade-off between seedling survival at low-light (forest understorey) and RGR at high light (small or large gap)? Is there a significant relationship between seedling survival and RGR along the resource gradient for each species? M AT E R I A L S A N D M E T H O D S Species

The five tree species studied are common in cooltemperate deciduous broad-leaved forests of northern Japan, and differ in seed mass, successional status and leaf phenology (Kikuzawa, 1983; Koike, 1988; Seiwa and Kikuzawa, 1991, 1996; Seiwa et al., 2006). Betula platyphylla var. japonica and Alnus hirsuta are small-seeded, shade-intolerant pioneer species that occur on abandoned arable or open woodlands. Cercidiphyllum japonicum is also a small-seeded species but is mid-successional, mainly occurring in small gaps on riverbanks. Acer mono and Quercus mongolica var. grosseserrata are shadetolerant climax species that occur in a wide range of habitats from moderately dense forest to the forest edge. As all of the species belong to different genera, they are henceforth referred to by their generic names. Mean seed reserve (dry mass of seed minus testa) of Quercus, Acer, Alnus, Cercidiphyllum and Betula was 1097, 46.3, 0.870, 0.173 and 0.102 mg, respectively. Methods of seed collection and determination of the seed reserve can be found in Seiwa and Kikuzawa (1996). Study area

This study was conducted in a deciduous broad-leaved forest on flat ground at 100 – 110 m a.s.l. in the Hokkaido Forest Experimental Station (HFES), Bibai, Hokkaido, Japan (438150 N, 1418500 E). In the year of the study (1988) the mean temperature was 7.1 8C and the annual rainfall was 1042 mm. At this location, 30 % of the annual precipitation falls as snow. Snow cover lasted from late November to early April. Mean monthly temperatures

Seiwa — Seedling Growth and Survival in Temperate Trees range from 27.9 8C in February to 22.3 8C in August. Quercus was the dominant canopy tree and dwarf bamboo (Sasa senanensis) was the dominant understorey species (details in Seiwa and Kikuzawa, 1996). The density of trees with stems 2 cm or more in diameter, at chest height, was about 2300 trees ha21. The mean height of trees forming the forest canopy was 16 m. A small gap (7  10 m) and a large gap (50  70 m) were created by felling trees, shrubs and Sasa in October 1987. Study sites (i.e. forest understorey, FU; small gap, SG; and large gap, LG) were located within 50 m of each other. All three sites have the same kind of soil, topography and vegetation. The primary difference among sites was whether a forest or a gap occurred on the site. Photosynthetic photon flux density (PPFD) was measured at microsites with bare soil in each site, at 10 cm above the ground, using a KIP photo sensor (Koito Co. Ltd) calibrated with a LI-COR quantum sensor (LI-90SB). The PPFD was highest in LG and lowest in FU (Seiwa and Kikuzawa, 1996). Before canopy leaf emergence, the relative PPFD in SG and FU was 83 and 63 %, respectively, of the PPFD in LG. The relative PPFD declined with leaf expansion of canopy trees and reached a minimum of 26 % at SG and 4 % at FU in early July. After September, relative PPFD increased again and reached 70 and 50 % in early November (Seiwa and Kikuzawa, 1996). In each site, two microsites were established: one with leaf litter intact and one with leaf litter removed (to bare mineral soil, by removing the A0 layer). The predominant leaf litter components were leaves of Quercus and Sasa. Leaf litter accumulation was highest in FU and lowest in LG. Soil moisture in the top 4 cm of soil increased in the order LG (15 – 30 %), SG (21 – 49 %) and FU (30 – 58 %). Soil moisture was higher in microsites with litter than in those with bare soil throughout the growing season in both LG and SG, but little difference was observed in FU (Seiwa and Kikuzawa, 1996). A known number of seeds of each of the five study species (Quercus 50, Acer 100, Alnus 737, Cercidiphyllum 1530 and Betula 6278) were sown on each of six replicated quadrats (0.5  1 m) in mid-October 1987. Four replicates were used for measurements of seedling demography and height growth, and the other two replicates were used for harvesting, to determine the RGRM. Details of the experimental design are provided in Seiwa and Kikuzawa (1996). Measurements

