Why do fast- and slow-growing grass species differ so little in their rate of root respiration, considering the large differences in rate of growth and ion uptake?

Authors

  • I. Scheurwater,

    1. Department of Plant Ecology and Evolutionary Biology, Utrecht University, PO Box 80084, 3508 TB Utrecht, The Netherlands, and,
    Search for more papers by this author
  • C. Cornelissen,

    1. Department of Plant Ecology and Evolutionary Biology, Utrecht University, PO Box 80084, 3508 TB Utrecht, The Netherlands, and,
    Search for more papers by this author
  • F. Dictus,

    1. Department of Plant Ecology and Evolutionary Biology, Utrecht University, PO Box 80084, 3508 TB Utrecht, The Netherlands, and,
    Search for more papers by this author
  • R. Welschen,

    1. Department of Plant Ecology and Evolutionary Biology, Utrecht University, PO Box 80084, 3508 TB Utrecht, The Netherlands, and,
    Search for more papers by this author
  • H. Lambers

    1. Department of Plant Ecology and Evolutionary Biology, Utrecht University, PO Box 80084, 3508 TB Utrecht, The Netherlands, and,
    2. Plant Sciences, Faculty of Agriculture, The University of Western Australia, Nedlands, WA 6907, Australia
    Search for more papers by this author

I. Scheurwater Fax: + 31 30 2518366; e-mail: F.I.Scheurwater@biol.uu.nl

Abstract

Herbaceous plants grown with free access to nutrients exhibit inherent differences in maximum relative growth rate (RGR) and rate of nutrient uptake. Measured rates of root respiration are higher in fast-growing species than in slow-growing ones. Fast-growing herbaceous species, however, exhibit lower rates of respiration than would be expected from their high rates of growth and nitrate uptake. We investigated why the difference in root O2 uptake between fast- and slow-growing species is relatively small. Inhibition of respiration by the build-up of CO2 in closed cuvettes, diurnal variation in respiration rates or an increasing ratio of respiratory CO2 release to O2 uptake (RQ) with increasing RGR failed to explain the relatively low root respiration rates in fast-growing grasses. Furthermore, differences in alternative pathway activity can at most only partly explain why the difference in root respiration between fast- and slow-growing grasses is relatively small. Although specific respiratory costs for maintenance of biomass are slightly higher in the fast-growing Dactylis glomerata L. than those in the slow-growing Festuca ovina L., they account for 50% of total root respiration in both species. The specific respiratory costs for ion uptake in the fast-growing grass are one-third of those in the slow-growing grass [0·41 versus 1·22 mol O2 mol (NO3)–1]. We conclude that this is the major cause of the relatively low rates of root respiration in fast-growing grasses.

INTRODUCTION

When herbaceous plants are grown with free access to nutrients, they exhibit inherent differences in relative growth rate (RGR) and rate of nutrient uptake (Poorter & Remkes 1990; Poorter et al. 1991; Garnier 1992; Van der Werf, Welschen & Lambers 1992). For example, fast-growing species exhibit RGR values that are 3-fold higher than those of slow-growing species (Poorter & Remkes 1990). Similarly, the rate of net nitrate uptake (NNUR) is 4- to 6-fold higher in fast-growing species than in slow-growing ones (Poorter et al. 1991). Rates of root respiration (rt, O2 uptake per unit root mass and time) are expected to be higher also, since more respiratory energy is needed for growth and ion uptake.

Although the measured rates of root respiration are indeed approximately 1·7-fold higher in fast-growing species than in slow-growing ones, they are not as high as predicted from their high rates of growth and ion uptake. To predict rates of root respiration, assumptions have to be made about the specific respiratory costs for energy-requiring processes. Poorter et al. (1991) calculated the expected rates of root respiration in fast-growing and slow-growing herbs using the differences in RGR and NNUR and assuming the same specific respiratory costs for ion uptake and for growth and maintenance of biomass in fast- and slow-growing species. To do this they used the specific respiratory costs determined for two slow-growing Carex species (Van der Werf et al. 1988). These calculations suggested that fast-growing species should exhibit 3-fold higher rates of respiration than their slow-growing counterparts, rather than the measured 1·7-fold higher rates.

In our study, we investigate why roots of fast-growing species respire at a lower rate than expected from their high RGR and NNUR. To address this question several possibilities have been explored, using nine grass species differing in maximum RGR. Firstly, we investigated whether fast- and slow-growing species differ in the extent to which respiration is inhibited by the CO2 that accumulates during measurements of O2 uptake in closed cuvettes. High CO2 concentrations have been reported to inhibit respiration (Bunce 1994; Qi, Marshall & Mattson 1994; Wullschleger, Ziska & Bunce 1994; Drake et al. 1996; Burton et al. 1997). The degree of inhibition of root respiration might be greater for fast-growing species, because of their higher rate of root respiration and, consequently, greater build-up of CO2. Other studies have found, however, that increased CO2 has little effect on root respiration (Williams et al. 1992; Bouma et al. 1997a,b) or even stimulates respiration (Williams et al. 1992). To our knowledge, no study has investigated the effect of high CO2 concentrations on root respiration rates of fast- and slow-growing species.

