Phenotypic variation in growth and biomass distribution for two advanced-1

generation pedigrees of hybrid poplar (Populus spp.) 2

3

Stan D. Wullschleger1, T.M. Yin2, S.P. DiFazio1, T.J. Tschaplinski1, L.E. Gunter1, M.F. 4

Davis3, and G.A. Tuskan1 5

6

1 Environmental Sciences Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 7

37831-6422, USA 8

2 The Key Laboratory of Forest Tree Genetic Engineering of the State Forest Administration, 9

Nanjing Forestry University, JS 210037, China 10

3 National Renewable Energy Laboratory, 1617 Cole Blvd., Golden, Colorado 80401, USA 11

12

13

Received __________ 2004 14

15

16

Address for correspondence: Stan D. Wullschleger 17

Environmental Sciences Division 18

Oak Ridge National Laboratory 19

Oak Ridge, Tennessee 37831-6422 20

Tel (865) 574-7839 21

Fax (865) 576-9939 22

E-mail wullschlegsd@ornl.gov 23

2

Abstract 1

Variation in the distribution of biomass among leaves, stems, and roots in forest trees is often 2

reported to be influenced by environment, nutrition, ontogeny or stage of development, and 3

genetics. However, although patterns of whole-tree distribution of dry mass are known to 4

vary in response to soil water and nitrogen availability, and as a function of developmental 5

stage, evidence that the above- and below-ground distribution of biomass varies as a function 6

of genetics is scarce. Therefore, to better assess the genetic control of biomass distribution in 7

trees, phenotypic variation in growth and patterns of biomass distribution to stems, branches, 8

leaves, coarse roots, and fine roots was examined in two advanced-generation hybrid poplar 9

(Populus trichocarpa x P. deltoides) families grown under field conditions during 2000-2002 10

in the Pacific Northwest. Family 331 was an inbred F2 (TD x TD) pedigree, whereas Family 11

13 was an outbred backcross BC1 (TD x D) pedigree. Fractional distribution of total biomass 12

(i.e., mass fraction) to shoots and roots averaged 0.62 ± 0.09 and 0.38 ± 0.09, respectively 13

across 247 genotypes in Family 331, and 0.57 ± 0.06 and 0.43 ± 0.06, respectively across 160 14

genotypes in Family 13. Phenotypic variation in the distribution of biomass to roots varied 15

considerably among genotypes, ranging from 0.13 to 0.64 for trees during their establishment 16

year. In contrast, fractional distribution of total biomass in two-year-old trees was 0.79 ± 0.04 17

to shoots and 0.21 ± 0.04 to roots, with little variation among genotypes. Allometric analysis 18

indicated that as trees increased in age, biomass was preferentially distributed to stems and 19

branches, whereas distribution of biomass to roots declined over time. Quantitative trait loci 20

(QTL) analysis for Family 13, based on a genetic linkage map consisting of 544 molecular 21

markers (SSR and AFLP) covering 19 chromosomes, indicated 31 QTL (LOD > 2.5) for all 22

traits measured. QTL were identified for shoot and root biomass, and for fractional biomass 23

3

distribution to leaves, stems, branches, cuttings, coarse roots, and fine roots. Six QTL for 1

biomass traits mapped to identical positions within the genome, whereas 10 QTL for 2

fractional distribution of biomass mapped to positions independent of biomass traits. This 3

suggests autonomous genetic control of allometric traits and raises the possibility that 4

distribution of dry mass to leaves, stems, branches, and roots can be independently 5

manipulated through selective breeding or genetic transformation. 6

7

Keywords: Above-ground biomass, below-ground biomass, carbon allocation, genetic map, 8

quantitative trait loci, QTL, root biomass, tree improvement. 9

10

Introduction 11

Trees, like all plants, demonstrate remarkable plasticity in the distribution of dry mass among 12

leaves, stems, and roots (Körner 1994, Poorter and Nagel 2000). Most often, this plasticity is 13

driven by resource availability, including light, soil water, and nutrients (Canham et al. 1996, 14

Ibrahim et al. 1997, Karin and Hawkins 1999, King et al. 1999, Kubiske et al. 1998, Retzlaff 15

et al. 2001, Wang et al. 1998). Such phenotypic plasticity, whether in the form of increased 16

leaf mass produced in response to low light or increased root proliferation in response to 17

drought, may promote the competitive ability of a given genotype over a range of resource 18

availabilities, enabling such plants to capture those resources that most strongly limit growth 19

and development (Aerts et al. 1991, Reynolds and D’Antonio 1996). 20

Although studied in less detail, genetics also plays a role in determining the distribution of 21

dry mass to leaves, stems, and roots (Ericsson et al. 1996, Zobel 1996). Provenance trials 22

(Oleksyn et al. 1998, Retzlaff et al. 2001, Wang et al. 1998) and studies with progeny from 23

4

seed of known parentage (Clark and Schlarbaum 2003, Li et al. 1991, McMillin and Wagner 1

1995) provide data on the above- and below-ground distribution of biomass for conifers and 2

hardwoods that can be interpreted in a genetic or ontogenetic context. In Populus – a genus 3

containing more than 30 species worldwide – a number of studies have reported significant 4

clonal variation in biomass distribution (Dickson et al. 1998, Heilman et al. 1994, Karim and 5

Hawkins 1999, Pregitzer et al. 1990, Rhodenbaugh and Pallardy 1993, Scarascia-Mugnozza et 6

al. 1997), providing circumstantial evidence that a genetic basis for above- and below-ground 7

dry mass distribution does exist. 8

One of the classic studies in this regard compared patterns of carbon allocation and 9

biomass distribution for two Populus clones with contrasting morphology and phenology 10

(Isebrands and Nelson 1983, Michael et al. 1988, Pregitzer et al. 1990). These studies 11

demonstrated that Populus x euramericana cv. ‘Eugenei’ consistently allocated more carbon 12

to leaves, branches, and stems, and less to roots than did the clone P. tristis x P. balsamifera 13

cv. ‘Tristis’ (Isebrands and Nelson 1983, Pregitzer et al. 1990). Preferential distribution of 14

biomass to shoots at the expense of roots, while increasing harvest index, predisposed 15

‘Eugenei’ to water stress (Michael et al. 1988). Although these and other early studies helped 16

characterize morphological plasticity in hybrid poplars and assisted in developing the 17

ideotype concept for tree breeders (Dickmann and Keathley 1996), most studies have 18

evaluated phenotypic variation in biomass distribution for only a small number of Populus 19

clones. As a result, inferences are limited and there is a need to examine growth and biomass 20

distribution using a larger number of clones, preferably from structured pedigrees (Stettler and 21

Bradshaw 1994), focusing not only on phenotypic variation among clones, but also on 22

5

identifying loci responsible for quantitative traits such as the total and fractional distribution 1

of dry mass to leaves, stems, and roots (Poorter and Nagel 2000, Tuberosa et al. 2002). 2

Interspecific hybridization and quantitative trait loci (QTL) analysis are techniques used 3

widely in Populus to better understand the genetic basis of complex phenotypes (Frewen et al. 4

2000, Bradshaw and Stettler 1995, Newcombe and Bradshaw 1996, Stirling et al. 2001, Wu 5

2000, Wu et al. 1998). Advanced-generation hybridization offers the opportunity to combine 6

traits not expressed in either parental species through segregation and recombination of genes 7

(Karim and Hawkins 1999). QTL analysis, along with the recent sequencing of the Populus 8

genome (Tuskan et al. 2004, Wullschleger et al. 2002), provides a means not only to verify a 9

genetic basis for biomass distribution in this genus, but also allows initial steps to be taken 10

towards identifying the genes responsible for such traits. In this paper, we report results from 11

a three-year investigation during which growth and biomass distribution to stems, branches, 12

leaves, coarse roots, and fine roots were determined for one- and two-year-old hybrid poplar 13

trees. Since our primary goal was to identify QTL associated with above- and below-ground 14

traits, two advanced-generation pedigrees (inbred F2 and outbred BC1) suitable for use in QTL 15

mapping were established from cuttings in the Pacific Northwest. Specific objectives were to 16