Five seedlings of each species were selected randomly and harvested on 2 September 1988 in each of the two microsites at each site. Seedlings attacked by herbivores and/or pathogens were discarded, resulting in a total of five to ten seedlings sampled per species. At harvest the seedlings were removed from the ground by hand and cutlass, and care was taken to collect as many roots belonging to a plant as possible. Dry mass was determined from materials dried at 70 8C for 60 h. RGRM (g d21) between germination and harvest was calculated as follows: RGRM ¼ (ln SM – ln SR)/t, where SM is

539

seedling dry mass, SR is dry mass of seed reserve, and t is seedling age (days) at harvest. RGRH (cm d21) was calculated as follows: RGRH ¼ (ln H2 – ln H1)/t2 – t1, where H2 is seedling height at the end of growing season, H1 is initial seedling height at the time (t1) when cotyledons or first foliage leaf had completely unfolded for epigeous and hypogeous species, respectively, and t2 is seedling age (days) at the time when the seedling growth stopped (seedling had shed all leaves). RGRH was calculated for all the seedlings that survived to the end of the first growing season for Quercus (n ¼ 137), Acer (n ¼ 201), Alnus (n ¼ 92), Cercidiphyllum (n ¼ 133) and Betula (n ¼ 217). Although RGRH was calculated for all five species studied in each microsite at each site, RGRM was not calculated for Alnus in all three sites and for Cercidiphyllum in either microsite in LG, because of high mortality (severe damage). Statistical analyses

Differences in RGRM and RGRH among sites for each microsite and between microsites at each site were examined separately using Kruskal– Wallis tests for each species. The relationship between seed mass and RGRM, and between seed mass and RGRH, was analysed by using a single regression analysis for each microsite in each site. Spearman’s rank correlation coefficients were calculated between seedling survival (%) at the end of the first growing season in FU and both RGRM and RGRH in both SG and LG. The analysis was conducted separately for each microsite (bare soil and litter). R E S U LT S Environmental conditions, seed mass and RGR

For each species studied, RGRH significantly increased with increasing light in each microsite (Kruskal– Wallis tests, P , 0.0001; Fig. 1B). In each microsite, RGRM also showed significant increases with increasing light in Betula (P , 0.0001), Acer (P , 0.05) and Quercus (P , 0.05; Fig. 1A), but not in Cercidiphyllum (0.14 , P , 0.35), possibly because of small sample size (no data are available for FU). In most of the species, microsite conditions ( presence or absence of litter) had little effect on either RGRM or RGRH, regardless of light condition. Exceptions were Betula and Acer, both of which showed higher RGRM and RGRH in bare soil in SG, and higher RGRH in bare soil in FU (Kruskal – Wallis tests, P , 0.05; Fig. 1). In contrast, RGRH was higher in the microsite with litter than in the microsite without litter for Quercus (P , 0.05; Fig. 1B). Both RGRM and RGRH were negatively correlated with seed reserve in each microsite at each site (Fig. 2). An exception was bare soil in FU for RGRH. Interspecific trade-off between seedling survival and RGR

Seedling survival was lower for the fast-growing, early successional species, Betula (8.7 – 43.8 %) and Alnus

540

Seiwa — Seedling Growth and Survival in Temperate Trees A

Large gap

0·06 RGRM (g g–1 d–1)

Small gap 0·04

**

Forest understorey

*

0·02

*

0

RGRH (cm cm–1 d–1)

B 0·01

* **

0·005

* **

0

L B LB L B

**** LBLBLB

Betuala platyphylla Cercidiphylum var. japonica japonicum Low

LBLBLB

L B LB L B

LBLBLB

Alnus hirsuta

Acer mono

Quercus monogolica var. grosseserrata

Seed mass shade tolerance

High

F I G . 1. RGRM (A) and RGRH (B) in two microsites at three sites (mean + s.e.). Histograms and means are shown for seedlings in microsites with bare soil (B) or leaf litter (L) at large gap, small gap and forest understorey sites. Asterisks indicate significant differences between microsites. Each site was tested separately using a Kruskal–Wallis test. Significant differences are indicated as follows: * P , 0.05, ** P , 0.01, **** P , 0.0001.