Secondly, we assess whether diurnal changes in root respiration contribute to the discrepancy between the measured and theoretical rates of respiration. In Poorter et al. (1991), the theoretical rates of root respiration were calculated assuming no diurnal variation in root respiration and rates of respiration were measured during the light period only. Although Veen (1977) did not observe differences in root respiration of maize between the light and dark period, some other studies have reported diurnal variation in root respiration (Neales & Davies 1965; Hansen 1980; Lambers, Layzell & Pate 1980). To our knowledge there is no information on whether fast- and slow-growing species differ in their respiration rates during the light and dark period. The difference in measured root respiration rates between fast- and slow-growing species might be greater than that reported by Poorter et al. (1991) if the contrasting species were to differ in the degree of diurnal variation. For example, the difference in root respiration per day between fast- and slow-growing species would have been greater if the respiration were substantially lower in the dark than in the light for slow-growing species, but not for fast-growing species.

A third possible explanation of the relatively low rate of root respiration in fast-growing species might be that they exhibit a lower activity of the non-phosphorylating alternative pathway, and a higher activity of the phosphorylating cytochrome path, compared with that in slow-growing species. The alternative pathway decreases the efficiency of respiratory ATP-production (McIntosh 1994). Therefore, a low contribution of the alternative path in plant roots will result in a relatively high rate of ATP production per unit mass and time. The difference in the rate of root respiration between fast- and slow-growing species might therefore be greater when expressed in units of ATP production than the difference in rates calculated in units O2 uptake. Poorter et al. (1991) and Van der Werf et al. (1992) concluded that the activity of the alternative path in roots does not correlate with the RGR of a species. However, their conclusion was based on experiments using inhibitors such as cyanide and salicylhydroxamic acid (SHAM) to determine the partitioning of electrons between the alternative and the cytochrome path in root respiration. Recently, it has become evident that inhibitors cannot be used to assess the activity of the alternative path (Day et al. 1996; Hoefnagel et al. 1995; Millar et al. 1995). On the basis of inhibitor studies, however, we do know there is a wide variation in the cyanide resistance of root respiration, which is a rough estimate of the maximum activity of the alternative path. Moreover, there is substantial variation in the sensitivity to SHAM, which gives a minimum estimate of the activity of the alternative path (Atkin, Villar & Lambers 1995; cf. Lambers, Atkin & Scheurwater 1996). Therefore, although inhibitors cannot be used to give precise estimates of the contribution of the cytochrome and alternative path to root respiration, they do provide a range of possible activities.

A fourth possibility is that fast- and slow-growing species differ in their respiratory quotient (RQ, the ratio of respiratory CO2 release to O2 uptake). Transfer of electrons to acceptors other than O2 (e.g. NO3) increases the RQ (Lambers et al. 1996). A higher RQ can imply higher glycolytic activity and/or higher activity of the oxidative pentose phosphate pathway. If NADH from glycolysis and NADPH from the pentose phosphate pathway are used to reduce nitrate and nitrite, respectively, CO2 is produced without concomitant oxygen consumption, which increases the RQ. This might lead to a higher rate of ATP production per mole of O2 consumed. RQ values range from 0·75 to 1·7 (cf. Lambers et al. 1996). The RQ for the complete oxidation of hexose to CO2 and H2O is 1, whereas complete oxidation of substrates which are more oxidized than hexose yields an RQ > 1 (Lambers, Scheurwater & Millenaar 1997). Since sugars are the major carbon compounds transported in the phloem (Zimmermann & Ziegler 1975) it is unlikely that the roots of fast- and slow-growing herbaceous species differ much with respect to the major substrate they respire. During synthesis of biomass, both carboxylating and decarboxylating reactions occur, which also affect the RQ. We investigated whether grass species with a high RGR exhibit a higher RQ and, possibly, have a slightly higher rate of ATP production per mole of O2, compared with slow-growing grasses.

Finally, fast- and slow-growing species might differ in the efficiency of their use of respiratory energy. As stated earlier, root respiration (rt, mmol O2 g–1 DM d–1) provides energy for ion uptake and growth and maintenance of biomass in plant roots, which can be described by the following equation (cf. Van der Werf, Poorter & Lambers 1994):

inline image

where rm (mmol O2 g–1 DM d–1), cu[mol O2 (mol NO3)–1] and cg (mmol O2 g–1 DM) are the specific respiratory costs for maintenance, ion uptake and growth, respectively. If fast-growing species have lower specific respiratory costs for growth, maintenance and/or ion uptake compared with slow-growing species, this would explain the relatively low rates of root respiration observed for the fast-growing species.

What is known about the specific respiratory costs of growth, maintenance and ion uptake in fast- and slow-growing species? Specific respiratory costs for growth and maintenance of biomass have frequently been determined in different plant organs (for reviews see Amthor 1984; Lambers 1985), including leaves (Villar & Merino 1994), roots (Blacquière 1987) and fruits (Marcelis & Baan Hofman-Eijer 1995). Less information is available on the separation of specific respiratory costs for root growth into costs for the synthesis of biomass and for the uptake of ions. Veen (1980) and Bouma, Broeckhuysen & Veen (1996) determined specific respiratory costs for growth, maintenance and ion uptake in maize and potato roots, respectively, using a multiple regression analysis with root RGR and NNUR as independent variables. The often fixed ratio between root growth and ion uptake was disturbed by pruning of the roots and/or the shoots. The specific costs for root growth, maintenance and ion uptake in two Carex species, differing in their rate of root respiration, have been determined by Van der Werf et al. (1988). They used the same analysis as Veen (1980), with the exception that variation in RGR and NNUR was due to ontogeny, rather than pruning. Poorter et al. (1991) calculated specific respiratory costs for growth in the roots of 24 herbaceous species differing in their maximum RGR, using information on the biochemical composition and values for the oxygen consumption necessary to construct the different compounds (cf. Penning de Vries, Brunsting & Van Laar 1974). The estimated growth costs increased slightly with increasing RGR. Moreover, Poorter et al. (1991) and Van der Werf et al. (1992) speculated that the specific respiratory costs for ion uptake are lower in roots of fast-growing species than in those of slow-growing species. To our knowledge, however, no other information exists on the relationship between specific respiratory costs and RGR. We therefore determined the specific respiratory costs for maintenance and growth plus ion uptake in the roots of two of our grasses, one fast- and one slow-growing species, using a linear regression approach (cf. Lambers & Van der Werf 1988). Assuming the same costs for growth as those previously determined for these species by Poorter et al. (1991), we then arrived at the specific costs for ion uptake in these two species.

