Oecologia (1995) 101:439-447
ORIGINAL
9 Springer-Verlag 1995
PAPER
G.M. Berntson 9E.J. Farnsworth 9F.A. Bazzaz
Allocation, within and between organs, and the dynamics
of root length changes in two birch species
Received: 5 April 1994 / Accepted: 21 October 1994
Spatial and temporal dynamics of biomass allocation within and between organs were investigated in
seedlings of two birch species of contrasting successional status. Seedlings of Betula alleghaniensis Britt (yellow birch) and B. populifolia Marsh (gray birch) were
grown for 6 weeks at two nutrient levels in rectangular
plexiglass containers to allow non-destructive estimates
of root growth, production and loss. Leaf area and production were simultaneously monitored. Yellow birch responded more to nutrient level than gray birch in terms
of total biomass, shoot biomass, leaf area and root
length. Yellow birch also flexibly altered within-organ
allocation (specific leaf area, specific root length and
specific soil amount). In contrast, gray birch altered between-organ allocation patterns (root length:leaf area and
soil amount:leaf area ratios) more than yellow birch in
response to nutrient level. Yellow birch showed greater
overall root density changes within a very compact root
system, while gray birch showed localized root density
changes as concentric bands of new root production
spread through the soil. Species differ critically in their
responses of standing root length and root production
and loss rates to nutrient supply. Early successional species such as gray birch are hypothesized to exhibit higher
plasticity in varied environments than later successional
species such as yellow birch. Our results suggest that different patterns of allocation, within and between plant
organs, do not necessarily follow the same trajectories.
To characterize thoroughly the nature of functional flexibility through ontogeny, within- and between-organ patterns of allocation must be accounted for.
Abstract
Key words
Allocation 9Betula
9 Biomass
9 Nutrients
Root length dynamics
G.M. Berntson (~) 9E.J. Farnsworth 9F.A. Bazzaz
Harvard UniversityBiologicalLaboratories, 16 DivinityAve.,
Cambridge, MA 02138, USA
Introduction
Whole-plant responses to resource limitation are often
described by patterns of biomass allocation to different
organs (e.g., Chapin 1980; Bloom et al. 1985; Tilman
1988). The mass of an organ may be a good proxy for
the cost of construction, but mass alone is not always indicative of an organ's capacity to acquire resources. Beyond mass, the physical arrangement of roots and the
physiological capacity for uptake within the root system
determine the capacity for water and nutrient uptake
within the soil (Nye and Tinker 1977; Barber 1984;
Clarkson 1985; Aerts et al. 1991). The spatial placement
of organs in relation to resource supply determines the
potential "functional" value of biomass allocation.
The arrangement of mass within organs can vary substantially, independently of patterns of biomass allocation between organs. For example, specific leaf area
(SLA, leaf area per unit leaf mass) and specific root
length (SRL, root length per unit mass) vary among species according to the environment and life span of modules (e.g., SLA: Reich et al. 1991, 1992; SRL: Fitter
1991; Eissenstat 1991, 1992). Variations in allocation
within an organ can lead to patterns of biomass partitioning between different organs that show little relationship
to the relative "functional strength" of these organs (sensu K6rner and Renhardt 1987). For example, relative
biomass allocation between roots and shoots shows weak
relationships with the relative amount of root length supported per unit leaf area (Kummerow 1983; K6rner and
Renhardt 1987; Larigauderie et al. 1991; Aerts et al.
1992).
Many studies explore the functional implications of
variations in shoot and root architecture for resource acquisition (e.g., Givnish 1986 and references therein), but
these studies often use disparate terminology to relate architecture and biomass. To simplify our discussion of
these relationships we make the distinction between the
allocation of biomass between different components
(e.g., roots or leaves) and the manner in which biomass
is allocated within an organ. Within-organ allocation en-
440
compasses the way mass is physically arranged within an
individual organ (e.g., SLA and SRL). The relative
"functional strength" of different organs may not parallel
between-organ patterns of allocation, depending on how
mass is allocated within an organ. In this paper we assess
the roles these different modes of plasticity play in the
occupation of space and thus the potential acquisition of
resources.
Below-ground resource acquisition can be complicated to quantify, due to the variety of below-ground resources (e.g., mobile versus immobile nutrients) and
their differing spatial and temporal patterns of supply.
Morphology and physiological activity of individual
roots vary with growth and position (Clarkson 1985),
and there may be trade-offs in growth rates versus uptake
rates (Jackson et al. 1990; Jackson and Caldwell 1991).
Young roots growing into unexploited soil can also contribute disproportionately to whole-plant nutrient acquisition (Robinson et at. 1991; Robinson 1991). Here, we
monitor three attributes of root growth of relevance to resource uptake: total root length, the amount of soil occupied by the root system (soil amount), and the loc~[tion
and rate of relative root production and loss (root relative
density change, RRDC).
Static observations of mass may greatly underestimate the total investment into an organ over time. Plants
are modular; short life spans of individual modules such
as root segments relative to the entire plant lead to turnover within organs. A recent review concluded that allocation of biomass to fine roots in temperate forest ecosystems is unchanged by increased nitrogen availability,
while root turnover is increased (Hendricks et at. 1993).
This pattern is analogous to leaf turnover responses to
increased nutrient levels (Mooney and Gulmon 1982). To
better understand the dynamics of investment and function, we also determine the total production and loss of
leaves and roots.