(1) characterize the extent of phenotypic variation in growth and biomass distribution for one- 17

and two-year-old hybrid poplar pedigrees, (2) derive allometric relationships among 18

individual plant components and total tree biomass, and (3) identify QTL associated with the 19

fractional distribution of biomass to leaves, stems, branches, and roots as a means to revealing 20

a genetic basis for growth and biomass distribution in young hybrid poplar trees during the 21

initial two years of stand establishment. 22

23

6

Material and Methods 1

Site description and establishment of field plantings 2

Field plantings were established in 2000 and 2001 at a site near Wallula, WA. Trees were 3

grown under drip irrigation and received ca. 360 mm of supplemental irrigation per year. 4

Irrigation lines were placed within rows with an emitter at each planting location. Mean 5

annual precipitation at the study site is 495 mm. All trees were fertilized with an operational 6

prescription of macro- and micro-nutrient solution that was routinely injected into the 7

irrigation lines. Trees received ca. 80 kg ha-1 nitrogen per year. Soil classification is 8

predominantly active dune and is associated with the Quincy soil series, consisting of 9

excessively drained and moderately drained, coarse-textured soils derived from old alluvium 10

and windblown sand (USDA 1971). Mean annual temperature is 12.1 oC. 11

Two populations of hybrid poplars (Populus trichocarpa Torr. & Gray x P. deltoids Bartr.) 12

were used to study phenotypic variation in growth and biomass distribution. Family 331 is an 13

inbred F2 (TD x TD) pedigree (Bradshaw and Stettler 1993), whereas Family 13 is an outbred 14

backcross BC1 (TD x D) pedigree (Yin et al. 2004). Dormant cuttings for these two families 15

were collected from either 8-year-old trees near Clatskanie, OR (Family 331) in late 1999 or 16

from 5-year-old trees outside Thief River Falls, MN (Family 13) in late 2000. Eight cuttings 17

were collected from each of 328 genotypes in Family 331 and 171 genotypes in Family 13. 18

Cuttings were 15 to 20 cm in length with each having 4 to 5 dormant buds. All cuttings were 19

stored in a cold-room and planted the following spring in four replicated blocks. Cuttings 20

were planted in three blocks at a 1.1 x 3 m spacing (ca. 3050 plants ha-1) and in one block at a 21

2.2 x 3 m spacing (ca. 1525 plants ha-1). Two cuttings were planted at each position. In the 22

case that both cuttings survived and produced sprouts, one was randomly removed early in the 23

7

season. Plants from the narrow and wide spacing were harvested as one- and two-year-old 1

trees, respectively. Genotypes from Family 331 were harvested after one growing season 2

(2000), whereas genotypes from Family 13 were harvested after one (2001) and two (2002) 3

seasons of growth. 4

5

Determination of growth and biomass distribution 6

Heights and diameters were measured at final harvest in November of each year. In the case 7

of one-year-old trees, measurements of height and diameter were made on intact trees in the 8

field. In the case of two-year-old trees, diameters were measured in the field, whereas 9

measurements of height were made after trees had been severed at ground level and moved to 10

a central processing area. Stem diameters were measured with a digital caliper (Brown and 11

Sharpe, North Kingstown, RI). 12

Harvesting the above-ground portion of the plants entailed cutting the stem at ground level 13

and separation into stems, branches (if there were any), and leaves. Below-ground portions of 14

each tree, including the original cutting, coarse roots (>2 mm), and fine roots (<2 mm) were 15

carefully excavated and sampled by hand. Roots were removed from a conical volume of soil 16

having a radius of 0.5 m centered on the tree stem and nominally 0.5 m in depth. Field 17

methods largely followed those used by Friend et al. (1991). A hole this size was more than 18

adequate, allowing access to coarse and fine roots in one-year-old trees. Coarse roots and fine 19

roots were recovered at the time of excavation by carefully sieving soils through 3 x 3 mm 20

mesh screens. In the case of two-year-old trees, the size of the hole was sufficient to enable 21

large coarse roots to be identified. These roots were completely extracted from the soil, 22

which in some instances required following roots for several meters both laterally and 23

8

horizontally until root diameter was less than 1 cm. Below-ground biomass resulting from the 1

original cutting was separated from other below-ground fractions and then processed along 2

with the rest of the sampled and subsampled biomass. 3

Fresh mass of all plant components was determined in the field with a battery-operated 4

digital balance (Intercomp, Inc., Minneapolis, MN). A random subsample of stems, branches, 5

leaves, coarse roots, and fine roots was taken, transported to Oak Ridge National Laboratory 6

via overnight express, dried in forced-ventilation ovens at 70oC for at least three days, and 7

then weighed. Dry mass-to-fresh mass ratios for each sample were used to calculate the total 8

dry mass of each component (Tuskan and Rensema 1992). The relative distribution of 9

biomass to leaves, stems, branches, cutting, coarse roots, and fine roots was estimated as a 10

fraction of total dry mass for each tree. 11

12

Allometric analysis 13

Since quantitative metrics that describe the relationship of one plant component to another 14

(i.e., mass fraction and root/shoot ratio) are subject to ontogenetic and environmental 15

influences, biomass data were also analyzed using the equation, 16

log (Y) = log (β) + α log (X) (1) 17

where X and Y are inter-dependent variables of plant biomass, and the parameters β and α are 18

regression coefficients (Niklas and Enquist 2002). Estimates of β and α obtained in this 19

manner could then be inserted into an allometric equation of the form, 20

Y = β X α (2) 21

9

where β is the allometric constant and α is the exponential or scaling coefficient. All 1

allometric relationships involving Eq. 1 were analyzed using Model Type I least-square 2

regressions fitted to log-log transformed data (Systat Software, Inc., Point Richmond, CA). 3

4

Statistical analysis 5

All growth and biomass distribution parameters were analyzed statistically by analysis of 6

variance (ANOVA) to test for significant (P ≤ 0.05) differences among genotypes. Analyses 7

were restricted to one-year-old trees in Family 331 and 13, since genotypes were not 8

replicated in two-year-old trees. Potential shifts in the distribution of total biomass among 9

leaves, stems, branches, and roots due to either pedigree or age were identified by testing for 10

significantly different slopes in the relationship between whole-tree biomass and each plant 11

component. In addition to testing allometric relationships for leaves, stems, branches, and 12

roots plotted against tree biomass, correlation matrices among all plant components were also 13

examined. Box plots were used to graphically depict variation among genotypes for total and 14

fractional distribution of biomass by percentiles. Lower and the upper portions of the box 15

indicate the 25th and 75th percentiles, respectively. The values of the bars are the 5th and 95th 16

percentile, and the 50th percentile (median) is given by the horizontal line within the box. 17

18

Construction of a genetic linkage map for Family 13 19

Ninety-two microsatellite (SSR) and 24 AFLP primer pairs generated 556 markers that were 20

used to construct a genetic linkage map from 171 genotypes in Family 13. GenScan and 21

Genotyper software (Applied Biosystems, Foster City, CA) were used to extract data and 22

score the electrophoragrams, and further processed by PERL scripts. Markers were assigned 23

10

to linkage groups at a minimum threshold of LOD 10.0 and a maximum recombination 1

fraction of 0.30. The resulting linkage map includes 544 molecular markers (SSR and AFLP) 2

mapped onto 19 linkage groups, equivalent to the Populus chromosome number, with all 3

markers displaying internally consistent linkage patterns. Complete details of genetic map 4

construction can be found in Yin et al. (2004). 5

6

QTL analysis 7

Phenotypic data were plotted by QTL Cartographer Windows v2.0. The normality of the trait 8

data was tested by the statistic S, 9

S =246

24

23 nknk+ , (3) 10

where 3k is the coefficient of skewness and 4k is the coefficient of the kurtosis. The S 11

statistic is distributed as a 2χ value with two degrees of freedom and the critical value for 12

rejection of normality is 5.99 at the 5% level. In the case that raw data for a particular trait 13

did not follow a normal distribution, a log-transformation was performed. For traits that 14

could not be normalized in this manner, we removed outlining data points (outside σµ 2± ) 15

and retested normality. QTL analysis was performed using Mapmaker/QTL 1.1 software 16

(Lincoln and Lander 1992), setting a LOD score threshold of 2.5 for accepting the presence of 17

a QTL in the mapped interval (Paterson et al. 1988). Detected QTL were then examined 18

using QTL Cartographer to test for the presence of multiple QTL. Empirical statistical 19

significance thresholds were set by permutations at significance level P ≤ 0.05 repeated 300 20

times (Churchill and Doerge 1994). The QTL effect was expressed as the percentage of 21

phenotypic variation explained (PVE) as provided by Mapmaker/QTL 1.1. Only additive 22

11

effects of the detected QTL are reported due to the nature of the Family 13 pedigree (i.e., a 1