MATERIALS AND METHODS

Plant growth

The following monocotyledonous species were used in the experiments: Dactylis glomerata L., Deschampsia flexuosa L., Festuca ovina L., Holcus lanatus L., Poa alpina L., P. annua L., P. compressa L., P. costiniana J. Vickery, P. fawcettiae L. and P. trivialis L. Seeds of D. flexuosa and F. ovina were collected in a heathland area in the Netherlands at the ‘Edese Heide’ and ‘Uddelse Heide’, respectively. F. ovina seeds were kindly provided by Dr H. Poorter, Utrecht University, the Netherlands. Seeds of D. glomerata were commercially obtained from Van Engelen Zaden B.V. (Vlijmen, the Netherlands) and H. lanatus and P. annua seeds from Kieft B.V. (Blokker, the Netherlands). Seeds of the other Poa species were kindly provided by Dr O.K. Atkin, Australian National University, Canberra, Australia and obtained as described in Atkin, Botman & Lambers (1996).

Growth room conditions were as follows: irradiance: 500 μmol m–2 s–1 (using high-pressure mercury lamps, HPI-T 400 W, Philips B.V., Eindhoven, the Netherlands); temperature: 20 °C; relative humidity: 70%; light period: 14 h. Seeds of D. flexuosa were stratified for 6 weeks prior to germination. All seeds were germinated in the growth room in Petri dishes on moistened filter paper. Subsequently, the seedlings were transferred to sand moistened with half the strength of the following nutrient solution: 795 mmol m–3 KNO3, 602·5 mmol m–3 Ca(NO3)2, 270 mmol m–3 MgSO4, 190 mmol m–3 KH2PO4, 41 mmol m–3 Fe-EDTA, 20 mmol m–3 H3BO3, 2 mmol m–3 MnSO4, 0·85 mmol m–3 ZnSO4, 0·25 mmol m–3 Na2MoO4, and 0·15 mmol m–3 CuSO4 (Poorter & Remkes 1990). After 7–14 d of establishment the seedlings were transferred to 32 dm3 tanks containing an aerated nutrient solution, as described above. The pH of the nutrient solution was adjusted regularly to about 5·8. The nutrient solution was changed once or twice a week, depending on the size of the plants.

Root respiration

O2 uptake

The rate of O2 uptake by the roots was measured by placing excised, intact roots in an airtight cuvette, containing an air-saturated nutrient solution similar to that in which the plants were grown, except that Fe was absent. This nutrient solution was buffered with 10 mol m–3 MES (pH 5·8). Excision does not have a major effect on root respiration when assessed within half an hour after severing of the shoot, either in slow-growing or in fast-growing plants (Lambers, Van der Werf & Bergkotte 1993). Different sizes of cuvette were used, depending on the size of the roots. Root respiration was measured polarographically, using a Clark-type oxygen electrode (Yellow Springs Instruments, OH, USA) (Lambers et al. 1993).

CO2 release

The rate of root CO2 release was measured using intact plants. Roots were placed in a separate cuvette from the shoots. The root compartment contained an aerated nutrient solution similar to that in which the plants were grown, except that Fe was absent. This nutrient solution was buffered with 10 mol m–3 MES (pH 5·8). Root CO2 release was measured using an infra-red gas analyser (ADC, model 225 MK3, Hoddesdon, UK) in a differential mode (Poorter & Welschen 1993).

Nitrogen analysis

The total N content of the oven-dried samples was determined with a C-H-N analyser (Carlo-Erba, model 1106, Milan, Italy) using combustion gas chromatography (Pella & Colombo 1973).

Inhibition of respiration by the build-up of CO2

The effect of the inorganic carbon concentration (CO2 plus bicarbonate, from now on referred to as CO2 concentration) in the root medium on the root respiration of D. glomerata, F. ovina, H. lanatus, P. alpina and P. annua was measured in the IRGA set-up (see above). The CO2 concentration of the root medium was calculated according to Umbreit (1957) and was increased by increasing the CO2 concentration of the incoming air from ambient (on average 15·5 mmol m–3 or 372 cm3 m–3) to 29·4 mmol m–3 (705 cm3 m–3) and 38·7 mmol m–3 (930 cm3 m–3). Four plants were used per species. At each concentration respiration was allowed to achieve a steady state before rates were recorded (≈ 15 min).

Differences in root respiration within a species, between concentrations, were tested with one-way analysis of variance with repeated measures, using the SAS statistical package (SAS 1988).

Respiratory quotient

The respiratory quotient (RQ) was determined in the middle of the light period. Root respiration rates were measured as described above. Four replicates per O2 uptake and four different replicates per CO2 release measurement were used to determine the RQ for P. alpina, P. compressa, P. fawcettiae and P. trivialis. For D. glomerata, D. flexuosa, F. ovina and H. lanatus eight replicates were measured.

The standard deviation of the RQ was calculated using the approximation as proposed by Armitage & Berry (1991).