(17.6 – 51.1 %), and the mid-successional species, Cercidiphyllum (4.0 – 50.5 %), than for the shade-tolerant, late successional species, Acer (58.1 –89.0 %) and Quercus (84.4 – 100 %), across the light gradient (Seiwa and Kikuzawa, 1996). As a result, seedling survival in FU was negatively correlated with both RGRM and RGRH in each microsite in both SG and LG (Spearman’s rank correlation; r2S . 0.81, P , 0.037; Fig. 3). An exception was the relationship between seedling survival in FU and RGRM in the microsite with litter in SG (r2S ¼ 0.64, P ¼ 0.20), possibly because of the small number of species observed (n ¼ 4). Relationship between seedling survival and RGR within a species

Along the environmental gradients, both RGRM and RGRH were negatively correlated with seedling survival in Acer (Fig. 4; r , 20.748, P , 0.05), and positively correlated in Betula (r . 0.755, P , 0.05). The other species showed no clear tendency. DISCUSSION RGR under different environmental conditions

In most of the species studied, both RGRM and RGRH increased as the amount of light at the site increased, regardless of their successional status. Such a lightdependent increase in RGRM has been observed for

several woody species in both tropical and temperate forests (e.g. Kitajima, 1994; Osunkoya et al., 1994; Veenendaal et al., 1996). Interestingly, Loach (1970) and Poorter (1999) reported that maximum RGRM was observed at intermediate light levels, particularly for late successional, shade-tolerant species. The discrepancy between these studies was probably because of differences in light intensity between shade-house experiments and natural conditions; the shade-house experiments usually included extremely high light levels compared with those in the natural habitats of the study species, particularly for shade-tolerant species. This study also revealed that fast-growing species in FU are also fast growers in SG and LG (Figs 1 and 2), suggesting that fast-growing, early successional species in temperate forests of Japan always grow more quickly than slow-growing, late successional species, regardless of light condition. Similar relationships have been observed for several tropical and temperate tree species (e.g. Popma and Bongers, 1988; Kitajima, 1994; Osunkoya et al., 1994; Veenendaal et al., 1996; Reich et al., 1998a; Poorter, 1999; Bloor and Grubb, 2003; cf. Grubb et al., 1996). In most of the species studied, the microsite condition ( presence or absence of litter) had little effect on the seedling RGRM and RGRH at most of the sites. However, the presence of litter had negative effects for both Betula and Acer in SG and FU, probably because of the greater initial seedling height required to penetrate the deep litter. This would reduce the RGR, particularly the RGRH. In contrast, presence of litter had a positive effect on the RGR in

Seiwa — Seedling Growth and Survival in Temperate Trees Large gap, litter Large gap, bare soil A

Small gap, litter 0·06

Small gap, bare soil Forest understorey, litter

0·05

Forest understorey, bare soil RGRM (g g–1 d–1)

0·04 0·03 0·02 0·01 0 –0·01 –0·02 0·1

1

10 100 Seed reserve (mg)

1000

B

1987; Walters and Reich, 2000; Dalling et al., 2002; Paz and Martinez-Ramos, 2003). In a comparative study of 31 hardwood species in a temperate forest of Japan (Seiwa and Kikuzawa, 1991), shorter leaf longevity and a higher leaf turnover rate were closely associated with small-seed mass, particularly under high-light conditions. Seedlings of small-seeded species attained maximum heights similar to those of large-seeded species at the end of growing season, by unfolding their leaves one by one for a longer period of time. In the large gap in this experiment, small-seeded species also attained maximum height similar to that of large-seeded species at the end of the second growing season, despite their initial disadvantage (Seiwa and Kikuzawa, 1996). In temperate forests, high resource levels favouring high RGRM and RGRH may be obligatory for seedlings of small-seeded tree species, to allow them to escape potentially high mortality rates that result, in part, from shading by neighbours, particularly in competitive early successional habitats. Performance trade-off and species coexistence

0·01

RGRH (cm cm–1 d–1)

541

0·005

0 0·1

1

10 100 Seed reserve (mg)

1000

F I G . 2. Relationship between seed reserve and RGRM (A) and RGRH (B) for four and five hardwood species, respectively, in two microsites at three sites.

large-seeded species (Quercus) in LG. Here, litter accumulation prevents the soil from severe drying in summer (Seiwa and Kikuzawa, 1996), resulting in enhanced seedling growth for species that are highly sensitive to water deficiency. Seed size and RGR

The small-seeded species tended to show greater RGRM and RGRH than the large-seeded species, which resulted in negative relationships between seed mass (dry mass of seed reserve) and RGR. Similar relationships have been observed in both herbaceous and woody plants (e.g. Tilman, 1988; Shipley and Peters, 1990; Westoby et al., 1992; Maranon and Grubb, 1993; Walters et al., 1993; Gleeson and Tilman, 1994; Kitajima, 1994; Osunkoya et al., 1994; Huante et al., 1995; Cornelissen et al., 1996; Grubb et al., 1996; Reich et al., 1998a, b; cf. Thompson,