‘Pseudotestcross’ from a backcross pedigree). 2

3

Results 4

Phenotypic analysis 5

Frequency distributions showed considerable variation among genotypes for total biomass 6

and the distribution of biomass to shoots and roots (Fig. 1). The magnitude of this variation 7

was similar between Family 331 and 13, with approximately a 55-fold difference between the 8

smallest and largest genotypes within a pedigree. The number of genotypes in a particular 9

size class (i.e., binned according to dry mass), however, was skewed in Family 331, whereas 10

total, shoot, and root biomass in Family 13 were more normally distributed. One-year-old 11

trees with a total biomass < 200 g dry mass represented 65% of all trees harvested in Family 12

331, but only 20% of the genotypes harvested in Family 13 (Fig. 1). 13

Total biomass averaged (± s.d.) across one-year-old genotypes was 197 ± 167 g dry mass 14

in Family 331 and 367 ± 224 g dry mass in Family 13 (Fig. 2). Shoot and root biomass were 15

similarly larger in Family 13. Root-to-shoot ratio averaged 0.65 ± 0.27 in Family 331 and 16

0.77 ± 0.19 in Family 13. Fractional distribution of total biomass (i.e., mass fraction) to 17

shoots and roots averaged 0.62 ± 0.09 and 0.38 ± 0.09, respectively in Family 331, and 0.57 ± 18

0.06 and 0.43 ± 0.06, respectively in Family 13 (Fig. 3). Phenotypic variation in the 19

fractional distribution of biomass to roots was considerable, ranging from 0.13 to 0.64 for 20

trees during their establishment year. Total biomass distributed to above-ground components 21

averaged across the two pedigrees (i.e., 406 genotypes) was 0.29 ± 0.07 for leaves, 0.21 ± 22

0.06 for stems, and 0.10 ± 0.07 for branches. Likewise, distribution of whole-tree biomass to 23

12

below-ground components averaged 0.20 ± 0.08 for coarse roots, 0.17 ± 0.07 for cuttings, and 1

0.03 ± 0.03 for fine roots (Fig. 3). 2

Strength of correlations among the various above- and below-ground components of total 3

biomass was variable in both Family 331 and 13 (Table 1). Nonetheless, all correlations were 4

significant. The strongest correlations were among leaves, stems, branches, cuttings, and 5

coarse roots. Correlations among these individual plant components and fine roots, however, 6

while still significant, were noticeably weaker in both pedigrees (Table 1). Allometric 7

analysis of the log-log transformed data indicated significant relationships between shoot, 8

root, and all plant components when plotted against total biomass (Fig. 4). There were no 9

differences between Family 331 and 13 in either the slope or intercept of these relationships. 10

The most significant regression model was between shoot and total biomass, whereas the least 11

significant model was between fine roots and total biomass (Table 2). 12

Two-year-old genotypes of Family 13 had an average total, shoot, and root biomass of 13

3455 ± 2115, 2759 ± 1720, and 695 ± 431 g dry mass, respectively (Fig. 5a), representing an 14

9.4, 12.8, and 4.6-fold increase in total and distributed biomass between first- and second-year 15

growth. Phenotypic variation was large for total and shoot biomass, but remarkably small for 16

roots. Root-to-shoot ratio averaged 0.27 ± 0.07 across genotypes. In contrast to one-year-old 17

genotypes of this pedigree, the distribution of shoot biomass for two-year-old trees was stems 18

> leaves > branches (Fig. 5b). Fractional distribution of total biomass was 0.79 ± 0.04 to 19

shoots and 0.21 ± 0.04 to roots, again with little variation among genotypes (Fig. 5c). Total 20

biomass was distributed 0.24 ± 0.03 to leaves, 0.35 ± 0.07 to stems, 0.20 ± 0.05 to branches, 21

0.09 ± 0.02 to cuttings, and 0.12 ± 0.04 to coarse roots (Fig 5d). Fine roots were not 22

harvested in two-year-old genotypes. 23

13

Correlations among all individual plant components harvested were strong and significant 1

for two-year-old genotypes (Table 3). The strongest correlations were between leaves and 2

branches, followed by leaves and coarse roots. Allometric analysis also indicated significant 3

relationships between shoot, root, and all plant components when individually plotted against 4

total biomass (Table 4). 5

6

QTL detection 7

All growth and biomass distribution traits in Family 13 were associated with at least one 8

QTL, with above-ground and below-ground having similar numbers of QTL (Table 5). There 9

were 14 and 17 QTL detected for one- and two-year-old trees, respectively. Generally, these 10

QTL mapped to alternate positions within the genome, with the exception that a single QTL 11

for fractional stem biomass occurred in roughly the same position on chromosome VI in both 12

the first- and second-year datasets. In addition, 6 QTL for multiple biomass traits mapped to 13

identical positions within the genome in three cases, e.g., total, shoot, root, cutting, coarse 14

root, and branch biomass traits for two-year-old trees mapped to the same position on 15

chromosome XIII (Fig. 6). Total and fractional biomass QTL for one-year-old trees occurred 16

solely on chromosomes I, II, III, IV, and VII; QTL for two-year-old trees occurred solely on 17

chromosomes XII, XIII, and XIV; chromosome VI shared QTL for both first- and second-18

year traits. LOD scores for significant QTL ranged from 2.52 to 8.40 and averaged 3.9 (Table 19

5). The percent phenotypic variance explained (PVE) by any single QTL ranged from 7.5% 20

to 18.3% and averaged 11.2%. Percent PVE did not vary substantially between QTL for 21

either one- or two-year-old trees (12.2 and 10.3% average, respectively). The allele with a 22

positive phenotypic effect came from the P. deltoides grandparent for 23 out of the 31 QTL 23

14

detected (Table 5). Ten QTL involving biomass distribution mapped to positions independent 1

of biomass traits (Fig. 6). 2

3

Discussion 4

Field-grown hybrid poplars in this study exhibited patterns of growth and biomass distribution 5

similar to that reported in earlier investigations (Michael et al. 1988, Ceulemans et al. 1996, 6

Scarascia-Mugnozza et al. 1997). Total biomass at the end of the establishment year was 7

largely distributed to shoots, consistent with a root/shoot ratio that averaged 0.65 and 0.77 in 8

Family 331 and 13, respectively. Previous studies report similar root/shoot ratios for one-9

year-old hybrid poplars (Michael et al. 1988, Ibrahim et al. 1997, Pregitzer et al. 1990, 10

Dickson et al. 1998), and several document that the distribution of biomass to shoots and roots 11

varied among clones (Heilman et al. 1994, Dickson et al. 1998, Scarascia-Mugnozza et al. 12

1997, Karim and Hawkins 1999). Most studies, however, have evaluated variation in dry 13

mass distribution for a relatively small number of clones or genotypes. As a result, inferences 14

concerning the genetic control of biomass distribution in hybrid poplar have been limited. 15

Variation among several hundred full-sib genotypes for above- and below-ground biomass 16

distribution, as measured in our study, was considerable. Root/shoot ratios ranged from 0.23 17

to 1.92 in Family 331 and from 0.24 to 1.25 in Family 13. Interestingly, approximately 9 to 18

14% of the one-year-old genotypes harvested in each family distributed biomass equally to 19

roots and shoots (i.e., root/shoot ratio ≥ 1.0), suggesting that selection for increased 20

distribution of biomass to roots is feasible. Most of these genotypes were small and results 21

could reflect artifacts associated with low rates of biomass production. However, a few – 22

notably clones 1093, 1732, 1878, and 1706 in Family 331 and clones 827, 695, 609, and 659 23

15

in Family 13 – were large trees when harvested at the end of the establishment year. Thus, 1

while these root/shoot ratios are high, they nonetheless reflect considerable phenotypic 2

plasticity in terms of the fractional distribution of biomass to roots, achieving above average 3

stem growth while maintaining above average root growth. 4

Above-ground biomass in one-year-old trees was distributed to leaves > stems > branches, 5

whereas first-year below-ground biomass was distributed to coarse roots > cutting > fine 6

roots. This pattern of biomass distribution was consistent across both Family 331 and Family 7