In this study, we investigate patterns of biomass allocation within plants, and the manner in which the biomass is physically arranged within and between leaves
and roots. Because we manipulated below-ground resources, we place special emphasis on determining the
spatial and temporal patterns of total and relative root
production and loss. This detailed documentation of
growth represents a critical first step in elucidating the
relationship between biomass and adaptive architecture.
The functional value of different root architectures and
growth "strategies" may only be confirmed by actually
quantifying the spatial dynamics of uptake during
growth. As an experimental system to explore these
ideas we studied two species of birch of contrasting life
history (Betula alleghaniensis Britt and B. populifolia
Marsh). The primary questions we focus on in this study
are:
1. What are the relationships between patterns of allocation within and between organs?
2. How are root dynamics and the spatio-temporal distribution of root production affected by nutrient supply?
Methods
Species
Betula populifolia Marsh (gray birch) is a fast growing, shade-intolerant early successional tree usually found in recently disturbed
sites, typically with low soil nutrients and/or moisture (Burns and
Honkala 1990). B. alleghaniensis Britton (yellow birch) is an intermediately shade-tolerant, mid- to later-successional tree often
found in moist to wet sites typically with higher nutrient availability (Burns and Honkala 1990; Harlow et al. 1979). Yellow birch
seedlings often establish on moss-covered rocks, fallen logs and
other shallow rooting zones. Research on the morphological and
physiological plasticity of these species was recently summarized
by Bazzaz and Wayne (1994). They concluded that the earlier successional gray birch showed greater plasticity in more characters
than the later successional yellow birch. Given existing evidence
on the physiological and morphological flexibility of early successional species (Bazzaz 1979, 1983, 1987), we expected that gray
birch would show greater within- and between-organ plasticity
than yellow birch.
Planting design
Seeds of Betula populifolia and B. alleghaniensis were collected
from at least three separate trees for each species in the fall of
1989 near Petersham, Mass. Seeds were stored dry at 4~ until
September 1992, when they were sown into flats filled with a
2:1:1 mixture of vermiculite, peat and sand, and watered daily.
Flats were exposed to 400 ~tE m -2 light 18 h a day, with day/night
temperatures set at 28/20~
Seedlings of both species had
emerged after 2 weeks, and had produced 2-4 leaves after 3
weeks. At that time, a total of 24 comparably sized seedlings of
each species were bare-rooted and transplanted into growth containers which were placed in one of four glasshouse modules.
Seedlings were randomly assigned to one of two nutrient treatments and one of four blocks, where each block was a separate
glasshouse module. Individuals that died within the 1st week were
replaced with new seedlings. Final sample sizes were 12 seedlings/species/nutrient level. Lighting in the glasshouse modules
was supplemented with metal halide lights to provide 12 h of light
at 300-400 gE m-L
Plants were grown in containers of 22x36x2 cm (internal dimensions, 1.7 1 volume), constructed of plywood, wood spacers
and clear plexiglas. The wood portion of the pots was lined with
polyethylene, which was sealed against the plexiglas with silicon
sealant. The plexiglas covered one of the two large sides of each
pot to allow non-destructive observations of root growth (e.g.,
Berntson and Woodward 1992; Fitter and Stickland 1992). Opaque
plastic covered the plexiglas to prevent exposing roots to light.
Containers were held at 30 ~ angles from vertical with the plexiglas
facing down to encourage root growth along the plexiglas surface.
The upper surface of each container was painted white to maximize albedo and thereby minimize fluctuations in soil temperature.
The soil was a 2:2:1 mixture of turface (a 2:1 clay), sand and
sterilized soil. This mixture allowed good drainage while maintaining a high cation exchange capacity. Plants were watered daily
and fertilized twice weekly with 90 ml of a balanced nutrient solution (Peter's 20:20:20 N-P-K) at either 0.25 or 1.0 g 1-1. These fertilization rates were scaled to nitrogen supply rates of 35.5 and
150 Kg nitrogen h ~ 1 y1-1, which cover a range of nitrogen supply
rates via mineralization plus deposition in New England forests
(Abet et al. 1989).
Above-ground growth
At weekly intervals (on the same days root tracings were made)
plant height, stem diameter, and the number and length of all
leaves (_+1 ram) were measured. Leaf length data were used to es-
441
timate leaf area non-destructively using exponential regressions of
leaf length versus leaf area (r2=0.975 for gray birch, r2=0.973 for
yellow birch).
Root growth: spatial and temporal distributions
of production and loss
Starting 2 weeks after transplanting, roots visible through the
plexiglas were traced weekly onto acetate sheets. These images
were then digitized into a computer using a flatbed scanner at 30
pixels cm q (LaCie Silverscanner, LaCie, Ore.). Traces were made
by hand using fine point (0.5 mm) permanent markers. This tracing method was able to resolve roots less than 1 m m apart. Each
digitized image was skeletonized (eroded to a single pixel in
width). A 22x36 grid of 1-cm 2 cells was laid over the tracings and
the length of root in each cell determined by counting the number
of pixels. Total root length (cm) was determined by summing the
number of pixels throughout the grid. The amount of soil occupied
by the root system (soil amount, cm 2) was calculated by counting
the number of cells that had some root present. Root density (cm
cm -~) was calculated by taking the average number of pixels per
cell for all those cells that had some root present.