Diurnal fluctuations in root respiration

Diurnal variation in root respiration was measured as oxygen consumption (see above), at seven points during the 24 h day in D. glomerata, F. ovina and H. lanatus. For D. flexuosa only four points were chosen: 45 min, 4 h 15 min and 7 h 45 min after the lights were switched on and 45 min after the lights were switched off. For the three other species the additional points were: 11 h 45 min after the lights were switched on and 4 h 15 min and 7 h 45 min after the lights were switched off. At each time eight replicates were used per plant species.

Differences in root respiration within a species, between time points, were tested with one-way analysis of variance using the SAS statistical package (SAS 1988). If significant differences existed a Duncan a posteriori test was performed (Winer 1971).

Calculation of the RGR

The RGR of the plants used in the three above-mentioned experiments was calculated from two harvests with six to eight replicates each. The first and second harvest took place 3 to 4 d before and after measuring the rate of root respiration, respectively. RGR was determined as the slope of the natural logarithm of total plant dry mass versus time (Hunt 1982).

Specific respiratory costs

Specific respiratory costs for growth and maintenance of root biomass and for ion uptake were determined for the roots of D. glomerata and F. ovina. These species were chosen because they differ ≈ 2-fold in maximum RGR. Plants were grown closely together (132 plants per m2), resulting in increased mutual shading with time and hence departure from exponential growth. At regular intervals, every 4–5 d for D. glomerata and every 7 d for F. ovina, six plants were harvested and the rate of root respiration was measured as O2 uptake (see above). On the first and the last days of the experiments a double harvest was taken. Each time a plant was removed it was replaced by another plant to maintain the level of mutual shading. Dry mass of shoot and root was determined after oven-drying for 48 h at 70 °C.

The RGR of the roots was calculated according to Poorter (1989), skipping one harvest each time and using the average of the ln-transformed root dry mass values. The RGR value of an interval was assigned to each day of that interval, which resulted in two or three RGR values per day. These values were averaged and used in a stepwise regression (SAS 1988), to obtain an equation for the time trend of RGR. With this equation the RGR for each harvest day was calculated.

The net rate of nitrogen uptake per unit root dry mass (NNUR, mmol N g–1 DM d–1) was calculated, skipping one harvest each time, using the following equation (White 1972):

inline image

where lnMn¯¯¯¯¯¯ is the average of the natural logarithm of the root mass values at time tn (day) and Mn¯¯¯ and Nn¯¯¯ are the average root mass (g DM) and the average total amount of nitrogen in the plant (mmol N plant–1) at that time, respectively. Subsequently the same approach was followed as that for the RGR calculations.

With a linear regression approach (cf. Lambers, Szaniziawski & De Visser 1983) the specific respiratory costs for maintenance of biomass and the specific respiratory costs for growth, including ion uptake in the roots were determined, using the following equation:

inline image

Where the slope of the regression of the total rate of root respiration (rt, mmol O2 g–1 DM d–1) versus the RGR gives the specific respiratory costs for growth, including ion uptake (c(g+u), mmol O2 g–1 DM). The y-intercept of the regression line gives the specific respiratory costs for maintenance of biomass (rm, mmol O2 g–1 DM d–1). The significance of the regression was tested using the SAS statistical package and significant differences between the slopes of the regression lines were tested using an analysis of covariance (SAS 1988).

To separate costs for growth from those for ion uptake, we used the specific respiratory costs for growth as determined by Poorter et al. (1991), which were 6·50 and 6·25 mmol O2 g–1 DM for the fast-growing D. glomerata and the slow-growing F. ovina, respectively. Poorter et al. (1991) used a light intensity of 315 μmol m–2 s–1 to grow their plants, while we used 500 μmol m–2 s–1. All other growth conditions were the same, including the nutrient solution. Multiplying these costs for growth by the RGR values obtained for the two species resulted in the rate of root respiration necessary for growth. By subtracting the portion of root respiration associated with growth from the total root respiration, the rate of root respiration for maintenance and ion uptake was found. The slope of the regression of root respiration rates for maintenance plus ion uptake against the rates of net nitrate uptake gave the specific respiratory costs for ion uptake, expressed per mole of nitrate.

RESULTS AND DISCUSSION

Differences in rate of root respiration

Figure 1 shows the data on root respiration, as determined for nine grass species differing in their relative growth rate (RGR) when grown with free access to nutrients. Whenever RGR and root respiration were determined more than once, the species’ average value [± standard deviation (SD)] is given in the graph. For the Poa species the RGR values as determined in Atkin et al. (1996) were used. On average the fast-growing species had a mere 1·4-fold higher rate of root respiration than the slow-growing ones, despite their 2-fold higher RGR (Fig. 1) and their more than 3-fold higher rate of net nitrate uptake (NNUR) (data not shown). A similar result was obtained by Poorter, Remkes & Lambers (1990) and Poorter et al. (1991) who compared rates of root respiration of 24 herbaceous species which differed 3-fold in RGR and 4-to 6-fold in NNUR. The rate of root respiration differed only by a factor of 1·7 between fast- and slow-growing species. Furthermore, Atkin et al. (1996), comparing six alpine, subalpine and lowland Poa species, found a 2-fold difference in RGR, but the rate of root respiration was similar for all species.

Figure 1.