13, and with few exceptions, across genotypes. Not all genotypes, however, had branches. In 8

our study, 25 out of 246 one-year-old genotypes in Family 331 lacked branches, while 25 out 9

of 160 genotypes in Family 13 had no branches. Genotypes that lacked branches had less 10

above-ground (P < 0.001) and below-ground (P < 0.001) biomass than did those genotypes 11

that had branches during their establishment year. Although genotypes that lacked branches 12

were smaller, they tended to distribute a greater proportion of total biomass to roots (i.e., 0.47 13

vs. 0.40). Scarascia-Mugnozza et al. (1997) also reported differences in whole-tree patterns 14

of biomass distribution among branchy and less branchy hybrid poplar clones, with less 15

branchy clones having a greater proportion of total dry mass distributed to below-ground 16

components. Differences in total, shoot, and root biomass between groups were less in two-17

year-old genotypes. Still, the fractional distribution of biomass to roots remained slightly 18

higher in the less branchy compared to the branchy clones (i.e., 0.23 vs. 0.20; P = 0.083). It is 19

not known what proportion of total branch mass was contributed by proleptic versus sylleptic 20

branches in two-year-old genotypes. Given that sylleptic branches and leaves associated with 21

them play an important role in determining patterns of carbon allocation, canopy architecture, 22

and ultimately rates of tree productivity (Ceulemans and Isebrands 1996), quantitative 23

16

distinctions between these two branch types either during the season or at the time of final 1

harvest would have been informative. 2

Previous reports for two-year-old hybrid poplars indicate that total biomass is distributed 3

among plant components in a manner similar to what we observed in this study: stems 0.30 to 4

0.50, branches 0.13 to 0.20, leaves 0.12 to 0.27, above-ground biomass 0.62 to 0.81, and 5

below-ground biomass 0.19 to 0.38 (Horwath et al. 1994, Scarascia-Mugnozza et al. 1997, 6

Friend et al. 1991). Comparison of these estimates with values derived for one-year-old 7

genotypes reveal marked age-related changes in patterns of biomass distribution. King et al. 8

(1999) point out that as plants increase in size and age, the relative distribution of biomass 9

between organs frequently changes. Evidence for this is seen in early studies by Ovington 10

(1957) and in the more recent data compilations of Körner (1994) and Enquist and Niklas 11

(2002). Heilman et al. (1994) observed that root/shoot ratios (leafless biomass) among T x D 12

hybrids and their parental species was much less for 4-year-old plants than for plants during 13

their establishment year. In a study of four poplar clones, Scarascia-Mugnozza et al. (1997) 14

reported that compared to the establishment year, there was a relative decrease in below-15

ground components during the second growing season, and an increase in stem, branches, and 16

leaves for all clones. A similar change in biomass distribution was observed in our study for 17

one- and two-year-old genotypes, expressed predominantly as an increase in biomass 18

distribution to stems and branches from one year to the next. While total biomass for Family 19

13 increased almost 10-fold during the second growing season, there was less than a 5-fold 20

increase in root biomass, but more than a 12-fold increase in shoot biomass. Allometric 21

analysis indicated that as the average genotype increased in age there was a marked increase 22

in the fractional distribution of biomass to stems and branches, and a general decrease in 23

17

biomass distributed to leaves, cutting, and coarse roots. While multiple studies have 1

examined fractional distribution of biomass in hybrid poplar, few (if any) studies have used 2

allometric analysis to examine variation in biomass distribution among individual plant 3

components. This is despite recommendations that such analyses would help resolve and 4

otherwise avoid size-related complications in interpreting biomass distribution due to 5

environmental and/or ontogenetic effects (King et al. 1999). 6

The parental species used in this study, P. trichocarpa and P. deltoides, and their F1 7

hybrids, are known to differ substantially in many morphological, phenological, anatomical, 8

and physiological traits (Tschaplinski et al. 1998). Such variation is well documented for F1 9

progeny and is probably related to the large genetic differences of the two parental species 10

(Stettler et al. 1996). Our interest in using Family 331 and 13 was in characterizing the extent 11

of phenotypic variation for advanced-generation (F2 and BC1) plant materials. In a study 12

designed to examine nitrogen and phosphorus requirements of a three-generation Populus 13

pedigree, Karim and Hawkins (1999) reported that compared to parental species and F1 14

clones, the F2 generation had the poorest growth performance, on average, but exhibited 15

extensive variability in all traits examined. These authors characterized patterns of biomass 16

distribution from which we could calculate the fractional distribution of total biomass to roots 17

(i.e., root mass fraction). For 29 F2 clones (TD x TD) grown in the greenhouse, root mass 18

fraction averaged 0.24 ± 0.06, with variation among clones ranging from 0.10 to 0.39 19

(calculated from their Table 4). Similar variation in root mass fraction was observed for one-20

year-old genotypes from Family 331 (0.18 to 0.66) and from Family 13 (0.24 to 0.56), and for 21

two-year-old genotypes from Family 13 (0.11 to 0.32) in our field investigation. Moreover, 22

there was no indication in either our study or previous reports that variation in the distribution 23

18

of biomass to roots was negatively correlated with whole-tree productivity. Thus, within the 1

extent of phenotypic variation observed in our investigation, increased distribution of biomass 2

to roots in one- and two-year-old trees occurred without apparently compromising above-3

ground biomass production. 4

QTL have been identified in forest tree species for a wide variety of traits including growth 5

(Wu et al. 1998), chemical, and physical properties of wood (Sewell et al. 2000, Sewell et al. 6

2002, Verhaegen et al. 1997, Grattapaglia et al. 1996), vegetative propagation capacity 7

(Grattapaglia et al. 1994, Marques et al. 1999), phenology (Bradshaw and Stettler 1995, 8

Frewen et al. 2000), and tree architecture (Wu 1998). Our study extends this body of 9

evidence to include total and fractional biomass distribution traits. These results suggest that 10

regions of the genome associated with genetic control of the biomass traits of one- and two-11

year-old trees are for the most part independent of one another, which is consistent with 12

observations of QTL across years in other studies in Populus (Bradshaw and Stettler 1995, 13

Wu et al. 1998), Salix (Tsarouhas et al. 2002), and Eucalyptus (Verhaegen et al. 1997). This 14

may reflect ontogenetic variation in biomass accumulation (Wu et al. 2003), as demonstrated 15

by marked differences in biomass distribution between one- and two-year-old trees (i.e., 16

root/shoot ratios). Alternatively, first-year QTL may reflect establishment effects such as 17

differential rooting ability among progeny, while second-year QTL may be more directly 18

associated with long-term productivity (Wu et al. 1998). 19

The co-location of multiple QTL was not surprising given that many of the individual 20

components of plant biomass are highly inter-correlated and allometric relationships are 21

relatively invariable within species (Niklas and Enquist 2001, Wu et al. 2003). However, in 22

some cases traits were associated that had no allometric relationship (e.g., QTL for multiple 23

19

biomass traits were co-located with a QTL for fractional distribution of below-ground 1

biomass for Year 1 on linkage group IV, and QTL for percentage fine roots was co-located 2

with a QTL for stem biomass on linkage group III; opposite genetic effects). This suggests 3

that there may be a single gene or regulatory element that has broad pleiotrophic effect on 4

multiple biomass traits. Alternatively, multiple genes of different effect may be located closer 5

than the resolution of our QTL mapping, which may span several megabases of sequence and 6

encompass dozens of candidate genes (DiFazio et al., manuscript in preparation). Finally, 7

there may be a functional relationship between traits that are co-located (e.g., genotypes with 8

greater distribution of biomass to roots in Year 1 may grow better due to superior capture of 9

resources in Year 2). We are unable to differentiate these possibilities in the current study, 10

but with availability of the entire Populus genome, including the annotated gene models 11

associated with our genetic map, it will be possible to test candidate genes from within these 12

flanked regions to determine the genetic basis for multiple traits mapping to the same position 13

within the genome. 14

One of the more noteworthy findings of our study was the independent location of biomass 15

trait QTL and fractional biomass trait QTL. Wu et al. (2002) previously demonstrated that 16

allometric variation in Family 331 was under genetic control. However, to our knowledge, 17

our study with Family 13 is the first time that independent genetic control of above- and 18

below-ground biomass traits has been demonstrated in a forest tree. Therefore, if proven true, 19

it should be possible to independently modify distribution of biomass to individual plant 20

components. Confirmation of this will require additional independent assessments of QTL 21

effects to validate both QTL position and magnitude, as well as to expand the environmental 22

and genetic backgrounds under which our estimates were determined. 23

20

Conclusion 1

Considerable variation was observed for the distribution of whole-tree biomass to leaves, 2

stems, branches, and roots in two advanced-generation hybrid poplar pedigrees. We have 3

