Biomass Allocation and Leaf Chemical Defence in Defoliated Seedlings of Quercus serrata with Respect to Carbon–Nitrogen Balance

Abstract

• Background and Aims Both nutrient availability and defoliation affect the carbon–nutrient balance in plants, which in turn influences biomass allocation (e.g. shoot-to-root ratio) and leaf chemical composition (concentration of nitrogen and secondary compounds). In this study it is questioned whether defoliation alters biomass allocation and chemical defence in a similar fashion to the response to nutrient deficiency.

• Methods Current-year seedlings of Quercus serrata were grown with or without removal of all leaves at three levels of nutrient availability.

• Key Results Plant nitrogen concentration (PNC), a measure of the carbon–nutrient balance in the plant, significantly decreased immediately after defoliation because leaves had higher nitrogen concentrations than stems and roots. However, PNC recovered to levels similar to or higher than that of control plants in 3 or 6 weeks after the defoliation. Nitrogen concentration of leaves produced after defoliation was significantly higher than leaf nitrogen concentration of control leaves. Leaf mass per plant mass (leaf mass ratio, LMR) was positively correlated with PNC but the relationship was significantly different between defoliated and control plants. When compared at the same PNC, defoliated plants had a lower LMR. However, the ratio of the leaf to root tissues that were newly produced after defoliation as a function of PNC did not differ between defoliated and control plants. Defoliated plants had a significantly lower concentration of total phenolics and condensed tannins. Across defoliated and control plants, the leaf tannin concentration was negatively correlated with the leaf nitrogen concentration, suggesting that the amount of carbon-based defensive compounds was controlled by the carbon–nutrient balance at the leaf level.

• Conclusions Defoliation alters biomass allocation and chemical defence through the carbon–nutrient balance at the plant and at the leaf level, respectively.

INTRODUCTION

When plants are subject to herbivory, they often significantly reduce growth and reproduction (Crawley, 1983). Responding to herbivory, plants alter the chemical composition in the remaining or newly flushing leaves (induced defence, Karban and Myers, 1989; Karban and Baldwin, 1997). Reduction in nitrogen concentration and accumulation of secondary compounds such as phenolics are believed to be effective in avoiding further herbivory (Karban and Baldwin, 1997). Such induced defence sometimes continues for several years after damage by herbivory (Tuomi et al., 1984).

It has been argued that the carbon–nutrient balance within plants is responsible for the induction of chemical defence (carbon–nutrient balance hypothesis; Bryant et al., 1983; Tuomi et al., 1984, 1990). As leaf nitrogen concentration (LNC) is higher than that of roots and stems, loss of leaves decreases the nitrogen concentration of the whole plant. For new leaves to regenerate after herbivory, nitrogen may be more limiting than carbon. This will lead to a lower LNC than in plants not defoliated by herbivores, and to a greater amount of secondary compounds that are produced from excessive carbon. This hypothesis has been supported by several studies that have shown that concentrations of phenolics are altered by the manipulation of light and nutrient availability (Bryant et al., 1987a, b; Price et al., 1989; Gebauer et al., 1998; Koricheva et al., 1998, Hemming and Londroth, 1999; Dormann, 2003), although it does not explain accumulation of other secondary compounds (Koricheva et al., 1998, Hamilton et al., 2001; Koricheva, 2002; Nitao et al., 2002).

Carbon–nutrient balance plays an important role for understanding plant growth. One well-known phenomenon is alteration in biomass allocation between root and shoot in response to changes in the balance between carbon and nutrient availability, which is observed in many species (Brouwer, 1962; Wilson, 1988). Plants need both carbon and nutrients for growth. As the role of root and shoot are to acquire nutrients and carbon, respectively, plants tend to allocate biomass to the root or shoot to balance the uptake rate of carbon and nutrients. More allocation to roots at the expense of shoot growth is effective to compensate for lower rates of nutrient uptake at low nutrient availabilities, while more allocation to the shoot ameliorates growth under low light conditions (Brouwer, 1962). Hirose (1986, 1987) has shown that the leaf mass ratio (LMR, leaf mass per plant mass) is a linear function of plant nitrogen concentration (PNC), irrespective of nutrient availability and size of individual plants, and suggested that the shoot-to-root ratio is regulated via the carbon–nitrogen balance within a plant. Mathematical models have predicted an optimal shoot-to-root ratio that maximizes the relative growth rate under given environmental conditions (Hirose, 1988a; Hilbert, 1990; van der Werf et al., 1993; Ishizaki et al., 2003; Osone and Tateno, 2003). Hirose (1988b) further showed that other parameters of growth analysis such as relative growth rate, LNC and specific leaf area are also strongly correlated with PNC, suggesting that PNC can be used for the scaling of plant growth.