Net production and net loss of roots were determined by comparing the length of root present in each cell of the grid between
successive tracings. An increase in length within a cell was interpreted as a net production of roots in the cell, while a decrease
was interpreted as a net loss. Net production of roots for an entire
root system was calculated as the sum of all increases in root
length throughout the grid; net root loss was calculated as the sum
of all decreases in root length. This method of determining root
production and loss is similar to sequential coring in that it uses
differences in the amount of root present between two intervals to
estimate production and loss (e.g., McClangherty et al. 1982; Fairley and Alexander 1985; Vogt et al. 1989). Our method is similar
to other methods relying on window observations in that sequential measurements are made on the same roots through time (e.g.,
Hendrick and Pregitzer 1992, 1993; Pregitzer et al. 1993). This
means that our method avoids the problems of significant overestimation of root production and loss resulting from random sam-
Fig. 1 A - F Biomass and functional characters at harvest
(week 6). Bars are 1 SEM.
Shaded bars are gray birch,
open bars are yellow birch.
Boxes hold ratio of high to low
nutrient treatment means for
each species and the significance level of Bonferroni-corrected LSM comparison between the nutrient treatments
for each species: *** P<0.001,
** P<0.01, * P<0.05; ~
P<0.10; ns, P _>0.10
pling variation (sensu Singh et al. 1984; Vogt et al. 1986; Lanenroth et al. 1986). However, because this method examines net root
production within subsets of the entire root system it is possible
that it results in an underestimation of actual root production and
loss rates.
Spatial distributions of new root production and loss were
characterized by calculating the relative change in root density
(root relative density change, RRDC) within each individual cell
in the grid. RRDC was calculated for each cell as
[log(D2)-log(D~)]/t, where D i is the length of root within a given
cell at time i and t is the interval of time between sequential traces.
The spatial distribution of RRDC was reduced to a single dimension by calculating average RRDC for all the cells that were a given distance from the base of the root system. In order to minimize
the effect of small-scale variations in RRDC, the one-dimensional
array of averaged RRDC values were smoothed by taking a running average (n=3). From the smoothed, one-dimensional array of
RRDC, five parameters were calculated to characterize spatial distributions RRDC: magnitude and position of maximum positive
RRDC (+mag, +pos) and negative RRDC ( - m a g and -pos) and the
spread of positive RRDC (width of zone of net root production).
Harvest and correction factors
Six weeks after transplanting, fresh roots were removed from the
soil using a pin board to maintain their spatial orientation (B/Shin
1979; Tatsumi et al. 1989). Intact root systems were laid out on a
board with a 3.9x3.9 cm grid and the number of sections having a
portion of root present were counted. The spread of root systems
relative to the amount of soil occupied by roots estimated from the
tracings were compared, and correction factors (soil amount actual/soil amount tracing) were calculated. The correction factors
were 1.70 cm 2 cm -2 for gray birch and 2.86 cm 2 cm 2 for yellow
birch. Correction factors for the root length present at the surface
of the containers versus the total actual roots within the soil matrix
were also calculated (root length actual/root length tracing). The
root length correction factors were 48.07 cm cm q for gray birch
and 132.22 cm cm q for yellow birch. Correction factors for each
species were used to correct all the measurements of root length
1,.6
14 :
a
I
0.6
0.4
0.
I
~I
150
[
100
~
50
0
I ***106'1% [
200
E
0.8
150 r
0.6"
I
100
0.4"
0.2"
5o
0
0
500
0.3
0.2
- "
F
8i:i
400
300
I
I
200
100
0
Nutrients : Low
High
Species : Gray Birch
Low
High
Yellow Birch
Low
High
Gray Birch
Low
High
Yellow Birch
<
442
and root area made using the tracings. The root length correction
values are substantially larger than those reported by Gross et al.
(1992). However, our plants were older and larger and the containers used in this experiment were much thicker (greater soil volume/observed surface area) than those used by Gross et al. (1992).
Derivation of the correction factors for root length involved selecting several root systems at random. For derivations of the correction factors for amount of soil, all plants were used. Regressions
(of the form y=mx) yielded an r 2 of 0.965 for gray birch and 0.936
for yellow birch. Previous studies have found similarly strong relationships between observed length and actual length (r2=0.93;
Berntson and Woodward 1992).
Plants were harvested after 6 weeks, and the mass of roots,
shoots/petioles and leaves were determined after oven-drying at
80~ for 2 days. All biomass values presented here (whole plant,
individual organ and within- and between-organ allocation figures)
refer to data obtained from harvested plants. Specific leaf area
(SLA) was calculated as the ratio of leaf area to leaf mass. Due to
the multiform behavior of tree root systems, we analyzed fine
(ephemeral, non-lignified, nutrient-absorbing) roots separately
from coarse (large-diameter, lignified) roots involved mainly in
solute transport and structural support. Within air-dried samples,
coarse roots were >__0.5 ram, round (larger diameter fine roots became flattened), relatively stiff and strong, and light red-orange.