shown that biomass distribution is not fixed, but varies with genotype and age. While similar 4

results have been reported in the past, our study is unique because of the number of genotypes 5

investigated and the use of these data to identify QTL associated with whole-tree biomass and 6

the relative distribution of total biomass to individual plant components. Our studies have 7

shown that extensive genetic variability does exist in biomass distribution among F2 hybrids 8

and that there is a genetic basis for this variation. Thus, phenotypic variation as observed in 9

our study, especially as it relates to biomass distribution to roots, is an asset for future tree 10

improvement programs. Interestingly, Heilman et al. (1994) point out that “…for practical 11

reasons, the process of clonal selection [in poplar] does not usually involve collection of data 12

on root systems”, with tree plant breeders and geneticists focusing almost exclusively on traits 13

observable above-ground. While such a statement is undoubtedly true and understandable, 14

the importance of root growth to carbon sequestration, phytoremediation, and to an improved 15

understanding of basic tree biology (Pregitzer 2002), will require a focused and informed 16

approach to investigating root-shoot relationships. Studies should seek to explore these 17

relationships for trees encompassing a range of ages, taking advantage of existing field 18

plantings and the availability of new technologies (Nadezhdina and Cermak 2002) for 19

quantifying variation in biomass distribution for trees grown under a range of conditions. In 20

summary, our results suggest that (1) extensive phenotypic variation does exist among 21

advanced generation hybrid poplar progeny for both growth and biomass distribution traits, 22

(2) there are positive allometric relationships among biomass traits, and (3) QTL exist for all 23

21

measured biomass traits, indicating coordinated as well as independent genetic control of 1

growth and biomass distribution. 2

3

Acknowledgments 4

Special thanks to Peggy Payne, Larry Miller, Tom Agens, Evan Purser, Robin Rayborn, 5

Andrea Silletti, Steve Thomas, Joanne Ledford, Ramie Wilkerson, Kenneth Lowe, Kim 6

Mangrini, and others who provided assistance with plantation establishment, silvicultural 7

management, field harvests, and processing of samples. Our appreciation is extended to Dr. 8

H.D. (Toby) Bradshaw at the University of Washington, Dr. Brian Stanton at GreenWood 9

Resources, and Dr. Jon Johnson at Washington State University for providing us with clonal 10

material for Family 331 experiments. Dr. Don Riemenschneider, U.S. Forest Service, kindly 11

provided clonal material for Family 13. We acknowledge support from the U.S. Department 12

of Energy, Office of Science, Biological and Environmental Research Program. This project 13

was supported under the Carbon Sequestration Program. Oak Ridge National Laboratory is 14

managed by UT-Battelle, LLC, for the U.S. Department of Energy under contract No. DE-15

AC05-00OR22725. 16

17

18

19

20

21

22

23

22

References 1

Aerts, R., Boot, R.G.A., and van der Aart, P.J.M. 1991. The relationship between above- and 2

belowground biomass allocation patterns and competitive ability. Oecologia 87: 551-559 3

Bradshaw, H.D., and Stettler, R.F. 1993. Molecular genetics of growth and development in 4

Populus. I. Triploidy in hybrid poplars. Theor. Appl. Genet. 86: 301-307. 5

Bradshaw, H.D., and Stettler, R.F. 1995. Molecular genetics of growth and development in 6

Populus. IV. Mapping QTLs with large effects on growth, form, and phenology traits in a 7

forest tree. Genetics 139: 963-973. 8

Canham, C.D., Berkowitz, A.R., Kelly, V.R., Lovett, G.M., and Ollinger, S.V. 1996. Biomass 9

allocation and multiple resource limitation in tree seedlings. Can. J. For. Res. 26: 1521-10

1530. 11

Ceulemans, R., and Isebrands, J.G. 1996. Carbon acquisition and allocation. 1996. In Biology 12

of Populus and its implications for management and conservation. Edited by R.F. Stettler, 13

H.D. Bradshaw, Jr., P.E. Heilman, and T.M. Hinckley. NRC Research Press, National 14

Research Council of Canada, Ottawa, ON, Canada. pp. 355-399. 15

Ceulemans, R., Shao, B.Y., Jiang, X.N., and Kalina, J. 1996. First- and second-year 16

aboveground growth and productivity of two Populus hybrids grown at ambient and 17

elevated CO2. Tree Physiol. 16: 61-68. 18

Churchill, G.A., and Doerge, R.W. 1994. Empirical threshold values for quantitative trait 19

mapping. Genetics 138: 963-971. 20

Clark, S.L., and Schlarbaum, S.E. 2003. Growth comparisons between open-pollinated 21

progeny of sugar maple grown under shade and in full sunlight. HortSci. 38: 302-303. 22

23

Dickmann, D.I., and Keathley, D.E. 1996. Linking physiology, molecular genetics, and the 1

Populus ideotype. In Biology of Populus and its implications for management and 2

conservation. Edited by R.F. Stettler, H.D. Bradshaw, Jr., P.E. Heilman, and T.M. 3

Hinckley. NRC Research Press, National Research Council of Canada, Ottawa, ON, 4

Canada. pp. 491-514. 5

Dickson, R.E., Coleman, M.D., Riemenschneider, D.E., Isebrands, J.G., Hogan, G.D., and 6

Karnosky, D.F. 1998. Growth of five hybrid poplar genotypes exposed to interacting 7

elevated CO2 and O3. Can. J. For. Res. 28: 1706-1716. 8

Enquist, B.J., and Niklas, K.J. 2002. Global allocation rules for patterns of biomass 9

partitioning in seed plants. Science 295: 1517-1520. 10

Ericsson, T., Rytter, L., and Vapaavuori, E. 1996. Physiology of carbon allocation in trees. 11

Biomass Bioenergy 11: 115-127. 12

Frewen, B.E., Chen, T.H.H., Howe, G.T., Davis, J., Rohde, A., Boerjan, W., and Bradshaw, 13

H.D., Jr. 2000. Quantitative trait loci and candidate gene mapping of bud set and bud flush 14

in Populus. Genetics 154: 837-845. 15

Friend, A.L., Scarascia-Mugnozza, G., Isebrands, J.G., and Heilman, P.E. 1991. 16

Quantification of two-year-old hybrid poplar root systems: morphology, biomass, and 14C 17

distribution. Tree Physiol. 8:109-119. 18

Grattapaglia, D., and Sederoff, R.R. 1994. Genetic linkage maps of Eucalyptus grandis and 19

Eucalyptus urophylla using a pseudo-testcross-mapping strategy and RAPD markers. 20

Genetics 137: 1121-1137. 21

24

Grattapaglia, D., Bertolucci, F.L.G., Penchel, R., and Sederoff, R.R. 1996. Genetic mapping 1

of quantitative trait loci controlling growth and wood quality traits in Eucalyptus grandis 2

using a maternal half-sib family and RAPD markers. Genetics 144: 1205-1214. 3

Heilman, P.E., Ekuan, G., and Fogle, D. 1994. Above- and below-ground biomass and fine 4

roots of 4-year-old hybrids of Populus trichocarpa x Populus deltoides and parental 5

species in short-rotation culture. Can. J. For. Res. 24: 1186-1192. 6

Horwath, W.R., Pregitzer, K.S., and Paul, E.A. 1994. 14C allocation in tree-soil systems. Tree 7

Physiol. 14: 1163-1176. 8

Ibrahim, L., Proe, M.F., and Cameron, A.D. 1997. Main effects of nitrogen supply and 9

drought stress upon whole-plant carbon allocation in poplar. Can. J. For. Res. 27: 1413-10

1419. 11

Isebrands, J.G., and Nelson, N.D. 1983. Distribution of 14C-labelled photosynthates within 12

intensively cultured Populus clones during the establishment year. Physiol. Plant. 59: 9-13

18. 14

Karim, S.A., and Hawkins, B.J. 1999. Variation in response to nutrition in a three-generation 15

pedigree of Populus. Can. J. For. Res. 29: 1743-1750. 16

King, J.S., Albaugh, T.J., Allen, H.L., and Kress, L.W. 1999. Stand-level allometry in Pinus 17

taeda as affected by irrigation and fertilization. Tree Physiol. 19: 769-778. 18

Körner, C. 1994. Biomass fractionation in plants: A reconsideration of definitions based on 19

plant functions. In J. Roy and E. Garnier (eds.), A Whole Plant Perspective on Carbon-20