Many authors have investigated changes in biomass allocation in response to defoliation. Defoliation accelerates shoot growth at the expense of root growth (for review see Wilson, 1988). Consequently, the shoot-to-root ratio after defoliation recovers to a value similar to that before the defoliation (Brouwer, 1962; Fick et al., 1971; Huxley and Sammerfield, 1976; Fondy and Geiger, 1980; Mariko and Hogetsu, 1987; Mihaliak and Lincoln, 1989). However, these studies did not consider the possibility that defoliation might also have altered the carbon–nutrient balance within the plant. If both nutrient deficiency and defoliation reduce PNC, we may expect that defoliation changes the shoot-to-root ratio in response to the reduced PNC. Alternatively, is the shoot-to-root ratio after defoliation independent of the change in carbon–nutrient balance?

The question addressed in the present paper is whether the plant response to defoliation can be scaled with PNC, as the response to nitrogen availability can be. Current-year seedlings of Quercus serrata were raised at different nutrient availabilities with and without defoliation and PNC was used as a measure of the balance of nitrogen and carbon in the plant. The first hypothesis to be tested is that changes in PNC with defoliation alter the LMR in a similar way to that observed with nutrient deficiency. The concentration of total phenolics and condensed tannin in leaves were determined to test the second hypothesis, that changes in PNC induce production of defensive chemicals both in defoliated and nitrogen-deficient plants.

MATERIALS AND METHODS

Quercus serrata is a deciduous broad-leaved tree species, which a typical dominant species in secondary forests in Japan. Current-year seedlings of Q. serrata were used in the experiment. Acorns of Q. serrata collected at Kanazawa City on 15 November, 1998, were stored at 4 °C. On 18 May 1999, one acorn was sown per pot (1·5 L volume, 25 cm depth) filled with washed river sand. Seedlings were grown outdoors in the experimental garden of Tohoku University (38°30′N, 141°57′E). A commercial nutrient solution (N : P : K = 5 : 10 : 5, Hyponex, Murakami bussan, Kamigori, Japan) was applied every week at three nutrient levels: 0 (only tap water), 0·5, and 5 mg nitrogen per week (low, middle and high nutrient treatment, respectively). Plants were watered daily.

On 25 July, at a stage when cotyledons were considered no longer to be an important carbohydrate source (Kabeya and Sakai, 2003), manual defoliation was applied and all leaves of target individuals were removed. From this date, plants were harvested four times at 3-weeks intervals (six plants for each treatment). Plants were separated into leaves, stems and roots, and dried for at least 3 d at 70 °C after determining leaf area with a leaf area meter (Li-3000, LiCor, Lincoln, NE). Dried samples were weighed and milled. Nitrogen concentration was determined with an NC-analyzer (NC-80, Shimadzu, Kyoto). Phenolic compounds were extracted from the milled samples with 50 % methanol at 90 °C for 8 min three times. Tannin concentration was determined spectrophotometrically at 550 nm (UV-160A, Shimadzu) after 0·5 mL of the extract with 3·5 mL of a mixture of n-buthanol and HCl (15 : 1) was placed at 95 °C for 2 h (Bate-Smith, 1977). The concentration of total phenolics was determined spectrophotometrically at 700 nm after the extract was reacted with a phenolic reagent and Na₂CO₃ solution (20 % W/V) (Julkunen-Tiitto, 1985).

Total non-structural carbohydrates (TNC) in roots were determined for milled samples (Kabeya et al., 2003). Subsamples were suspended in KOH, placed in boiling water for 30 min, and acetic acid was added after cooling. To digest starch to glucose, amyloglucosidase solution was added to subsamples and incubated at 55 °C for 30 min. Digested samples were centrifuged, and all sugars in the supernatant were analysed by the phenol–sulfuric acid method. Absorbance was read at 490 nm on a spectrophotometer (UV-160A, Shimadzu). A glucose solution was used for the standard.

Analysis of variance (ANOVA) and analysis of covariance (ANCOVA) were performed with StatView ver. 5 (SAS Institute Inc.).