This field study revealed that seedling survival in the forest understorey was negatively correlated with both RGRM and RGRH in both small and large gaps in a deciduous broad-leaved forest in a temperate region of Japan. This clear trade-off between low-light survival and highlight growth strongly supports the idea of light gradient partitioning (i.e. species coexistence) in spatially heterogeneous light environments (Kitajima, 1994; Kobe et al., 1995; Pacala et al., 1996; Kobe, 1999; Walters and Reich, 2000; Dalling and Hubbell, 2002; Baraloto et al., 2005; Sanchez-Gomez et al., 2006). This study further revealed that variation in RGRH among the study species was much greater in the large gap than in the small gap (Figs 1 and 3), resulting in a more obvious performance trade-off in the large gap. This suggests that a wider light gradient may lead to greater niche partitioning. In other words, large disturbances may be necessary to create sufficient variation in light availability, allowing many tree species to coexist as a result of resource partitioning. In shaded forest understorey, seedling persistence is strongly affected by the abundance and the activity of herbivores and pathogens, which are usually a major source of mortality (e.g. Seiwa, 1998; Nakashizuka, 2000). Thick, lignified leaves with a low specific leaf area (SLA) reduce the risk of mortality from herbivore damage or fungal infection in the large-seeded species (Quercus, Acer; K. Seiwa and K. Kikuzawa, unpubl. res.). Large-seeded and late successional species produce this type of leaf by using their large seed reserves, and the low leaf turnover (at the expense of a reduced potential growth) enhances seedling survival. In contrast, small-seeded, early successional species show rapid leaf turnover (Seiwa and Kikuzawa, 1996) and a high SLA (K. Seiwa and K. Kikuzawa, unpubl. res.). This enhances growth but lower investment in defence increases the risk of herbivore damage and fungal infection (Coley et al., 1985), resulting in lower survival under shaded conditions when compared with the greater growth ability seen in high-light

542

Seiwa — Seedling Growth and Survival in Temperate Trees Seedling survival in forest understorey (%)

A

B

100

100

Qm

Qm

Am 80

80

Am

Qm

Am 60

60

40

40

Cj Cj

Bp

Bp 20

20 Bp

Bp 0 –0·02

0

0·02

0·04

0 –0·02

0·06

RGRM in large gap (g g–1 day–1)

Seedling survival in forest understorey (%)

Am

0

0·02

0·04

0·06

RGRM in small gap (g g–1 day–1)

D

C 100

100

Qm Am

80

Qm Am Qm

80

Qm

Am

Am 60

Ah

40

60

Cj

Ah

Cj Cj

40

Cj Ah

Ah

Bp

Bp

20

20

Bp

Bp 0

0 0

0·005

0

0·01

RGRH in large gap (cm cm–1 d–1)

0·005

0·01

RGRH in small gap (cm cm–1 d–1)

F I G . 3. Relationship between seedling survival in forest understorey (FU) and RGRM in both large (A) and small gap (B), and between seedling survival in FU and RGRH in both large (C) and small gap (D) for four (RGRM) or five (RGRH) hardwood species in each microsite. Microsites with bare soil are indicated by open symbols while microsites with leaf litter are indicated by closed symbols. Species are abbreviated as follows: Qm, Quercus mongolica var. grosseserrata; Am, Acer mono; Cj, Cercidiphyllum japonicum; Ah, Alnus hirsuta; Bp, Betula platyphylla var. japonica.

A

B Quercus Acer Alnus Cercidiphylum Betula

Seedling survival (%)

100 80 60 40 20 0 –0·02

0

0·02 0·04 RGRM (g g–1 d–1)

0·06

0

0·005 0·01 RGRH (cm cm–1 d–1)

F I G . 4. Relationship between seedling survival and RGRM (A) and RGRH (B) for four and five hardwood species, respectively, under the six different environmental conditions (two microsites at three sites).

Seiwa — Seedling Growth and Survival in Temperate Trees environments. Dalling and Hubbell (2002) showed that a trade-off between growth and susceptibility to herbivores contributed to the coexistence of neotropical pioneer species with different light requirements. However, a recent study of neotropical tree species showed that variation in growth and mortality could not be attributed to differences in foliar herbivore damage (Pearson et al., 2003). Further empirical studies examining resource allocation for growth and defence in response to multiple limiting resources are needed to determine how these factors impact the coexistence of temperate tree species.