Nitrogen Interactions, pp. 173-185. SPB Academic Publishing, The Hague, The 21

Netherlands. 22

25

Kubiske, M.E., Pregitzer, K.S., Zak, D.R., and Mikan, C.J. 1998. Growth and C allocation of 1

Populus tremuloides genotypes in response to atmospheric CO2 and soil N availability. 2

New Phytol. 140: 251-260. 3

Li, B.L., Allen, H.L., and McKeand, S.E. 1991. Nitrogen and family effects on biomass 4

allocation of loblolly pine seedlings. For. Sci. 37: 271-283. 5

Lincoln, S., and Lander, E. 1992. Systematic detection of errors in genetic linkage data. 6

Genomics 14: 604-610. 7

Marques, C.M., Vasquez-Kool, J., Carocha, V.J., Ferreira, J.G., O’Malley, D.M., Liu, B.H., 8

and Sederoff, R. 1999. Genetic dissection of vegetative propagation traits in Eucalyptus 9

tereticornis and E. globules. Theor. Appl. Genet. 99: 936-946. 10

McMillin, J.D., and Wagner, M.R. 1995. Effects of water stress on biomass partitioning of 11

ponderosa pine seedlings during primary root growth and shoot growth periods. For. Sci. 12

41: 594-610. 13

Michael, D.A., Isebrands, J.G., Dickmann, D.I., and Nelson, N.D. 1988. Growth and 14

development during the establishment year of two Populus clones with contrasting 15

morphology and phenology. Tree Physiol. 4: 139-152. 16

Nadezdina, N., and Cermak, J. 2002. Instrumental methods for studies of structure and 17

function of root systems of large trees. J. Exp. Bot. 54: 1511-1521. 18

Newcombe, G., and Bradshaw Jr., H.D. 1996. Quantitative trait loci conferring resistance in 19

hybrid poplar to Septoria populicola, the cause of leaf spot. Can. J. For. Res. 26: 1943-20

1950. 21

Niklas, K.J., and Enquist, B.J. 2001. Invariant scaling relationships for interspecific plant 22

biomass production rates and body size. PNAS 98: 2922-2927. 23

26

Niklas, K.J., and Enquist, B.J. 2002. Canonical rules for plant organ biomass partitioning and 1

annual allocation. Amer. J. Bot. 89: 812-819. 2

Oleksyn, J., Modrzynski, J., Tjoelker, M.G., Zytkowiak, R., Reich, P.B., and Karolewski, P. 3

1998. Growth and physiology of Picea abies populations from elevational transects: 4

common garden evidence for altitudinal ecotypes and cold adaptation. Func. Ecol. 12: 5

573-590. 6

Ovinton, J.D. 1957. Dry matter production by Pinus sylvestris L. Ann. Bot. 21: 287-314. 7

Paterson, A.H., Lander, E.S., Hewitt, J.D., Peterson, S., Lincoln, S.E., and Tanksley, S.D. 8

1988. Resolution of quantitative traits into Mendelian factors by using a complete linkage 9

map of restriction fragment length polymorphisms. Nature 335: 721-726. 10

Poorter, H, and Nagel, O. 2000. The role of biomass allocation in the growth response of 11

plants to different levels of light, CO2, nutrients and water: a quantitative review. Aust. J. 12

Plant Physiol. 27: 595-607. 13

Pregitzer, K.S. 2002. Fine roots in trees – a new perspective. New Phytol. 154: 267-273. 14

Pregitzer, K.S., Dickmann, D.I., Hendrick, R., and Nguyen, P.V. 1990. Whole-tree carbon and 15

nitrogen partitioning in young hybrid poplars. Tree Physiol. 7: 79-93. 16

Retzlaff, W.A., Handest, J.A., O-Malley, D.M., McKeand, S.E., and Topa, M.A. 2001. 17

Whole-tree biomass and carbon allocation of juvenile trees of loblolly pine (Pinus taeda): 18

influence of genetics and fertilization. Can. J. For. Res. 31: 960-970. 19

Reynolds, H.L., and D’Antonio, C. 1996. The ecological significance of plasticity in root 20

weight ratio in response to nitrogen: Opinion. Plant Soil 185: 75-97. 21

Rhodenbaugh, E.J., and Pallardy, S.G. 1993. Water stress, photosynthesis and early growth 22

patterns of cuttings of three Populus clones. Tree Physiol. 13: 213-226. 23

27

Scarascia-Mugnozza, G.E., Ceulemans, R., Heilman, P.E., Isebrands, J.G., Stettler, R.F., and 1

Hinckley, T.M. 1997. Production physiology and morphology of Populus species and 2

their hybrids grown under short rotation. II. Biomass components and harvest index of 3

hybrid and parental species clones. Can. J. For. Res. 27: 285-294. 4

Sewell, M.M., Bassoni, D.L., Megraw, R.A., Wheeler, N.C., and Neale, D.B. 2000. 5

Identification of QTLs influencing wood properties traits in loblolly pine (Pinus taeda L.). 6

I. Physical wood properties. Theor. Appl. Genet. 101: 1273-1281. 7

Sewell, M.M., Davis, M.F., Tuskan, G.A., Wheeler, N.C., Elam, C.C., Bassoni, D.L., and 8

Neale, D.B. 2002. Identification of QTLs influencing wood property traits in loblolly pine 9

(Pinus taeda L.). II. Chemical wood properties. Theor. Appl. Genet. 104: 214-222. 10

Stettler, R.F., and Bradshaw, H.D. 1994. The choice of genetic material for mechanistic 11

studies of adaptation in forest trees. Tree Physiol. 14: 781-796. 12

Stettler, R.F., Zsuffa, L., and Wu, R. 1996. The role of hybridization in the genetic 13

manipulation of Populus. In Biology of Populus and its implications for management and 14

conservation. Edited by R.F. Stettler, H.D. Bradshaw, Jr., P.E. Heilman, and T.M. 15

Hinckley. NRC Research Press, National Research Council of Canada, Ottawa, ON, 16

Canada. pp. 87-112. 17

Stirling, B., Newcombe, G., Vrebalov, J., Bosdet, I., and Bradshaw, H.D. 2001. Suppressed 18

recombination around the MXC3 locus, a major gene for resistance to poplar leaf rust. 19

Theor. Appl. Genet. 103:1129-1137. 20

Tsarouhas, V., Gullberg, U., and Lagercrantz, U. 2002. An AFLP and RFLP linkage map and 21

quantitative trait locus (QTL) analysis of growth traits in Salix. Theor. Appl. Genet. 105: 22

277-288. 23

28

Tschaplinski, T.J., Tuskan, G.A., Gebre, G.M., and Todd, G.E. 1998. Drought resistance of 1

two hybrid poplar clones grown in a large-scale plantation. Tree Physiol. 18: 653-658. 2

Tuberosa, R., Sanguineti, M.C., Landi, P., Giuliani, M.M., Salvi, S., and Conti, S. 2002. 3

Identification of QTLs for root characteristics in maize grown in hydroponics and analysis 4

of their overlap with QTLs for grain yield in the field at two water regimes. Plant Mol. 5

Biol. 48: 697-712. 6

Tuskan, G.A., and Rensema, T.R. 1992. Clonal differences in biomass characteristics, coppice 7

ability and biomass prediction equations among four Populus clones grown in eastern 8

North Dakota. Can. J. For. Res. 22:348-354. 9

Tuskan, G.A., DiFazio, S.P., and Teichmann, T. 2004. Poplar genomics is getting popular: 10

The impact of the poplar genome project on tree research. Plant Biol. 6: 2-4. 11

USDA Soil Conservation Service. 1971. Soil survey: Benton County Area, Washington. 12

Series WA605, pp 50. 13

Verhaegen, D., Plomion, C., Gion, J.M., Poitel, M., Costa, P., and Kremer, A. 1997. 14

Quantitative trait dissection analysis in Eucalyptus using RAPD markers. 1. Detection of 15

QTL in interspecific hybrid progeny, stability of QTL expression across different ages. 16