Because the length of coarse roots was less than 0.5% of the total
length of the root systems, fine root mass was used in all measures
of allocation. Specific root length (SRL) was calculated as the ratio of root length to fine root mass. Specific soil amount (SSA)
was calculated as the ratio of soil amount to fine root mass.
tained from type II regressions (geometric mean regression, Sokal
and Rohlf 1981) of numerator and denominators for all between
and within organ ratios (e.g. R/S, SLA, SRL) were not significantly different from zero. Thus, directly examining ratios between
two variables was equivalent to examining the slopes of these relationships. Nutrient and species were treated as fixed effects, and
block as a random effect in the ANOVA model. To examine the effect of nutrient level on each species individually, least-square
means were computed and compared by t-tests with associated
Bonferroni probabilities (adjusted for two comparisons). Root dry
weight to leaf dry weight ratios were arcsine-square-root transformed prior to analyses. Assumptions of homoscedasticity and
normality were tested on all variables using Scheffr-Box tests (Sokal and Rohlf 1981) and normal-probability plots (Velleman
1989). The sequential, repeated measurements of growth and below-ground spatial distribution were analyzed using a repeatedmeasures multivariate ANOVA with measurements at each time
interval as independent variables (SuperAnova v l . l l , Gagnon et
al. 1989). Time was treated as a fixed, repeated measure where the
numerator and denominator degrees of freedom in all F-tests involving time (time and timextreatment interactions) were corrected using Greenhouse-Geisser epsilon correction factors (Gagnon
et al. 1989).
Statistical analyses
Gray birch seedlings were significantly larger in terms of
total biomass and all measured functional characters,
than yellow birch (Fig. 1, Table 1). High-nutrient plants
Allocation between and within organs was examined by performing ANOVAs on ratios of the relevant parameters. Intercepts ob-
Results
Allocation within and between organs
(harvest measurements)
Table 1 Two-way ANOVA results for final harvest data (df degrees of freedom, P probability, SS sum of squares). Significant terms
(P<0.05) are in bold type. Variables marked with * were log transformed prior to analysis
Species *
df
Species
1
Nutrient
1
Nutrient
1
Whole plant
mass *
P
SS
<0.001
1.12
<0.001
2.732
0.310
0.033
Leaf mass *
P
SS
P
SS
P
SS
P
SS
P
SS
P
SS
P
SS
P
SS
P
SS
P
SS
P
SS
<0.001
0.921
<0.001
1.680
<0.001
0.480
<0.001
1.200
<0.001
2.142
<0.001
0.133
0.451
Fine
root mass *
Leaf area *
Root length *
Soil amount *
Fine root mass/
leaf mass
Root length/
leaf area
Root area/
leat area
Specific
leaf area
Specific
fine root length
Specific
soil amount
1665
0.016
7.27
0.032
68.32
0.063
2.44• 10-5
<0.001
1.06xlO 6
<0.001
3.051
<0.001
1.655
<0.001
1.763
0.003
0.846
0.284
0.141
<0.001
0.295
0.057
11032
<0.001
23.18
0.004
130.04
0.026
3.59• 10-5
<0.001
1.58x10 -6
0.369
0.025
0.189
0.066
0.156
0.057
0.925
0.001
0.613
0.031
0.035
0.035
0.443
1723
0.081
3.71
0.007
113.75
O. 1024
1.86x 10-5
0.981
37.80
Block
3
0.165
0.168
0.190
0.154
0.062
0.291
0.021
0.300
0.079
0.606
0.711
0.165
0.087
0.050
0.660
4627
0.638
0.66
0.006
65.90
0.005
9.85x 10 -5
0.049
5.68x10 -5
Residual
41
1.251
1.233
1.507
1.128
3.403
4.888
0.291
117737
1.16
13.88
2.73 • 10 -6
2.73•
-6
443
5
Fig. 2 A,B,C Between-organ
allocation and D,E,F within-organ allocation at harvest (week
6). Shading, symbols and layout same as Fig. 1.
i
600
25
300
~gG
.~
ns-2.4%
0
g
e~
r~
***-47.8%
0
1000
150
E
-2 ~
lO0-
~ ~
600 ,4 .-.
I
400 ~ ~
2oo ~
50"3 8%
.2% I
~
:~
~
0.7
0 . 6 ~
0.5
0.4
~ o.3
0.2
T
0.
~
.3% I
Nutrients : Low
High
Species : GrayBirch
Table 2 Results of repeated measures analysis of variance for
non-destructive measurements of functional characters, root turnover and summary statistics of RRDCV s (dfdegrees of freedom, P
df
n
1
Leaf
Area
P
SS
<0.001
5.06
0.127
0.25
0.449
0.06
0.023
1.08
Root
Length
Soil
Amount
Relative
Loss Ratio
+Mag
P
SS
P
SS
P
SS
P
SS
P
SS
P
SS
P
SS
P
SS
0.120
1.34
0.341
0.44
0.001
2.25
<0.001
2.25
0.245
10.55
0.871
0.001
0.505
0.52
0.434
13.5
<0.001
10.68
<0.001
13.44
0.008
0.42
0.080
0.42
<0.001
591.51
0.109
0.062
0.007
9.19
<0.001
1569.8
0.278
0.65
0.331
0.46
0.015
0.01
0.824
0.01
0.815
0.42
0.726
0.003
0.789
0.08
0.490
10.5
0.050
4.52
0.409
1.41
0.130
1.57
0.013
1.57
0.102
50.18
0.084
0.164
0.465
3.00
0.027
220.2
+Pos
-Mag
-Pos
Spread
S
1
NxS
1
aoo ~
B
3
Low
High
YellowB~'ch
FI
...47.o% !
..