543

AC KN OW L E DG E M E N T S I thank Bill Shipley and two anonymous reviewers for comments that improved the quality of this paper. I also thank Kihachiro Kikuzawa, Tatsuhiro Asai, Norio Mizui, Naoto Ueno and Aya Imaji who contributed field work and improvement of the manuscript. This work was supported by a grant from the Ministry of Education, Science and Culture (No. 17380087, Seiwa).

L IT E RAT URE C IT E D Within-species variation along an environmental gradient

In Acer, both RGRM and RGRH increased with increasing the light availability but survival decreased along the same gradient, resulting in a negative correlation between survival and RGR along the light gradient. Higher seedling survival in the forest understorey was attributable to earlier germination, which enabled the seedlings to acquire abundant light in the spring prior to canopy closure and to avoid the attack of pathogens and herbivores (Seiwa and Kikuzawa, 1996; Seiwa, 1998). In contrast, a substantial number of the seedlings died in large gaps because of spring drying, although their growth was promoted by the abundant light (Seiwa and Kikuzawa, 1996; Seiwa, 1998), resulting in the trade-off. Such an intraspecific trade-off between survival and growth suggests that similar probabilities of seedling establishment exist in light-abundant large gaps and shaded forest understoreys. This idea is supported by previous observations of Acer in a wide range of habitats (e.g. Abe et al., 1995; Nagamatsu et al., 2002). The results of intraspecific trade-offs, however, seem to be inconsistent with the results of the interspecific analysis (Fig. 3) in which Acer was restricted to optimal performance within a narrow range of an environmental gradient (i.e. shaded forest understorey). Thus, in a forest community, the habitat width or conditions of tree species should be evaluated for both inter- and intraspecific performance trade-offs. In contrast, seedlings of Betula survived and grew better in more abundant light, suggesting that it is a typical pioneer species and may be exposed to strong selective pressures that result in adaptations under high-light conditions, such as large gaps. This result is in accordance with those of the interspecific analysis of performance trade-offs. For the other three species, Quercus, Alnus and Cercidiphyllum, there is no clear trend in the intraspecific relationships between growth and survival, probably because of the small sample sizes (no replicates). In conclusion, the results strongly support the idea of resource gradient partitioning (i.e. species coexistence) in spatially heterogeneous environments in temperate forests, and suggest that further species diversity would be promoted by increased spatial heterogeneity. The contrasts observed within species between two extremes of the resource continuum, as well as the contrast among species at the two extremes, are useful for evaluating habitat requirements.

Abe S, Masaki T, Nakashizuka T. 1995. Factors influencing sapling composition in canopy gaps of a temperate deciduous forest. Vegetatio 120: 21– 32. Baraloto C, Goldberg DE, Bonal D. 2005. Performance trade-offs among tropical tree seedlings in contrasting microhabitats. Ecology 86: 2461– 2472. Bazzaz FA. 1991. Habitat selection in plants. American Naturalist 137: S116–S130. Bloor JM, Grubb PJ. 2003. Growth and mortality in high and low light: trends among 15 shade-tolerant tropical rainforest tree species. Journal of Ecology 91: 77–85. Canham CD, Kobe RK, Latty EF, Chazdon RL. 1999. Interspecific variation in tree seedling survival: effects of allocation to roots versus carbohydrate resource. Oecologia 121: 1 –11. Chesson PL. 2000. Mechanisms of maintenance of species diversity. Annual Review of Ecology and Systematics 31: 343– 366. Coley PD, Bryant JP, Chapin FS. 1985. Resource availability and plant antiherbivore defense. Science 230: 895–899. Cornelissen JHC, Diez PC, Hunt R. 1996. Seedling growth, allocation and leaf attributes in a wide range of woody plant species and types. Journal of Ecology 84: 755–765. Dalling JW, Hubbell SP. 2002. Seed size, growth and gap microsite conditions as determinants of recruitment success for pioneer species. Journal of Ecology 90: 557–568. Dalling JW, Winter K, Hubble SP. 2002. Variation in growth responses of Neotropical pioneers to simulated forest gaps. Functional Ecology 18: 725–736. Davies SJ. 2001. Tree mortality and growth in 11 sympatric Macaranga species in Borneo. Ecology 82: 920– 932. Facelli JM, Pickett STA. 1991. Plant litter: its dynamics and effects on plant community structure. Botanical Review 57: 1– 32. Gilbert B, Wright SJ, Muller-Landau HC, Kitajima K, Hernandez A. 2006. Life history trade-offs in tropical trees and lianas. Ecology 87: 1281–1288. Givnish TJ. 1982. On the adaptive significance of leaf height in forest herbs. American Naturalist 120: 353–381. Gleeson SK, Tilman D. 1994. Plant allocation, growth rate and successional status. Functional Ecology 8: 543–550. Green PT, Juniper PA. 2004. Seed-seedling allometry in tropical rain forest trees: seed mass-related patterns of resource allocation and the ‘reserve effect.’ Journal of Ecology 92: 397– 408. Grime JP, Hunt R. 1975. Relative growth rate: its range and adaptive significance in a local flora. Journal of Ecology 63: 393–422. Grubb PJ, Lee WG, Kollmann J, Willson JB. 1996. Interaction of irradiance and soil nutrient supply on growth of seedlings of ten European tall-shrub species and Fagus sylvatica. Journal of Ecology 84: 827–840. Huante P, Rincon E, Acosta I. 1995. Nutrient availability and growth rate of 34 woody species from a tropical deciduous forest in Mexico. Functional Ecology 9: 849– 858. Hubbell SP, Foster RB. 1992. Short-term dynamics of a neotropical forest: why ecological research matters to tropical conservation and management. Oikos 63: 48–61. Kaerke CM, Kruger EL, Reich PB. 2001. Trade-offs in seedling survival, growth, and physiology among hardwood species of contrasting successional status along light-availability gradients. Canadian Journal of Forest Research 31: 1602–1616.