. Rate of root respiration (mmol O2 g–1 DM d–1) against the RGR (mg g–1 d–1). Whenever RGR and root respiration were determined more than once, the species’ average value (± standard deviation) is given. For the Poa species the RGR values as determined in Atkin et al. (1996) were used. Species: Dactylis glomerata (▴), Deschampsia flexuosa (•), Festuca ovina (◆), Holcus lanatus (▪), Poa alpina (•), P. compressa (▾), P. costiniana (▪), P. fawcettiae (▴) and P. trivialis (◆).

Inhibition of root respiration by the build-up of CO2

What might be the cause of the relatively low rates of root respiration in the fast-growing species, considering their rather high rates of growth and ion uptake? We have explored several possible causes. Firstly, respiration may have been inhibited by the accumulation of CO2 in the closed cuvettes during the measurements. During our measurements, the inorganic carbon concentration (CO2 plus bicarbonate) increased from 14·5 to ≈ 25–140 mmol m–3 in the first 15 min, depending on the size of the roots, the rate of O2 uptake and the volume of the cuvette. The relatively low rates of O2 uptake exhibited by our fast-growing species (Fig. 1) might therefore have been due to greater inhibition by CO2 than in the slow-growing species. To test whether this was the case, we exposed roots to three different CO2 concentrations. Root respiration rates were measured as CO2 release per unit mass and time using an infra-red gas analyser. The IRGA system allowed us to maintain the CO2 concentration at a constant level. The CO2 level of the air entering the cuvette was increased from ambient (on average 15·5 mmol m–3 or 372 cm3 m–3) to 29·4 mmol m–3 (705 cm3 m–3) and then to 38·7 mmol m–3 (930 cm3 m–3). This resulted in a change in the CO2 concentration of the root medium from 14·6 to 27·7 and 36·4 mmol m–3, respectively. Depending on the species, the rate of root respiration was either slightly stimulated or inhibited (Fig. 2). The CO2-induced changes in respiration did not exceed 10%, however, and were not significant (P < 0·05, Fig. 2). This contrasts with several studies that have reported substantial inhibition of respiration by high CO2 concentrations (Amthor 1991; Bunce 1994; Drake et al. 1996; Wullschleger et al. 1994) including root respiration (Burton et al. 1997; Nobel & Palta 1989; Qi et al. 1994). However, other studies report no effect on respiration or a stimulation of respiration upon exposure to CO2-enriched air. In Williams et al. (1992), doubling of the ambient CO2 concentration (from 350 to 700 cm3 m–3) resulted in a significant increase in leaf respiration in barley and no change in leaf respiration of Agrostis capillaris and Poa alpina. Furthermore, Bouma et al. (1997a,b), determining rates of root respiration in soil, found that root respiration rates of citrus and bean were not affected by soil CO2 concentrations in a range of 400–25000 cm3 m–3.

Figure 2.

. The change in the rate of root respiration (%) (± SE, n = 4), after increasing the CO2 concentration, versus the CO2 concentration (CO2 plus bicarbonate) of the root medium (mmol m–3). Species: Dactylis glomerata (▴), Festuca ovina (◆), Holcus lanatus (▪), Poa annua (▾) and P. alpina (•).

The inorganic carbon concentration in our closed cuvettes after 15 min of O2 uptake sometimes exceeded the maximum that could be reached in the IRGA system. Therefore, to further assess the effect of the gradual build-up of CO2 on root O2 uptake in our closed cuvettes, we compared the rates of O2 uptake at different times after placing roots in the cuvettes. We assumed that a gradual build-up of CO2 is fully responsible for any decrease of the rate of O2 uptake with time. The rate of root O2 uptake decreased slightly with time in all species. Taking the rate 5–7 min after the measurements were started as 100%, the rates decreased to 95 ± 4·8% (SD, n = 20) and then 91 ± 6·4% (n = 20), at 10–12 and 15–17 min after the measurements were started, respectively. However, the repeated measures ANOVA (SAS 1988), with repeated measurements on the rate of root respiration, showed that the interaction term (species × time) was not significant. Therefore, while the rate of root O2 uptake decreased by on average 9% over 15 min in all species, the fast- and slow-growing grass species did not respond differently in this respect. We conclude that the lower rates of root respiration in fast-growing grass species than in slow-growing species were not the result of greater inhibition by the build-up of CO2 in the cuvettes.

Diurnal variation in root respiration

The data in Fig. 1 pertain to measurements made on intact roots detached from the shoots during the light period only. The relatively small differences in respiration rates between the slow- and fast-growing species (Fig. 1) might have been much greater when expressed per day, if respiration rates of the slow-growing species were substantially lower during the dark period or if the fast-growing species exhibited substantially higher rates during the dark period. As stated in the Introduction, both diurnal variation in root respiration and no diurnal variation in root respiration have been reported in separate studies. Therefore, we measured the diurnal pattern of root respiration as the diurnal variation in oxygen consumption for two fast-growing and two slow-growing grass species (Fig. 3). Diurnal fluctuations in root respiration rate were observed, but there was no distinct difference between the light and dark periods, except for H. lanatus. This fast-growing grass species had higher rates of root respiration during the light than during the dark period (Fig. 3), which results in an even smaller difference in the daily rate of respiration between fast- and slow-growing grasses. Moreover, the daily rates of respiration calculated by integrating the data in Fig. 3 were similar to the daily rates calculated from the single light-period measurements (Fig. 1): the difference in daily rates of respiration, calculated by the two methods, was ≈ 10% for H. lanatus and even less for the other three species. Therefore, the fast- and slow-growing grass species did not differ enough in their diurnal pattern of root respiration to explain the relatively low rates of root respiration of fast-growing grass species (Fig. 1).

Figure 3.