Low
High
GrayBirch
r~
o
Low
High
YellowBirch
probability, SS sum of squares). Significant terms (P<0.05) are in
bold type. All dependant variables were log transformed prior to
analysis. S=species, N=nutrients, B=block, T=time
Grp
41
TxN
4
TxS
4
T
4
<0.001
72.41
<0.001
1.72
<0.001
0.59
0.042
0.15
0.159
0.23
<0.001
21.84
201.44
<0.001
19.59
58.64
<0.001
5.30
3.32
<0.001
5.30
3.32
<0.001
310.84
934.97
0.002
0.941
0.473
<0.001
47.21
31.63
<0.001
889.6
2895.8
0.244
0.54
0.626
0.15
0.008
0.10
0.633
0.10
0.382
9.85
0.345
0.067
0.005
15.60
0.304
22.7
0.002
2.16
0.036
0.78
0.002
0.30
0.270
0.30
<0.001
98.64
0.553
0.035
0.959
0.19
<0.001
640.7
0.322
0.44
0.043
0.75
0.029
2.01
0.001
2.01
0.668
3.89
0.217
0.099
0.006
15.13
0.097
47.0
0.813
0.63
0.846
0.42
0.535
1.07
0.174
1.07
0.070
64.67
0.002
0.806
0.037
20.21
0.159
92.7
0.42
of both species were significantly larger in all aspects except for soil amount. Yellow birch showed greater enh a n c e m e n t with increased nutrients for whole plant biomass (244%) and leaf b i o m a s s (261%, Fig. 1A,B) than
gray birch (164% and 185%, respectively). Gray birch
exhibited greater differences in b e t w e e n organ allocation
due to nutrient regime than did yellow birch (Fig. 2 A - C ) .
o
1200
1000 ~
800 ~
600 ~
400 . ~
2o0 [
TxNxS TxB
4
12
Residual
164
2.10
15.51
10.74
9.34
9.34
422.22
2.593
123.25
776.4
The proportion of b i o m a s s allocated to roots and the relative a m o u n t of root length supported and a m o u n t of soil
per unit leaf area was higher in low n u t r i e n t treatments
for both species.
Gray birch exhibited greater and significant reductions in soil a m o u n t ( - 5 0 . 9 % ) and root length to leaf area
( - 3 3 . 9 % ) ratios u n d e r elevated nutrients than yellow
444
Trajectories of Nutrient Enhancement
[(High - Low)/Low]
200-
I
I
[
I
I
I
150- Gray Birch
.l
-"
50- ~
o-50- I
I
I
I
I
200~
I
[
I
[
I
1
.D'
-50
.0
i
r
1
~
I
I
I
E
4
5
Age (Weeks)
6
o 30
,~
~
I --50 r~
~
.30
h.-lo
o,
20-
1
2
I
3
[
4
lO-
0-
~-
.-lo
~
I
1 "50
~~176
[]
"~
I
~D
150- Yellow B
-50'
I
.~"-"40
~1, "100
," zJS
100-
~9.
I
[
5
6
"-30
Age (weeks)
Fig. 3 Trajectories of effects of nutrient treatments for leaf area,
root length and root area expressed as percent difference between
nutrient treatments relative to low. The left vertical axes are the
scales for the relative changes in leaf area, the right vertical axes
are the scales for the relative changes in root length and root
area
birch (-33.5% and -18.2%, respectively; Fig. 2A,B).
Both species showed significantly reduced fine root mass
to leaf mass ratios under high nutrients, with greater reductions on the part of gray (-32.6%) than yellow birch
(-21.3%, Fig. 2C).
In contrast, yellow birch altered w i t h i n - o r g a n allocation more than gray birch. Nutrient regimes did not significantly affect specific root length or specific leaf area
in gray birch, but these ratios were significantly reduced
with high nutrients in yellow birch (-34.8% and -47.8%
respectively, Fig. 2C,E). While both species showed significant reductions in specific soil amount (SSA, soil
amount/root mass) under high nutrients, this pattern was
more pronounced and more highly significant for yellow
(47.0%) than gray birch (34.1%).
Overall, nutrient-level differences were greater than
species-level differences in leaf area throughout the experiment (Table 2). Species-level disparities were greater
for both root length and soil amount. Gray birch had longer roots, distributed over a greater area than yellow
birch, regardless of nutrient level (Table 2). Both leaf area and soil amount showed significant time by species by
nutrient effects. For leaf area, this interaction resulted
from a leveling off of new leaf production in high nutrients at the very end of the experiment, producing a decline in the relative enhancement with high nutrients
(Fig. 3). The magnitude of leaf area enhancements in
high nutrients was twice or more that of root length enhancements for both gray and yellow birch by week 6 of
the experiment (Fig. 3). No leaf turnover was observed
through the course of the experiment. For soil amount,
this three-way interaction term resulted from initially
[
I
2
3
Fig. 40ntogeny of root relative loss ratio (loss/production). Each
symbol represents the amount of root length lost relative to root
length produced for the week previous to time where the symbol is
placed. (A yellow birch, high nutrients; /~ yellow birch, low
nutrients; 9 gray birch, high nutrients; @ gray birch, low
nutrients)
slow enhancement in root area in high nutrients followed
by a rapid increase and the reverse temporal pattern of
enhancement for yellow birch (Fig. 3).