Theor. Appl. Genet. 95: 597-608. 17

Wang, J.R., Hawkins, C.D.B., Letchford, T. 1998. Relative growth rate and biomass 18

allocation of paper birch (Betula papyrifera) populations under different soil moisture and 19

nutrient regimes. Can. J. For. Res. 28: 44-55. 20

Wu, R.L. 1998. Genetic mapping of QTLs affecting tree growth and architecture in Populus: 21

implication for ideotype breeding. Theor. Appl. Genet. 96: 447-457. 22

29

Wu, R.L. 2000. Quantitative genetic variation in leaf size and shape in a mixed diploid and 1

triploid population of Populus. Genet. Res. 75: 215-222. 2

Wu, R., Bradshaw, H.D, Jr., and Stettler, R.F. 1998. Developmental quantitative genetics of 3

growth in Populus. Theor Appl genet. 97: 1110-1119. 4

Wu, R.L., Ma, C.X., Littell, R.C., and Casella, G. 2002. A statistical model for the genetic 5

origin of allometric scaling laws in biology. J. Theor. Biol. 219: 121-135. 6

Wu, R.L., Ma, C.X., Lou, X.Y., and Casella, G. 2003. Molecular dissection of allometry, 7

ontogeny, and plasticity: A genomic view of developmental biology. BioScience 53: 8

1041-1047. 9

Wullschleger, S.D., Jansson, S., and Taylor, G. 2002. Genomics and forest biology: Populus 10

emerges as the perennial favorite. Plant Cell 14: 2651-2655. 11

Yin, T.M., DiFazio, S.P., Gunter, L.E., Riemenschneider, D., and Tuskan, G.A. 2004. Large-12

scale heterospecific segregation distortion in Populus revealed by a dense genetic map. 13

Theor. Appl. Genet. 109: 451-463. 14

Zobel, R.W. 1996. Genetic control of root systems. In Plant roots: The hidden half. Edited by 15

Y. Waisel, A. Eshel, and U. Kafkafi. Marcel Dekker, Inc. New York. pp. 21-30. 16

17

18

19

20

21

22

23

30

Table 1. Correlation matrices for the distribution of biomass among all plant components for 1

one-year-old progeny from two pedigrees of hybrid poplar. 2

___________________________________________________________________________ 3

Stem Branches Leaves Cutting Coarse Fine 4

roots roots 5

___________________________________________________________________________ 6

Family 331 7

Stem 1.00 8

Branches 0.83 1.00 9

Leaves 0.88 0.84 1.00 10

Cutting 0.90 0.85 0.86 1.00 11

Coarse roots 0.86 0.82 0.80 0.91 1.00 12

Fine roots 0.60 0.59 0.62 0.68 0.66 1.00 13

14

Family 13 15

Stem 1.00 16

Branches 0.72 1.00 17

Leaves 0.83 0.84 1.00 18

Cutting 0.90 0.79 0.86 1.00 19

Coarse roots 0.79 0.79 0.79 0.84 1.00 20

Fine roots 0.38 0.40 0.44 0.40 0.37 1.00 21

___________________________________________________________________________ 22

23

31

Table 2. Parameters of the regression log (Y) = log (β) + α log (X) for one-year-old plants 1

from two hybrid poplar pedigrees grown near Wallulu, WA. Genotypes from Family 331 and 2

Family 13 were combined (n = 406). In all cases P < 0.001. 3

___________________________________________________________________________ 4

Plant components β α r2 F 5

___________________________________________________________________________ 6

Shoot vs. total 0.5187 1.0268 0.983 24,241 7

Root vs. total 0.4468 0.9738 0.958 9,292 8

Leaves vs. total 0.3573 0.9560 0.948 7,350 9

Stem vs. total 0.2164 0.9896 0.926 5,059 10

Branches vs. total 0.0100 1.4126 0.755 1,097 11

Cutting vs. total 0.5302 0.7620 0.913 4,260 12

Coarse roots vs. total 0.0504 1.2544 0.910 4,057 13

Fine roots vs. total 0.0355 0.9162 0.611 631 14

Shoot vs. root 1.6346 0.9841 0.894 3,406 15

___________________________________________________________________________ 16

17

18

19

20

21

22

23

32

Table 3. Correlation matrices for the distribution of biomass among all plant components for 1

two-year-old progeny from Family 13. 2

___________________________________________________________________________ 3

Stem Branches Leaves Cutting Coarse 4

roots 5

___________________________________________________________________________ 6

Stem 1.00 7

Branches 0.78 1.00 8

Leaves 0.84 0.96 1.00 9

Cutting 0.87 0.79 0.89 1.00 10

Coarse roots 0.75 0.75 0.90 0.80 1.00 11

___________________________________________________________________________ 12

13

14

15

16

17

18

19

20

21

22

23

33

Table 4. Allometric coefficients and regression parameters describing relationships among 1

individual components and total tree biomass. Data are for two-year-old plants from Family 2

13. In all cases P < 0.001. 3

___________________________________________________________________________ 4

Plant components β α r2 F 5

___________________________________________________________________________ 6

Shoot vs. total 0.6529 1.0241 0.995 25,832 7

Root vs. total 0.4315 0.9057 0.903 1,327 8

Leaves vs. total 0.3167 0.9621 0.966 4,095 9

Stem vs. total 0.2787 1.0298 0.910 1,442 10

Branches vs. total 0.0852 1.1004 0.890 1,155 11

Cutting vs. total 0.3330 0.8257 0.903 1,327 12

Coarse roots vs. total 0.1482 0.9685 0.734 392 13

Shoot vs. root 3.9326 0.9980 0.856 844 14

___________________________________________________________________________ 15

16

17

18

19

20

21

22

23

34

Table 5. Main characteristics and effects of QTL with an LOD > 2.5 for biomass traits 1

measured in Family 13. LOD significance thresholds for each QTL were calculated based on 2

permutations. 3

Trait ChromosomeOrigin of

'Positive' AllelePVE (%)

LOD Score

LOD Threshold

-----------------------------------Year 1-----------------------------------

Above ground IV P. deltoides 17.0 4.07 2.02 %Below ground I P. trichocarpa 17.9 8.40 1.96 IV P. trichocarpa 18.3 4.30 1.99 Coarse root IV P. deltoides 10.5 3.21 2.00 %Coarse root I P. trichocarpa 9.1 4.86 2.43 Cutting IV P. deltoides 9.7 3.25 2.10 %Fine root II P. trichocarpa 7.5 2.87 1.86 III P. trichocarpa 11.9 3.04 1.76 Leaves IV P. deltoides 17.1 4.72 2.12 %Leaves II P. trichocarpa 9.5 4.11 2.18 Stem III P. deltoides 10.0 4.27 2.18 VII P. deltoides 8.2 3.79 1.93 %Stem VI P. trichocarpa 9.8 2.68 2.14 Total Biomass IV P. deltoides 14.2 4.05 1.81

-----------------------------------Year 2----------------------------------- Above ground XII P. deltoides 8.9 3.08 1.70 XIII P. deltoides 11.3 4.51 1.82 Below ground XII P. deltoides 11.1 4.95 1.99 XIII P. deltoides 9.4 4.68 1.82 %Below Ground VI P. deltoides 11.5 2.89 2.17

35

Branch XIII P. deltoides 12.5 4.60 1.65 %Branch XIV P. deltoides 8.3 3.06 2.02 Coarse root XII P. deltoides 11.8 4.08 1.82 XIII P. deltoides 8.1 2.66 1.94 %Coarse root VI P. deltoides 9.1 3.19 1.82 Cutting XII P. deltoides 10.2 2.52 2.01 XIII P. deltoides 10.1 3.36 2.04 Leaves XII P. deltoides 9.2 4.18 1.77 %Leaves VI P. deltoides 11.8 3.40 1.85 %Stem VI P. trichocarpa 10.8 3.64 1.80 Total Biomass XII P. deltoides 9.8 3.10 1.57 XIII P. deltoides 11.2 3.97 1.74

1

2

3

4

5

6

7

8

9

10

11

12

13

36

Figure Legends 1

Figure 1. Frequency distributions of total biomass (panels A and B), shoot biomass (panels C 2

and D), and root biomass (panels E and F) for one-year-old trees from two hybrid poplar 3

pedigrees. 4

5

Figure 2. Above- and below-ground biomass for two hybrid poplar pedigrees. Data for both 6

Family 331 (panels A and C) and Family 13 (panels B and D) are for one-year-old trees. 7

8

Figure 3. Relative distribution of total biomass (i.e., mass fraction) to shoots, roots, and 9

individual plant components for one-year-old trees harvested from two hybrid poplar 10

pedigrees. 11

12

Figure 4. Allometric relationships between individual plant components and total whole-tree 13

biomass. Data are for one-year-old trees from two hybrid poplar pedigrees. Family 331 data 14

are shown as open circles; Family 13 as closed circles. 15

16

Figure 5. Above- and below-ground biomass (panels A and B) and the relative distribution of 17

total biomass (panels C and D) to leaves, stems, branches, and coarse roots for two-year-old 18

progeny from Family 13. 19

20

Figure 6. Graphic representation of quantitative trait loci (QTL) for biomass traits in a TD x 21