RESULTS

Plant biomass growth was significantly affected by day, nutrient and defoliation treatments (Table 1, Fig. 1A). Control plants showed positive growth rates with leaf flushes 1–3 times through the experiment, and their final biomass tended to be higher under higher nutrient availabilities. In defoliated plants, negative or no growth was found for the first 3 weeks after defoliation. Subsequent growth rates tended to be higher in higher nutrient availabilities. Biomass was smaller in defoliated plants than in control plants.

Table 1.

F-values from analysis of variance for the effects of day, nutrients and defoliation on plant mass, the amount of nitrogen in the plant, plant nitrogen concentration (PNC), leaf nitrogen concentration (LNC), leaf mass ratio (LMR) and the concentration of total non-structural carbohydrates in roots (TNC). Data are from plants harvested at 21, 42 and 63 d after defoliation

	Degrees of freedom	Plant mass	Plant N	PNC	LNC	LMR	TNC
Day	2	30·31***	38·89***	9·343***	6·232**	12·25***	111·3***
Nutrient	2	72·5***	349·3***	263·8***	209·1***	164·8***	54·81***
Defoliation	1	142·2***	203·9***	23·97***	154·4***	108·3***	35·94***
Day × Nutrient	4	10·46***	27·62***	1·628ns	0·566ns	1·534ns	2·643*
Day × Def.	2	5·765**	4·339*	11·48***	1·512ns	45·99***	1·524ns
Nut. × Def.	2	37·64***	90·07***	3·305ns	0·881ns	3·11*	2·318ns
D. × N. × Def.	4	4·401**	4·175**	14·15***	3·935**	11·65***	1·241ns
Error	90

^*P < 0·05;

^**P < 0·01;

^***P < 0·001;

^nsnot significant.

Fig. 1.

Changes in (A) biomass and (B) the amount of nitrogen in plants after defoliation. Open and closed symbols indicate control and defoliated plants, respectively. Circles, triangles and squares indicate the low, middle and high nutrient treatments, respectively. Mean and s.d. are shown (n = 6). The arrow indicates the day of defoliation. Values for defoliated plants on the day of defoliation were calculated as those of the stems and roots in control plants. Note that plant mass and nitrogen content are expressed on a log scale.

The amount of nitrogen in the plant was significantly affected by day, nutrient and defoliation treatments (Table 1, Fig. 1B). In the low and middle nutrient treatments, increase in plant nitrogen was small, while in the high nutrient treatment plant nitrogen increased considerably with time. Due to the loss of leaves, defoliated plants had a lower nitrogen content when compared within the same nutrient treatment throughout the experimental period.

Figure 2 shows plant nitrogen concentration (PNC, plant nitrogen per plant mass), which we use as a measure of nitrogen–carbon balance in the plant. Because leaf nitrogen concentration (LNC) was higher than PNC (Figs 2 and 3), defoliation significantly decreased PNC; however, this recovered in 3 weeks (Fig. 2, Table 1). As control plants showed a steady decrease in PNC with time, at the end of the experiment there was no significant difference in PNC between defoliated and control plants in the low and middle nutrient treatments: defoliated plants had a significantly higher PNC than control plants in the high nutrient treatment.

Fig. 2.

Changes in plant nitrogen concentration after defoliation. (A) Low, (B) middle, and (C) high nutrient treatments. Significance levels were assessed for each date using a Student t-test (ns, not significant; ^**P < 0·01; ^***P < 0·001). Symbols are as described in Fig. 1.

Fig. 3.

Changes in leaf nitrogen concentration after defoliation. Symbols are as described in Fig. 1.

Defoliated plants produced leaves immediately after defoliation and the expansion of this first flush had almost finished by 3 weeks after defoliation. LNC was significantly affected by day, nutrient treatment and defoliation (Fig. 3, Table 1). New leaves of defoliated plants had a significantly higher LNC than leaves of control plants.

Leaf mass ratio (LMR, leaf mass per total plant mass) was significantly affected by day, nutrient availability and defoliation (Table 1). Figure 4 shows allometric relationships between leaf and plant mass. When compared at the same plant mass, defoliated plants grown under low and middle nutrient availabilities had relatively lower leaf mass. However, defoliated plants under high nutrient availability had comparable leaf mass at later stages.

Fig. 4.

Allometric relationship between leaf and plant mass in defoliated and control plants grown at (A) low, (B) middle, and (C) high nutrient availability. Open and closed symbols indicate control and defoliated plants, respectively. One point denotes one plant. Plants were harvested at 0, 21, 42 and 63 d after defoliation.