Spatio-temporal patterns of root production and loss
Relative loss ratios (net root loss/net root production)
were high in both species, with rates of 10-30% by week
4 (Fig. 4). This pattern differed between species and nutrient treatments (Table 2). Increasing nutrients resulted
in decreased relative loss ratios. Gray birch showed lower relative loss ratios than yellow birch at both nutrient
levels. Nutrients had a small effect on relative loss ratios
in gray birch, with a slight decrease in the high nutrient
treatment in weeks 5 and 6. The nutrient treatment
showed a much more pronounced effect in yellow birch.
High-nutrient yellow birch showed a two-fold decrease
in the relative loss ratio during weeks 2-4, but these effects declined toward the end of the experiment.
Graphical representation of the spatial distribution of
relative root density changes (RRDC) through time
(Fig. 5) illustrates that gray birch seedlings had larger
root systems, while yellow birch had more compact root
systems. RRDCs for both species were lower under low
nutrient conditions, resulting in a more diffuse pattern of
root production throughout the soil in which the roots
were distributed. Figure 5 also shows that areas of high
RRDC spread out as advancing fronts (concentric bands)
from the base of the root system. The magnitude (+mag,
-mag) and spatial location (+pos, -pos) of net root production and net root loss were differentially affected by
nutrient supply for the two species (Table 2, Fig. 5).
Gray birch in particular showed pronounced localized
relative root loss near the base, and high relative root
production as the root system grew through the soil. Soil
amount did not differ with nutrient level in gray birch,
445
Gray Birch
Low Nutrients
Yellow Birch
High Nutrients
Low Nutrients High Nutrients
Log(D~)-Log(D 1)
cln r -2
;-~ ~.:','..::~:' :i::.:
-0.26 ... -0.53
g:,x.x-' .4.~,:.,,
:J~~:!*"&~
2-3
ii~ i
0 ... -0.26
,.. i:: 1:?~gsii~?.~.?.
:::~:~. .~.:~:~: ".,:~:.~:
~i/':-;:.,'~. . . .
....
:~.~:~
0 ... 0.26
3-4
0.26 ... 0,53
0.53 ... 0.80
~ ~":-"-"~
~-i~
0.80 ... 1.07
1.07 ... 1.33
i~i~#i~-
9
':::
9
x
4-5
:: iiiiiitil
~i?i:g
N:~.L.:!~:.f~.:
!:...,......:.x,i!~!
5-6
~ ~,,;":"~:~"!~i~':!2"*~":o!!.~i~...,:~i~{~~ii.:.~!i~}ii
~ . ; . . , ~ !::.:%~"~!;
: :~i,!i. ~;~~!!" i.i.i.i.i.i. f. .i .!.~. .i .i .~. i. l. .i .~. . . . . . . . . .
!
,l...,a
=2
,w-I
Fig. g Graphical representation of the spatial distribution of relative root density change (RRDC) through time. Values less than 0
are designated with a dark outline. Each density map is the average of 12 plants
but the rate of new root production decreased throughout. Yellow birch concentrated most growth over a shallow zone neat the stem, showing lower rates of relative
root production and a more diffuse spread of net root
loss.
Discussion
In this experiment, we made simultaneous above and below-ground measurements of plant growth to address
two questions concerning the potential limitations of biomass as a currency for describing plant responses to environmental change. We grew two congeners of contrasting life histories and successional habitats to elucidate
how the disparity between within and between organ allocation may be expressed variably between species.
Within- and between-organ allocation
We found that yellow birch and gray birch consistently
differed from one another in their magnitudes of within
and between organ responses to the nutrient treatments.
Yellow birch showed greater within-organ responsive-
ness to nutrient enhancement. The amount of leaf area
per unit leaf mass and soil amount per unit root mass decreased with an increased supply of nutrients. In contrast, the length of root per unit root mass declined. Gray
birch, however, exhibited more flexible between organ
allocation, consistently showing less below-ground investment (biomass, root length or soil amount) with increasing nutrients.
We observed substantial discrepancies between within- and between-organ patterns of allocation in response
to altered nutrient supplies. These discrepancies were
more pronounced for yellow birch than for gray birch
due to the greater variation observed in within-organ patterns of allocation. Changes in biomass allocation between leaves and roots resulted in similar patterns of
change in leaf area to root length or soil amount ratios
for gray birch, but not for yellow birch. These observations demonstrate that differences in the plasticity of between-organ allocation relative to within-organ allocation may be characteristics of individual species.
While we also observed different patterns of leaf area,
root length and soil amount enhancement in response to
increased soil nutrients for the two species, we cannot
make inferences about ontogenetic patterns of within organ allocation from these data. Root and leaf biomass
were determined only at the end of the experiment. It is
possible that all of the patterns of within-organ allocation
presented here are unique to the time of harvest. Thus, it
is possible that the relationships between within- and between-organ allocation may themselves have pronounced
ontogenetic patterns.
446
Bazzaz (1979, 1983, 1987) has suggested that earlier
successional species typically show greater levels of phenotypic plasticity over a wide range of environmental
gradients. In apparent contradiction to this hypothesis,
the later successional yellow birch showed higher magnitude total biomass changes than the earlier successional
gray birch (Fig. 1). When we examine patterns of withinand between-organ allocation, however, it is not clear to
which species the epithet "more plastic" belongs (Table
3). Bazzaz and Wayne (1994, p. 381) observed of B. alleghaniensis and B. populifolia that "there do not seem to
be clear differences in the types (i.e., physiological versus morphological) of traits shown by pioneer and latersuccessional species" grown over an experimental light
gradient. We did find consistent differences in betweenand within-organ allocation between the two species in
response to two levels of nutrient supply.