544

Seiwa — Seedling Growth and Survival in Temperate Trees

Kikuzawa K. 1983. Leaf survival of woody plants in deciduous broad-leaved forests. 1. Tall trees. Canadian Journal of Botany 61: 2133– 2139. Kitajima K. 1994. Relative importance of photosynthetic traits and allocation patterns as correlates of seedling shade tolerance of 13 tropical trees. Oecologia 98: 419– 428. Kobe RK. 1999. Light gradient partitioning among tropical tree species through differential seedling mortality and growth. Ecology 80: 187–201. Kobe RK, Pacala SW, Silander JA. 1995. Juvenile tree survivorship as a component of shade tolerance. Ecological Applications 5: 517–532. Koike T. 1988. Leaf structure and photosynthetic performance as related to the forest succession of deciduous broad-leaved trees. Plant Species Biology 3: 77–87. Loach K. 1970. Shade tolerance in tree seedlings. II. Growth analysis of plants raised under artificial shade. New Phytologist 69: 273–286. Maranon T, Grubb PJ. 1993. Physiological basis and ecological significance of the seed size and relative growth rate relationship in Mediterranean annuals. Functional Ecology 7: 591–599. Nagamatsu D, Seiwa K, Sakai A. 2002. Seedling establishment of deciduous trees in various topographic positions. Journal of Vegetation Science 13: 35–44. Nakashizuka T. 2000. Species coexistence in temperate, mixed deciduous forest. Trends in Ecology and Evolution 16: 205–210. Osunkoya OO, Ash JE, Hopkins MS, Graham AW. 1994. Influence of seed size and seedling ecological attributes on shade-tolerance of rain-forest tree species in northern Queensland. Journal of Ecology 82: 149 –163. Pacala SW, Canham CD, Saponara J, Silander JA, Kobe RK, Ribbens E. 1996. Forest models defined by field measurements: estimation, error analysis and dynamics. Ecological Monographs 66: 1– 43. Paz H, Martinez-Ramos M. 2003. Seed mass and seedling performance within eight species of Psychotri (Rubiaceae). Ecology 84: 439–450. Pearson TRH, Burslem DFRP, Goeriz RE, Dalling JW. 2003. Regeneration niche partitioning in Neotropical pioneers: effects of gap size, seasonal drought and herbivory on growth and survival. Oecologia 137: 456–465. Poorter L. 1999. Growth responses of 15 rain-forest tree species to a light gradient: the relative importance of morphological and physiological traits. Functional Ecology 13: 396– 410. Poorter L, Arets EJMM. 2003. Light environment and tree strategies in a Bolivian tropical moist forest: an evaluation of the light partitioning hypothesis. Plant Ecology 166: 295– 306. Poorter L, Rose S. 2005. Light-dependent changes in the relationship between seed mass and seedling traits: a meta-analysis for rain forest tree species. Oecologia 142: 378– 387. Popma J, Bongers F. 1988. The effects of canopy gaps on growth and morphology of seedlings of rain forest species. Oecologia 75: 625–632. Reich PB, Tjoelker MG, Walters MB, Vanderklein DW, Buschena C. 1998a. Close association of RGR, leaf and root morphology, seed mass and shade tolerance in seedlings of nine boreal tree species grown in high and low light. Functional Ecology 12: 237–338. Reich PB, Walters MB, Tjoelker MG, Vanderklein DW, Buschena C. 1998b. Photosynthesis and respiration rates depend on leaf and root morphology and nitrogen concentration in nine tree species with differing relative growth rate. Functional Ecology 12: 395–405.