Figure 5A shows leaf mass ratio (LMR, leaf mass per total plant mass) plotted against PNC. In both defoliated and control plants, LMR was positively correlated with PNC across different days and N treatments (there was no interacting effect of day × nutrient treatment, ANCOVA, P > 0·05). Between defoliated and control plants there was a significant difference in the y-intercept (ANCOVA, P < 0·0001). It should be noted that defoliated plants grown at the high nutrient treatment had a LMR comparable to control plants, but when compared at the same PNC defoliated plants had a lower LMR.

Fig. 5.

(A) Leaf mass ratio (leaf mass per total mass) and (B) leaf nitrogen concentration as a function of plant nitrogen concentration. Open and closed symbols indicate control and defoliated plants, respectively. Circles, triangles and squares indicate low, middle and high nutrient availability, respectively. One point represents one plant. Plants were harvested at 0, 21, 42 and 63 d after defoliation. For leaf mass ratio, the regression lines are y = 0·045 + 25·5x (r² = 0·79, P < 0·0001) for control plants and y = −0·067 + 23·9x (r² = 0·79, P < 0·0001) for defoliated plants. For leaf nitrogen concentration, the regression lines are y = 0·0054 + 1·18x (r² = 0·75, P < 0·0001) for control plants and y = 0·011 + 1·03x (r² = 0·68, P < 0·0001) for defoliated plants.

Figure 5B shows LNC plotted against PNC. Like LMR, LNC was a positive function of PNC across day and nutrient treatments, and the regression was significantly different between defoliated and control plants (ANCOVA, P < 0·0001).

Tables 2 and 3 show concentrations of total phenolics and condensed tannins in leaves on a mass basis in defoliated and control plants. In plants harvested 6 weeks after defoliation, there was a significant difference in total phenolics and tannin concentrations between defoliation and nutrient treatments, without any interaction. Defoliated plants tended to have lower total phenolics and tannin concentrations in leaves. Figure 6 shows tannin concentration plotted against PNC (A) and LNC (B). When plotted against PNC, there was a significant difference in the slope of the regression between defoliated and control plants (P < 0·05), while the data points fell on one regression line when plotted against LNC (ANCOVA, P > 0·05). Similar trends were also found in the total phenolics–PNC and total phenolics–LNC relationships (data not shown).

Table 2.

Analysis of variance of total phenol and condensed tannin concentrations in leaves

	Degrees of freedom	Total phenol	Condensed tannin
Nutrient	2	7·408**	5·094*
Defoliation	1	4·955*	25·18***
Nut. × Def.	2	2·37ns	0·354ns
Error	16

Data are from plants harvested at 42 d after defoliation.

^*P < 0·05;

^**P < 0·01;

^***P < 0·001;

^nsnot significant.

Table 3.

Concentrations of total phenol and condensed tannin in leaves

Nutrient availability	Harvest (days after defoliation)	Defoliation	Total phenol	Condensed tannin
Low	0	Control	0·636 (0·093)	0·342 (0·116)
	42	Control	0·595 (0·063)	0·524 (0·193)
	42	Defoliated	0·485 (0·064)	0·273 (0·120)
Middle	0	Control	0·590 (0·100)	0·359 (0·109)
	42	Control	0·503 (0·045)	0·424 (0·049)
	42	Defoliated	0·386 (0·047)	0·208 (0·030)
High	0	Control	0·539 (0·171)	0·182 (0·023)
	42	Control	0·379 (0·073)	0·313 (0·038)
	42	Defoliated	0·400 (0·105)	0·149 (0·032)

Means (± s.d.) of relative values on a dry mass basis are shown.

Fig. 6.

Leaf tannin concentration as a function of (A) plant and (B) leaf nitrogen concentration. Tannin concentration is expressed as a relative value on a dry mass basis.

Concentration of total non-structural carbohydrate (TNC) in roots was significantly different depending upon day, nutrient and defoliation treatment (Fig. 7, Table 1). TNC concentration was higher in lower nutrient treatments. At low and middle nutrients, a large reduction in TNC concentration was found 3 weeks after defoliation and the reduction was larger in defoliated plants. At the final harvest, however, defoliated plants recovered their TNC concentrations to levels similar to those before defoliation.

Fig. 7.