. The diurnal variation in the rate of root respiration (nmol O2 g–1 DM s–1) (n = 8) for four grass species differing in RGR. The dark bar represents the 10 h dark period. Species: Dactylis glomerata (▴), Deschampsia flexuosa (•), Festuca ovina (◆) and Holcus lanatus (•).

Alternative pathway activity

We also investigated whether the fast-growing species produce their ATP more efficiently than their slow-growing counterparts. The upper and lower limits of alternative pathway activity (and hence a range for the rate of ATP production) were estimated by exposing roots to cyanide and SHAM, respectively (Lambers et al. 1997). We assumed that 1 mole of ATP is produced per atom of O consumed by the alternative pathway (compared to 3 moles ATP per atom of O via the cytochrome path): in this way, we could calculate the theoretical upper and lower limits of ATP synthesis in the fast- and slow-growing species. We applied this method to two of the species used in our study: the slow-growing F. ovina and the fast-growing D. glomerata. When expressed as the rate of O2 consumption, D. glomerata exhibited root respiration rates that were 1·2-fold higher than the rates exhibited by F. ovina. Based on its rates of growth and ion uptake, however, D. glomerata is expected to have a 2·3 times higher rate of root respiration than F. ovina (Poorter et al. 1991). The rate of ATP production of D. glomerata, estimated using the minimum estimate of the contribution of the alternative pathway (3%), was 27·9 mmol ATP g–1 DM d–1, which is 1·5-fold higher than that of F. ovina using a maximum estimate of the contribution of the alternative path (37%, i.e. 18·7 mmol ATP g–1 DM d–1). The difference between the species in respiration rates (expressed as rates of ATP production) might therefore be greater than that suggested by the rates of O2 consumption. However, the maximum possible differences in ATP production are too small to account fully for the expected differences in respiration rates. We therefore conclude that differences in alternative pathway activity can at most partly explain why roots of fast-growing grass species respire at a lower rate than expected when root respiration is expressed as O2 consumption.

Respiratory quotient

The degree to which differences in the respiratory quotient (RQ) contribute to the relatively small difference in root O2 uptake between fast- and slow-growing species was also investigated. Assuming that fast- and slow-growing herbaceous species do not differ in the major substrate they use in respiration, the factors affecting the RQ are the transfer of electrons to acceptors other than O2 (e.g. NO3 and NO2) and the carboxylating and decarboxylating reactions that occur during the synthesis of biomass (Lambers et al. 1996). Figure 4 shows a positive correlation between RQ and RGR. The average RQ value in the fast-growing species is 1·3 compared with 1·0 in the slow-growing species (Fig. 4). A higher RQ is probably due to a higher glycolytic activity and a higher activity of the oxidative pentose phosphate pathway. If NAD(P)H from these pathways is used for the reduction of nitrate to nitrite and subsequently to ammonium, CO2 is produced without concomitant O2 consumption, which leads to a higher RQ. Use of NADH produced in glycolysis for the reduction of nitrate to nitrite instead of O2 leads to a slightly higher ATP production per mole of O2 consumed. Assuming that hexose is the substrate used in respiration, however, taking into account that NAD(P)H from glycolysis and the pentose phosphate pathway are used to reduce nitrate to ammonium and using the average minimum estimate of alternative pathway activity for fast-growing species from Lambers et al. (1997), i.e. 17%, a RQ of 1·3 corresponds to an ATP production per mole of O2 consumed which is only marginally higher than that at a RQ of 1·0. Therefore, differences in RQ cannot explain why roots of fast-growing grass species respire at a lower rate than expected from their relatively high rates of growth and nitrate uptake.

Figure 4.

. The respiratory quotient (RQ) (± SE, n = 4 for the Poa species and n = 8 for the other species), the ratio of the moles of CO2 released and the moles of O2 consumed, versus RGR (mg g–1 d–1) in the roots of eight grass species (for individual species’ symbols see the legend of Figure 1). For the Poa species the RGR values as determined in Atkin et al. (1996) were used.

Specific respiratory costs

The lower than expected respiration rates exhibited by the fast-growing species might also result from more efficient use of respiratory energy. We expected the specific respiratory costs for growth, maintenance and/or the uptake of ions to be lower in the roots of the fast-growing species than in those of the slow-growing ones. Using a linear regression approach (evaluated in Lambers et al. 1983) (Fig. 5a,b), specific respiratory costs for growth, maintenance and ion uptake were estimated for the roots of a fast-growing (D. glomerata) and a slow-growing grass species (F. ovina). We were unable to use the multiple regression method (cf. Van der Werf et al. 1988) to determine the specific respiratory costs because this requires the variables RGR and NNUR to be independent, which was not the case in our study (the correlation coefficient of RGR against NNUR was 0·99 for both D. glomerata and F. ovina). The linear regression of total root respiration rate against RGR was significant for both species (P < 0·0001, Fig. 5a,b). The specific respiratory costs for maintenance (rm; ± standard error of the regression coefficient), represented by the y-intercept of the regression line, were slightly higher in the roots of the fast-growing species; 2·2 ± 0·18 mmol O2 g–1 DM d–1 (Fig. 5a) compared with 1·8 ± 0·23 mmol O2 g–1 DM d–1 for the slow-growing species (Fig. 5b). These maintenance costs are 4 times higher than those in roots of Carex species (Van der Werf et al. 1988) and cork oak (Mata et al. 1996) and twice as high as the maintenance costs in potato roots (Bouma et al. 1996). They are similar to the specific respiratory costs for maintenance in barley genotypes, however, as estimated by Bloom, Sukrapanna & Warner (1992). Specific respiratory costs for maintenance of biomass accounted for 50% of the total rate of root respiration in both D. glomerata and F. ovina. In general, the portion of total root respiration that is associated with maintenance processes increases with decreasing RGR and can even result in a rate of root respiration that is insensitive to changes in growth rate and/or rate of ion uptake (cf. Lambers et al. 1998). This was not the case, however, in the present species. The slope of the regression line was significantly (P < 0·05) steeper for the slow-growing F. ovina, indicating higher specific respiratory costs for growth, including ion uptake, in the roots of F. ovina compared with those for the fast-growing D. glomerata.