Spatio-temporal patterns of root production and loss
Our observations reinforce the caveat that static observations of standing root length need to be interpreted with
caution when attempting to describe the dynamics of allocation. Patterns of root turnover were complex, varying
with species, nutrient supply, and time. One interesting
result is that whole-plant relative loss ratios (net root
loss/net root production) appeared to be decoupled from
the degree of localization of root loss. Yellow birch
showed greater overall root relative loss ratios, while
gray birch showed more pronounced localized relative
root loss and production (compare Figs. 4 and 5). In gray
birch, new roots were produced in spreading concentric
bands while older roots near the stem were lost. Yellow
birch, on the other hand, distributed its roots within a
smaller amount of soil in a more diffuse spatial distributions of relative root loss and production. These differences in the spatial distribution of RRDC are interesting
in the light of the tendency of yellow birch to establish in
locations with shallow rooting space (Burns and Honkala
1990). The net effect of these differences is that yellow
birch showed little capacity to expand its range of soil
exploration, while gray birch moved its production of
new roots rapidly through the available soil.
Our observations suggest that both relative allocation
to roots and root turnover decrease with increasing nutrient supply, in contrast to Hendricks et al. (1993). Our
findings generally accord with those of Pregitzer et al.
(1993), who found that patch additions of nitrogen resulted in increased life-spans of individual roots. Pregitzer et al. (1993) observed community-level responses
of root growth, which could be the result of one or more
individual species. Our results clearly demonstrate that
different species can show very different patterns of root
production and loss in response to nutrient supply. For
example, at week 4 there was no effect of nutrient supply
on the relative loss ratio in gray birch, but in yellow
birch there was a 100% increase with the low nutrient
supply (Fig. 4). In comparing the root life span data of
Pregitzer et al. (1993) with our relative loss ratio data,
caution needs to be employed because they monitored
the status of individual roots and we monitored changes
in root length in small regions within the soil through
time.
Summary and implications
Previous studies suggest that partitioning of biomass
alone does not account for the functional strength between different compartments. Similarly, our study illustrates that gray and yellow birch exhibit variable relationships of within- and between-organ allocation, resulting from different relative plasticity of within- and
between-organ arrangements. We have yet to explore experimentally the implications of these divergent allocation patterns for actual nutrient uptake: a critical area for
further study.
This study has also demonstrated that even in the absence of above-ground senescence and turnover, roots
may show high rates of senescence. Further, the total
amount, spatial distribution of root production and loss
and response to nutrient supply can be quite variable between species. In order to better understand the patterns
of root production and loss in natural ecosystems, we
must take into account different species' inherent patterns of root growth and response to their environment.
Acknowledgements G.M.B. performed this research under appointment to the Graduate Fellowships for Global Change Program administered by Oak Ridge Institute for Science and Education for the U.S. Department of Energy, Office of Health and Environmental Research, Atmospheric and Climate Research Division.
E.J.F. was funded by a graduate fellowship from the Harvard Department of Organismic and Evolutionary Biology. S. Bassow, T.
Sipe, and R Wayne and two anonymous reviewers provided helpful comments on the manuscript.
References
Aber JD, Nadelhoffer KJ, Steudler P, Melillo JM (1989) Nitrogen
saturation in northern forest ecosystems. BioScience 39:
378-386
Aerts R, Boot RGA, Aart PJM van der (1991) The relation between above- and belowground biomass allocation patterns
and competitive ability. Oecologia 87:551-559
Aerts R, Caluwe H de, Konings H (1992) Seasonal allocation of
biomass and nitrogen in four Carex spieces from mesotrophic
and eutrophic fens as affected by nitrogen and light. J Ecol 80:
653-664
Barber SA (1984) Soil nutrient bioavailability: a mechanistic approach. Wiley, New York
Bazzaz FA (1979) The physiological ecology of plant succession.
Annu Rev Ecol Syst 10:351-371
Bazzaz FA (1983) Characteristics of populations in relation to disturbance in natural and man-modified ecosystems. In: Godron
HA, Gordon M (eds) Disturbance and ecosystems, components of response. Springer, Berlin Heidelberg New York, pp
259-275
Bazzaz FA (1987) Experimental studies on the evolution of niche
in successional plant populations. In: Gray AJ, Crawley MJ,
Edwards PJ (eds) Colonization, succession and stability.
Blackwell, Oxford, pp 245-271
447
Bazzaz FA, Wayne PM (1994) Coping with environmental heterogeneity: the physiological ecology of tree seedling regeneration across the gap-understory continuum. In: Caldwell M, Pearcy B (eds) Exploitation of environmental heterogeneity by
plants. Academic Press, New York, pp 349-390
Berntson GM, Woodward FI (1992) The root system architecture
of Senecio vulgaris L. under elevated CO 2 and drought. Funct
Ecol 6:324-333
Bloom AJ, Chapin FS III, Mooney HA (1985) Resource limitation
in plants - an economic analogy. Annu Rev Ecol Syst 16:
363-392
B6hm W (1979) Methods of studying root systems. Springer, Berlin Heidelberg New York
Burns RM, Honkala BH (1990) Silvics of North American Trees,
vol 2, Hardwoods. U.S. Department of Agriculture, Washington D.C.