Changes in the concentration of total non-structural carbohydrate in roots after defoliation. Symbols (slightly shifted to avoid overlap of bars) are as described in Fig. 1.

DISCUSSION

Plant and leaf nitrogen concentration in defoliated plants

The carbon–nutrient balance hypothesis implies reduction in LNC after defoliation, based on the idea that a loss of tissues having a higher nitrogen concentration results in nitrogen deficiency in the plant (Tuomi et al., 1990). In the present study, as was expected, PNC immediately after defoliation was lower than that of control plants (Fig. 2), owing to higher LNC than PNC (Fig. 3). However, PNC recovered in 3–6 weeks. Under the low and middle nutrient treatments, PNC in defoliated plants increased and achieved similar levels to control plants, while in the high nutrient treatment PNC became significantly higher in defoliated plants than in control plants (Fig. 2). Thus defoliation did not cause nitrogen deficiency in the long term.

Under the high nutrient treatment, increase in PNC after defoliation (Fig. 2) was caused by increased rates of nitrogen uptake (Fig. 1B). Increase in PNC was also found in the low and middle nutrient treatments (Fig. 2), although nitrogen uptake remained low after defoliation (Fig. 1B). This increase was caused by a loss of biomass (Fig. 1A), probably due to increased respiratory consumption. As plants cannot photosynthesize after defoliation, respiratory loss of carbon increases nitrogen concentrations. Quercus species accumulate carbohydrates in roots up to 40 % of dry mass (Sakai et al., 1997; Kabeya et al., 2003). Defoliation decreased the root TNC (Fig. 7), which might have been used for respiration and carbon skeleton supply for producing new leaves.

LNC was higher in defoliated plants (Fig. 3). This result is surprising because many studies reported that LNC in woody species decreased (e.g. Tuomi et al., 1984, 1988; Kamata et al., 1996; Kudo, 1996) or did not change (Volin et al., 2002; Cerasoli et al., 2004) after defoliation. However, an increase in LNC after defoliation seems to be common in herbaceous species (e.g. Ruess et al., 1983; Nowak and Caldwell, 1984; Chapin and McNaughton, 1989). An increase in LNC would enhance leaf photosynthesis, which is termed compensatory photosynthesis (Nowak and Caldwell, 1984; Anten and Ackerly, 2001). There seems to be no theory that consistently explains the response to defoliation across woody and herbaceous species. It may be possible to explain the difference between herbaceous and woody species in terms of the carbon–nutrient balance: nutrient availability might limit growth of woody species while carbon limits growth of herbaceous species. However, Chapin and McNaughton (1989) reported that phosphorus and nitrogen concentration in leaves of several grassland species increased after defoliation even under severe phosphorus limitation. Furthermore, the present study shows that LNC increased even in plants at the low nutrient treatment, which had not received fertilizer (Fig. 3).

We suggest two possible explanations for the increase in LNC. One is a developmental limitation. After defoliation, plants (particularly grazed grass species) immediately produce new leaves so as to recover photosynthetic activity. If plants need to produce photosynthesizing leaves immediately after defoliation, they will make smaller or thinner leaves that finish their expansion in a shorter period (Miyazawa et al., 1998). Such leaves may contain smaller amounts of cell wall materials and secondary compounds with a LNC that is higher than that before defoliation. This may not be the case for leaves that emerge one year after defoliation, where plants are producing leaves under low nutrient availability. This explains the above observations: studies for woody species tended to use leaves emerging in the year following the defoliation treatment (e.g. Tuomi et al., 1984; Kamata et al., 1996), while those for herbaceous species used leaves emerging in the same year as defoliation. Heichel and Turner (1983) observed that the photosynthetic capacity increased in leaves that emerged in the year of defoliation in woody species. However, increase in LNC and photosynthetic capacity after defoliation has been found even in a leaf that was partly defoliated (Alderfer and Eagles, 1976), which may be free from the developmental constraint.

The other possibility is a contribution of respiration. As discussed above, respiratory loss after defoliation may be important for the carbon–nutrient balance in the plants. Because woody tissues have a low respiratory activity compared with other tissues, woody species may have a smaller loss of carbon per plant mass than herbaceous species. Therefore, leaf production of woody species may be limited by nutrients rather than carbon, while that of herbaceous species is limited by carbon. This hypothesis explains our results on current-year seedlings, which may be closer to herbaceous plants in terms of the amount of woody tissues; most studies for woody species have used plants older than 3 years (e.g. Tuomi et al., 1984, 1988; Kamata et al., 1996). It should be noted that LNC was higher in defoliated plants when compared at the same PNC (Fig. 5B). Although LNC was strongly correlated with PNC, defoliation significantly altered the relationship between LNC and PNC. LNC does not necessarily represent nitrogen status when defoliated plants are compared with control plants.