Figure 5.

. Determination of the specific respiratory costs for growth in the fast-growing Dactylis glomerata and the slow-growing Festuca ovina. (a,b) The total rate of root respiration (rt, mmol O2 g–1 DM d–1) (± SE, n = 6) against RGR (mg g–1 d–1) for D. glomerata and F. ovina, respectively. Both regression lines are significant (P < 0·0001). For D. glomerata rt = 2·2 + 10·7E-3 (RGR and for F. ovina rt = 1·8 + 19·9E-3 × RGR. (c) The regression of root respiration for maintenance plus ion uptake (mmol O2 g–1 DM d–1) versus the net rate of nitrate uptake (NNUR) (mmol NO3 g–1 DM d–1). The dashed line represents D. glomerata and the dotted line F. ovina; rm, y-intercept, specific respiratory costs for maintenance of root biomass (mmol O2 g–1 DM d–1); cu, slope of the regression line, specific respiratory costs for ion uptake [mol O2 (mol NO3)–1].

How can we separate specific respiratory costs for ion uptake from those for growth? Poorter et al. (1991) described a method by which the costs associated with growth processes are estimated using information on the chemical composition of the roots and the oxygen consumption associated with constructing specific compounds. Using this approach, Poorter et al. (1991) arrived at specific respiratory costs for growth (cg) of 6·50 and 6·25 mmol O2 g–1 DM for the fast-growing D. glomerata and the slow-growing F. ovina, respectively. That is, specific respiratory costs for growth are slightly higher in the fast-growing grass species. Multiplying these growth costs by the RGR values provides an estimate of the root respiration necessary for growth.

We expected the fast-growing species to have lower specific respiratory costs for energy-requiring processes than slow-growing species. However, the fast-growing D. glomerata exhibited slightly higher costs for growth and maintenance of root biomass than the slow-growing F. ovina. Hence, specific respiratory costs for growth and maintenance of biomass cannot explain the relatively low rates of root respiration in fast-growing grasses. This suggests that fast-growing grass species have lower costs associated with ion uptake. By subtracting the portion of root respiration associated with growth from the total root respiration, the rate of root respiration for maintenance plus ion uptake can be calculated. Moreover, the costs associated with ion uptake can be determined by plotting the rate of root respiration for both processes against the rate of net nitrate uptake: the slope of such a plot provides an estimate of the specific respiratory costs for ion uptake (cu; expressed per mole of nitrate). 5Figure 5c provides such a plot for D. glomerata and F. ovina, with the specific respiratory costs for both species summarized in Table 1. The fast-growing species (D. glomerata) exhibits costs associated with ion uptake which are one-third of those in the slow-growing F. ovina[0·41 ± 0·013 and 1·22 ± 0·013 mol O2 (mol NO3)–1, respectively]. These results agree with the speculation presented in Poorter et al. (1991) and Van der Werf et al. (1992), that fast-growing species have lower specific respiratory costs for ion uptake. The present values are within the range of ion uptake costs calculated for the roots of Carex species, maize and potato (cf. Bouma et al. 1996). We conclude that the respiratory costs associated with ion uptake are clearly lower in fast-growing grasses than in slow-growing grasses.

Table 1.  . Specific respiratory costs for growth, maintenance and ion uptake in a fast-growing (Dactylis glomerata) and a slow-growing (Festuca ovina) grass species Thumbnail image of

How robust are our conclusions?

Poorter et al. (1991) used a light intensity of 315 μmol m–2 s–1 while the plants in our experiments were grown with 500 μmol m–2 s–1. What happens to the specific respiratory costs for ion uptake when the costs for growth as determined by Poorter et al. (1991) are an overestimation or an underestimation? To analyse this we added 25% to or subtracted 25% from the growth costs as determined by Poorter et al. (1991) and calculated the effect on the specific respiratory costs for ion uptake (Table 2). Literature data provide estimates for growth costs of roots from 4·9 to 10·9 mmol O2 g–1 DM and for ion uptake costs from 0·39 to 1·6 [mol O2 (mol NO3)–1] (cf. Bouma et al. 1996). These estimates were obtained by performing regression analyses with RGR and NNUR as independent variables and by calculating growth costs using information on the biochemical composition and values for the oxygen consumption necessary to construct the different compounds. The values of the specific respiratory costs for growth and ion uptake in Table 2 fit quite well into the range covered by literature data. Are the specific respiratory costs for ion uptake still substantially lower in the fast-growing species when the growth costs as determined by Poorter et al. (1991) are an overestimation or an underestimation? The costs for ion uptake (Table 2) are only 1·3-fold lower in the fast-growing species if the growth costs calculated by Poorter et al. (1991) are 25% overestimated in the fast-growing D. glomerata (cg– 25%) and 25% underestimated in the slow-growing F. ovina (cg + 25%), which is not a very likely scenario. In all other possible scenarios, specific respiratory costs for ion uptake are more than 2-fold lower in the fast-growing grass species than in their slow-growing counterparts. Therefore an overestimation or underestimation of the growth costs does not affect the conclusion of this study.