D hybrid poplar backcross pedigree. Vertical bars indicate the mapping interval for each 22

37

QTL, delineated by markers with significant trait associations. The peak position of the QTL 1

is indicated by a heavy horizontal line. For details on marker codes, see Yin et al. (2004). 2

Num

ber o

f Gen

otyp

es

0

40

80

120

Total Biomass (g)

0 500 1000 15000

15

30

45

A) Family 331

B) Family 13

Shoot Biomass (g)

0 250 500 750

Root Biomass (g)

0 250 500 750

C) Family 331 E) Family 331

D) Family 13 F) Family 13

0

250

500

750

0

75

150

225Bio

mas

s (g

)

0

250

500

750

1000

Bio

mas

s (g

)

0

75

150

225

300

Tota

l

Sho

ot

Roo

t

Leav

es

Ste

ms

Bra

nche

s

Coa

rse

Fine

A) Family 331

B) Family 13

C) Family 331

D) Family 13

Cut

tings

0.00

0.25

0.50

0.75

0.00

0.15

0.30

0.45Mas

s Fr

actio

n

0.00

0.25

0.50

0.75

1.00

Mas

s Fr

actio

n

0.00

0.15

0.30

0.45

0.60A) Family 331

B) Family 13

C) Family 331

D) Family 13

Sho

ot

Roo

t

Leav

es

Ste

ms

Bra

nche

s

Coa

rse

Fine

Cut

tings

Bio

mas

s (g

)

100

101

102

103

100

101

102

103

A) Leaves

C) Branches

B) Stem

D) Cutting

Total Biomass (g)

100 101 102 103100

101

102

103

100 101 102 103

E) Coarse roots F) Fine roots

Mas

s Fr

actio

n

0.00

0.25

0.50

0.75

Mas

s Fr

actio

n

0.00

0.15

0.30

0.45

Sho

ot

Roo

t

Leav

es

Ste

ms

Bra

nche

s

Coa

rse

Cut

tings

Bio

mas

s (g

)

0

2000

4000

6000

8000

Bio

mas

s (g

)

0

750

1500

2250

3000A) B)

C) D)

Tota

l

P_2852_A157.3

P_2385_A

BELO

W-1

CO

ARSE R

OO

T-1

P_204_C0.0S8_328.8P_2385_C11.6T4_1012.1S15_8 S5_3713.8T4_7 S6_1215.5S8_2917.9P_2786_A S12_1820.4T1_1322.3T7_423.5T3_13 T3_36S17_2124.1

S15_16 T12_1525.3T2_3026.5S13_2029.5S1_2036.5T9_1 S1_1943.2S3_1350.5S1_2452.9S2_754.1P_575_A59.1T12_2260.6S2_3285.0T7_995.7S2_6107.8S13_16 T5_25121.4T5_12124.3T10_4129.0T1_26 T7_13135.7P_93_A148.6S4_20150.2S7_13 S7_12T12_4152.8

S4_24 T3_10S6_4154.1

S3_1163.4S6_20 S13_31T7_15171.3

T2_31178.2S8_4180.8S8_28182.1O_30_A184.2T5_4193.5T3_17198.1T12_12206.8S5_29210.6P_2789_A219.9P_634_A S17_43226.5S17_33230.3S17_12232.7S4_19243.1

S17_26262.9

I

FINE R

OO

T-1LEAVES-1

S5_350.0P_2523_A5.4S7_2211.4S4_5219.8S7_1923.0O_260_A23.7P_2766_A28.4S15_2029.6S3_232.8T7_643.8T9_647.6T3_1257.0S5_563.8S6_1865.1P_2797_A72.1S4_1375.0S4_1475.3S2_281.5S4_1885.1O_207_A97.6P_684_A98.1O_286_A103.0

T9_9114.8P_422_A116.2S13_12126.7P_667_A127.9T1_33130.9T1_15142.8S12_11148.3T2_32 T5_6148.9T1_25151.7

T10_6163.8

II

FINE R

OO

T-1

STEM-1

P_2277_A0.0S6_15.1S5_118.7S13_2512.6S17_1614.4T4_921.6S8_1122.7S6_1023.9S1_526.3T6_330.5T2_3 T9_11P_2501_A31.0

S4_3638.1O_30_B53.0S13_18 T6_1059.7O_203_A61.4S4_1263.0T10_366.6T7_582.6T7_190.0T4_190.4S1_497.5S17_42 T3_19102.7

S13_34119.8

S6_14131.1T8_14132.9

III

AB

OVE-1

BELO

W-1

CO

AR

SE RO

OT-1

CU

TTING

-1

LEAVES-1

TOTA

L-1

S4_390.0O_349_A1.8O_356_A2.9STS1_A STS33.9P_204_E15.1

S1_733.5S15_1134.0T7_1137.7O_127_A49.1T3_18 S1_2350.1S8_19 S8-2054.8S8_962.1S5_1662.4S4_767.8P_2826_A T5_19T5-1870.2

T3_7 S13-2672.5S4_1580.6

S17_5102.2

S13_15111.7P_2020_A114.5

S15_9128.5S3_18 S3_19131.4

IV

AB

OVE-2

BELO

W-2

CO

AR

SE RO

OT-2

CU

TTING

-2

LEAVES-2

TOTA

L-2

S3_150.0T12_1317.1S13_1418.2S6_2219.4S6_15 T9_1220.6S13_1322.4S2_1524.8S12_21 W03_A25.3T1_2029.0T5_2332.7S6_1634.5P_495_A35.3T10_12 T11_11S3_30 P_204_FS8_21

36.2

S5_2436.8S13_2737.4W05_A43.0P_2885_A50.5S12_1052.4T1_3154.8T11_372.4T11_473.0

XII

BR

AN

CH

-2

S2_40.0S6_1112.7S1_1315.3T5_1119.9O_59_A24.4S2_1425.6T9_532.1S2_1235.4T8_345.6T1_3047.3S1_247.8T3_2553.8T9_354.7

T7_874.7T11_581.3T12_387.0T5_290.6P_571_A96.6

T9_7106.3S6_6115.8T2_4 S8_22126.3P_2515_A S17_25S5_6127.3

T2_14129.1

XIV

T6_10.0S8_133.5S2_248.2S3_279.3S5_2312.2S4_5120.1O_16_C22.0S2_11 T3_237.8T2_2041.9P_2658_A42.5S3_1645.0S17_645.3T12_747.2S4_3350.0T7_252.5S7_31 S17_3454.4S8_2 P_649_B55.0S4_3558.0T8_962.2S13_38 S12_2267.6T1_1768.1S13_4070.5T3_1572.2T7_1077.0T1_2277.5T1_2378.1

AB

OVE-2

BELO

W-2

BR

AN

CH

-2

CO

AR

SE RO

OT-2

CU

TTING

-2

TOTA

L-2XIII

STEM-1

P_607_A0.0T3_90.5S8_25r2.4P_562_A P_2557_AS13_216.5

S7_189.0T12_511.4S2_30 S6_23T7_314.5

P_2140_A25.9S1_1142.2W17_A S17_20T12_6 S4_643.7

S6_344.9S13_148.4P_2794_A53.7S15_157.7O_312_A T6_858.7S17_1965.8T8_1074.2T10_877.5S5_3480.8P_2730_A W16_A90.1

VII

BELO

W-2

CO

AR

SE RO

OT-2

LEAVES-2

STEM-2

O_268_A0.0T5_152.7O_279_A13.6S17_117.2P_2557_B20.7P_2556_A S6_221.1

O_363_A32.8

W07_A44.3

T2_6 S17_1859.2W12_A63.2T12_164.9S17_3265.5T3_468.3S4_4074.2O_190_A74.5P_2221_A75.7S4_4781.5S12_690.0P_2217_A97.8T3_33 S4_37102.0S13_29103.1T9_2109.0O_26_A118.0T3_5119.1T1_32131.5O_268_D133.7S7_2138.2

VI

Green – Above ground Brown – Below ground

Normal – P. trichocarpa positive Italicized – P. deltoides positive

Normal – Absolute Biomass Underlined – Proportional Biomass

STEM-1