Chapin FS III (1980) The mineral nutrition of wild plants. Annu
Rev Ecol Syst 11:233-260
Clarkson DT (1985) Factors affecting mineral nutrient acquisition
by plants. Annu Rev Plant Physiol 36:77-115
Eissenstat DM (1991) On the relationship between specific root
length and the rate of root proliferation: a field study using citrus rootstocks. New Phytol 118:63-68
Eissenstat DM (1992) Costs and benefits of constructing roots of
small diameter. J Plant Nutr 15:763-782
Faifley RI, Alexander IJ (1985) Methods of calculating fine root
production in forests. In: Fitter AH, Read DJ, Atkinson D,
Usher MB (eds) Ecological Interactions in Soil. Blackwell,
Oxford, pp 3 7 4 2
Fitter AH (1991) Characteristics and functions of root systems. In:
Waisel AEY, Kafkafi U (eds) Plant roots: the hidden half. Marcel Dekker, New York, pp 3-25
Fitter AH, Stickland TR (1992) Fractal characterization of root
system architecture. Funct Ecol 6:632-635
Gagnon J, Roth JM, Carroll M, Hofman R, Haycock KA, Plamondon J, Feldman DSJ, Simpson J (1989) SuperANOVA: accessible general linear modeling, v l. 11. Abacus Concepts, Berkeley
Givnish TJ (1986) On the economy of plant form and function.
Cambridge University Press, Cambridge
Gross KL, Maruca D, Pregitzer KS (1992) Seedling growth and
root morphology of plants with different life-histories. New
Phytol 120:535-542
Harlow WM, Harrar ES, White FM (1979) Textbook of dendrology. McGraw-Hill, New York
Hendrick RL, Pregitzer KS (1992) The demography of fine roots
in a northern hardwood forest. Ecology 73:1094-1104
Hendrick RL, Pregitzer KS (1993) Patterns of fine root mortality
in two sugar maple forests. Nature 361:59-61
Hendricks JJ, Nadelhoffer KJ, Abet JD (1993) Assessing the role
of fine roots in carbon and nutrient cycling. Trends Ecol Evol
8:174-178
Jackson RB, Manwaring JH, Caldwell MM (1990) Rapid physiological adjustment of roots to localized soil enrichment. Nature 344:58-60
Jackson RB, Caldwell MM (1991) Kinetic responses of Pseudoroegneria roots to localized soil enrichment. Plant Soil 138:
231-238
K6rner C, Renhardt U (1987) Dry matter partitioning and root
length/leaf area ratios in herbaceous perennial plants with diverse altitudinal distribution. Oecologia 74:411-418
Kummerow J (1983) Root surface/leaf area ratios in arctic dwarf
shurbs. Plant Soil 71:395-399
Larigauderie A, Ellis BA, Mills JN, Kummerow J (1991) The effect of root and shoot temperatures on growth of Ceanothus
greggii seedlings. Ann Bot 67:97-101
Lauenroth WK, Hunt HW, Swift DM, Singh JS (1986) Reply to
Vogt et al. Ecology 67:580-582
McClaugherty CA, Aber JD, Melillo JM (1982) The role of fine
roots in the organic matter and nitrogen budgets of two forested ecosystems. Ecology 63:1481-1490
Mooney HA, Gulmon SL (1982) Constraints on leaf structure and
function in reference to herbivory. BioScience 32:198-206
Nye PH, Tinker PB (1977) Solute movement in the soil-root system. Blackwell, Oxford
Pregitzer KS, Hendrick RL, Fogel R (1993) The demography of
fine roots in response to patches of water and nitrogen. New
Phytol 125:575-580
Reich PB, Uhl C, Walters MB, Ellsworth DS (1991) Leaf lifespan
as a determinant of leaf structure and function among 23 Amazonian tree species. Oecologia 86:16-24
Reich PB, Walters MB, Ellsworth DS (1992) Leaf life-span in relation to leaf, plant, and stand characteristics among diverse
ecosystems. Ecol Monogr 62:365-392
Robinson D (1991) Roots and resource fluxes in plants and communities. In: Atkinson D (ed) Plant root growth: an ecological
perspective. Blackwell, London, pp 103-130
Robinson D, Linehan DJ, Caul S (1991) What limits nitrate uptake
from the soil. Plant Cell Environ 14:77-85
Singh JS, Laurenroth WK, Hunt HW, Swift DM (1984) Bias and
random errors in estimators of net root production: a simulation approach. Ecology 65:1760-1764
Sokal RR, Rohlf FJ (1981) Biometry, 2nd edn. Freeman, San
Francisco
Tatsumi J, Yamauchi A, Kono Y (1989) Fractal analysis of plant
root systems. Ann Bot 64:499-503
Tilman D (1988) Plant strategies and the structure and dynamics
of plant communities. Princeton University Press, Princeton
Velleman PF (1989) Data desk: statistics guide. Odesta, Northbrook
Vogt KA, Grier CC, Gower ST, Sprugel DG, Vogt DJ (1986)
Overestimation of net root production: a real or imaginary
problem? Ecology 67:577-579
Vogt KA, Vogt DJ, Moore EE, Sprugel DG (1989) Methodolgical
considerations in measuring biomass, production, respiration
and nutrient resorption for tree roots in natural ecosystems. In:
Winship JGT, Winship LJ (eds) Applications of continuous
and steady state methods to root biology. Kluwer Academic,
Dordrecht, pp 217-232