Table 2.  . Change in specific respiratory costs for ion uptake [cu, mol O2 (mol NO3)–1] after a change in specific respiratory costs for growth (cg, mmol O2 g–1 DM) in the fast-growing Dactylis glomerata and the slow-growing Festuca ovinaThumbnail image of

What happens to the difference in specific respiratory costs for ion uptake between fast- and slow-growing grass species when the costs are calculated on an ATP basis? For D. glomerata we used the minimum estimate of the contribution of the alternative pathway (3%), while using the maximum estimate of the contribution of the alternative path (37%) for F. ovina. Furthermore, for both species a 9% higher rate of root respiration was assumed to account for inhibition by the build-up of CO2 in the cuvettes. Specific ATP requirements for growth are 28·4 and 27·2 mmol ATP g–1 DM for D. glomerata and F. ovina, respectively (H. Poorter and M. Bergkotte, Utrecht University, the Netherlands, unpublished results). Expressed on an O2 basis the specific respiratory costs for ion uptake in the fast-growing grass species were one-third of those in its slow-growing counterpart (Table 1). Nitrate is the main anion taken up by the present grass species; phosphate and sulphate uptake accounts for less than 10% of the total anion uptake. Specific costs for nitrate and phosphate uptake cannot be less than 2 mol of ATP per mole of nitrate or phosphate, since this is the ‘biochemical’ cost for nitrate and phosphate transport based on a proton/nitrate and proton/phosphate stoichiometry of 2 and a stoichiometry of 1 for the plasma-membrane ATPase. For sulphate the ‘biochemical’ cost is 3 mol of ATP per mole of sulphate (Clarkson 1998). Expressing costs on an ATP basis resulted in 1·5-fold lower costs in the fast-growing species than in the slow-growing species. However, the fast-growing D. glomerata still clearly exhibited lower specific respiratory costs for ion uptake than the slow-growing F. ovina[3·9 ± 0·12 and 5·8 ± 0·06 mol ATP (mol NO3)–1, respectively]. Further experiments have shown that a lower nitrate efflux per net nitrate taken up in fast-growing grass species than in slow-growing ones accounts for the lower specific respiratory costs for ion uptake, possibly in combination with lower activity of the non-phosphorylating alternative respiratory path (I. Scheurwater, D.T. Clarkson & H. Lambers, unpublished results; Lambers et al. 1998).

Is the difference in specific respiratory costs for nitrate uptake between fast- and slow-growing grass species similar to that in costs for ion uptake, expressed per mole of nitrate? As long as the relative contribution of the uptake of nitrate to the total amount of main anions taken up (N, P and S) is the same for fast- and slow-growing species, this is indeed the case. The relative contribution (mol percentage) of N was not determined in F. ovina, but was 96% in D. glomerata plants, assuming the sulphur content to be 50% of that of phosphorus (cf. Marschner 1988). If the specific costs for uptake of N, P and S are the same, the specific respiratory costs for nitrate uptake can be calculated by multiplying the costs for ion uptake, expressed per mole of nitrate (Table 1), by 0·96. If the specific respiratory costs for S and P uptake are 3-fold higher than those for N, the specific respiratory costs for ion uptake expressed per net mole of nitrate taken up must be multiplied by 0·89 to arrive at the specific costs for nitrate. We have no evidence that fast- and slow-growing grass species grown with full access to nitrate differ much in mol percentage N. A. Van der Werf (AB-DLO, Wageningen, the Netherlands, personal communication) found a mol percentage N of 92 and 88 for D. glomerata and F. ovina, respectively, for plants grown in pots. This small difference would only lead to 2·8-fold instead of 3-fold lower specific respiratory costs for nitrate uptake in the fast-growing D. glomerata compared with the slow-growing F. ovina.

Concluding remarks

Fast-growing grass species exhibit lower rates of root respiration than those expected from their high rates of growth and nitrate uptake. Inhibition of root respiration by the build-up of CO2 in closed cuvettes, diurnal variation in root respiration rates or an increasing RQ with increasing RGR cannot account for the relatively small difference in root respiration between fast- and slow-growing grass species. Furthermore, it appears that differences in the alternative pathway activity can at most partly explain why roots of fast-growing grass species respire at a lower rate than expected when root respiration is expressed as O2 consumption. Therefore, the relatively low rates of root respiration in fast-growing grass species is likely to be caused mostly by lower specific respiratory costs for growth, ion uptake and/or maintenance of root biomass. While the costs for growth and maintenance are similar, the respiratory costs associated with ion uptake are clearly lower in fast-growing grasses than in their slow-growing counterparts. Fast-growing grasses also exhibit lower specific respiratory costs for nitrate uptake than slow-growing grasses. Further experiments have shown that the lower specific costs associated with nitrate uptake in fast-growing grass species are due to a smaller efflux of nitrogen per net mole of nitrogen taken up in these species compared with slow-growing grasses.

Acknowledgements

The authors thank Marc Bergkotte, Yvonne Van Berkel and Franka Den Ouden for technical assistance. Owen Atkin and Adrie Van der Werf provided valuable comments on previous versions of this manuscript. Seeds of the Poa species and of Festuca ovina were kindly provided by Owen Atkin, Australian National University, Canberra, Australia and Hendrik Poorter, Utrecht University, the Netherlands, respectively. This study was financially supported by the Life Sciences Foundation (SLW), which is subsidized by the Netherlands Organization for Scientific Research (NWO).

Ancillary