Biomass allocation in defoliated plants

In control plants, a single, positive correlation was found between LMR and PNC irrespective of plant size and nutrient availability (Fig. 5A), as has been shown in previous studies (Hirose, 1986, 1987; Kachi and Rorison, 1989). This is regarded as an adaptive adjustment for resource acquisition: when nitrogen is limiting, plants allocate more biomass to roots to increase nitrogen uptake rates per plant (Hirose, 1987; Hilbert, 1990). For defoliated plants, a similar correlation was found but the regression was significantly different in the y-intercept, i.e. defoliated plants had a lower LMR when compared at the same PNC. The simplest interpretation of the difference may be that leaf growth in defoliated plants was insufficient to recover the leaf mass ratio to the level in control plants. However, plant mass of defoliated plants increased from 0·6 to 2·3 g in the high nutrient treatment (Fig. 1A), which seems enough to recover the balance between the leaf and other organs (Fig. 4).

The ratio of leaf to root mass is further analysed in Fig. 8, where mass ratios of leaves to roots that were newly produced after defoliation are plotted against PNC. For control plants, the ratio was simply calculated from attached leaves and roots at the time of the harvests. There was little difference between defoliated and control plants in the relationship between the leaf-to-root ratio and PNC. This result suggests that defoliated plants regulate biomass allocation so as to maintain the balance between newly produced tissues, rather than the balance between total tissues. Such allocation could occur if old tissues become less active after defoliation. When new leaves are produced after defoliation, resources are translocated from the roots, which may accelerate the senescence of old roots. Hodgkinson and Becking (1977) observed increased root mortality after defoliation in cowpea plants. After establishment of new leaves, defoliated plants might have lowered root activities per root mass and allocated more biomass to roots in order to balance root and leaf activities.

Fig. 8.

The ratio of leaf to root mass newly produced after defoliation as a function of plant nitrogen concentration. Open circles denote control plants (mean of 6 plants), where total mass of harvested roots and leaves is used. Closed circles denote defoliated plants, where the leaf mass 63 d after defoliation is used. The root mass 63 d after defoliation minus the minimum root mass observed though the experiment was used for defoliated plants. Minimum root mass was found at 21 d after defoliation (middle and high nutrient availability) and at 42 d after defoliation (low nutrient availability). The regression line is y = −0·185 + 72·7x (r² = 0·87, P < 0·0001).

Chemical defence

Although many studies have shown that defoliated plants accumulate defensive compounds such as tannins in leaves (Tuomi et al., 1984, 1987; Mihaliak and Lincoln, 1989; Kamata et al., 1996), we observed significantly lower concentrations of tannins and total phenolics in defoliated plants (Table 1). Together with the increased LNC, we suggest that defoliation did not induce defensive responses in the present experiment. It is remarkable that there was a strong correlation between tannin concentration and LNC irrespective of growth conditions and of defoliation treatment (Fig. 6B), suggesting that the carbon–nutrient balance at the leaf level controls the concentration of tannins. A similar correlation was found in previous studies (Haukioja et al., 1985; Tuomi et al., 1988; Dormann, 2003), although the generality of the carbon–nutrient balance hypothesis is suspected (Hamilton et al., 2001; Koricheva, 2002; Nitao et al., 2002).

Conclusions

The present study has shown that defoliation altered carbon–nitrogen balance and affected subsequent biomass allocation. Although PNC was reduced immediately after defoliation, it recovered to levels similar to or higher than that of control plants due to nitrogen uptake and respiratory consumption of biomass. When compared at the same PNC, LMR was lower in defoliated plants. However, the newly produced leaf-to-root ratio as a function of PNC was not different between defoliated and control plants. This result suggests that defoliated plants regulate biomass allocation so as to maintain the balance between newly produced tissues, rather than the balance between total tissues. As the concentration of defensive compounds was negatively correlated with LNC, we suggest that the carbon–nutrient balance strongly influences both biomass allocation and chemical defence in Quercus serrata. We further suggest that defoliation induces accumulation of these chemicals only when it decreases the nitrogen status in the plant.