Ross MA, Harper JL. 1972. Occupation of biological space during seedling establishment. Journal of Ecology 60: 70– 88. Sack L, Grubb PJ. 2001. Why do species of woody seedlings change rank in relative growth rate between low and high irradiance? Functional Ecology 15: 145– 154. Sanchez-Gomez D, Valladares F, Zavala MA. 2006. Performance of seedlings of Mediterranean woody species under experimental gradients of irradiance and water availability: trade-offs and evidence for niche differentiation. New Phytologist 170: 795 –806. Schreeg LA, Kobe RK, Walters MB. 2006. Tree seeding growth, survival, and morphology in response to landscape-level variation in soil resource availability in northern Michigan. Canadian Journal of Forest Research 35: 263–273. Shaw MW. 1968. Factors affecting the natural regeneration of sessile oak (Quercus petraea) in North Wales. II. Acorn losses and germination under field conditions. Journal of Ecology 56: 647–660. Seiwa K. 1998. Advantages of early germination for growth and survival of seedlings of Acer mono under different overstorey phenologies in deciduous broad-leaved forests. Journal of Ecology 86: 219–228. Seiwa K. 2000. Effects of seed size and emergence time on tree seedling establishment: importance of developmental constraints. Oecologia 123: 208–215. Seiwa K, Kikuzawa K. 1991. Phenology of tree seedlings in relation to seed size. Canadian Journal of Botany 69: 532– 538. Seiwa K, Kikuzawa K. 1996. Importance of seed size for the establishment of seedlings of five deciduous broad-leaved tree species. Vegetatio 123: 51– 64. Seiwa K, Kikuzawa K, Kadowaki T, Akasaka S, Ueno N. 2006. Shoot lifespan in relation to successional status in deciduous broad-leaved tree species in a temperate forest. New Phytologist 169: 537–548. Shipley B, Peters RH. 1990. The allometry of seed weight and seedling relative growth rate. Functional Ecology 4: 523– 529. Stamp N. 2003. Out of the quagmire of plant defense hypotheses. Quarterly Review of Biology 78: 23– 55. Swanborough P, Westoby M. 1996. Seedling relative growth rate and its components in relation to seed size: phylogenetically independent contrasts. Functional Ecology 10: 176–184. Thompson K. 1987. Seeds and seed bank. New Phytologist 106: 23– 34. Tilman D. 1988. Plant strategies and the dynamics and structure of plant communities. Princeton, NJ: Princeton University Press. Veenendaal EM, Swaine MD, Lecha RT, Walsh MF, Abebrese IK, Owusu-Afriyie K. 1996. Reponses of West African forest tree seedlings to irradiance and soil fertility. Functional Ecology 10: 501–511. Walters MB, Reich PB. 1996. Are shade tolerance, survival, and growth linked? Low light and nitrogen effects on hardwood seedlings. Ecology 77: 841–853. Walters MB, Reich PB. 2000. Seed size, nitrogen supply, and growth rate affect tree seedling survival in deep shade. Ecology 81: 1887– 1901. Walters MB, Kruger EL, Reich PB. 1993. Growth, biomass distribution and CO2 exchange of northern hardwood seedlings in high and low light: relationships with successional status and shade tolerance. Oecologia 94: 7 –16. Westoby M, Jurado E, Leishman M. 1992. Comparative evolutionary ecology of seed size. Trends in Ecology and Evolution 7: 368– 372. Wright SJ, Muller-Landau HC, Condit R, Hubbell SP. 2003. Gap-dependent recruitment, realized vital rates, and the size distribution of tropical trees. Ecology 84: 3174–3185.