Journal of Experimental Botany, Vol. 53, No. 375, pp. 1771-1779,
August 1, 2002
© 2002 Oxford University Press
Estimation of whole-plant transpiration of bananas using sap flow measurements
Received 5 September 2001; Accepted 15 April 2002
1 CSIRO Plant Industry, Darwin Laboratory, PMB44, Winnellie, NT 0822, Australia
2 Faculty of SITE, Northern Territory University, Darwin, NT 0909, Australia
3 To whom correspondence should be addressed. Fax: +61 8 89470052. E-mail: ping.lu{at}csiro.au
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResults and discussionConclusionsReferences |
|---|
Banana, one of the largest rhizomatous herbs in the world, is the fourth most important global food crop. It has a high water requirement, but the whole-plant water use in the field has not been determined satisfactorily. In this study, whole-plant water use in potted and field-grown banana plants (Musa ‘Cavendish’ cv. Williams) was successfully determined using a xylem sap flow method. This was achieved using Granier sensor probes implanted into the central cylinder of the banana corm. The whole-plant water use in field-grown bananas was 9–10 l plant–1 d–1. The values of daily total sap flow in potted plants correlated closely with gravimetric measurements (r2=0.92) and with changes in soil water status (r2=0.77). In well-watered, mature, field-grown plants, hourly sap flow also closely correlated with changes in solar radiation, vapour pressure deficit and evapotranspiration. The study indicates that sap flow measurement is a sensitive and accurate method for determining whole-plant water use in bananas under potted as well as field conditions.
Key words: Key words: Banana, corm, Musa, sap flow, water deficit, whole-plant water use.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionConclusionsReferences |
|---|
Bananas (family Musaceae), one of the largest herbs in the world, is the fourth most important global food crop (INIBAP, 2001). It is generally accepted that bananas need large amounts of water for high production; the recommended irrigation rates in the tropics are 1.0–1.4 of the Class A pan evaporation (Stover and Simmonds, 1987; Diczbalis and Toohill, 1993). However, it is difficult to quantify water use in these plants and the basis for these high values is not well understood. Banana plants range in height from 0.8 m to more than 15 m (Turner, 1994). Each contains a flattened, modified stem called the corm with the root system, a false stem called the pseudostem consisting of concentric layers of leaf sheaths and a crown of large leaves (Fig. 1). The huge tree-like size and non-woody pseudostem make banana one of the most difficult plants for studies on whole-plant water use. Methods used so far in field-grown bananas, such as gravimetric and soil water balance methods, suffer from uncertainties in quantifying the drainage and poor time resolution (Turner, 1987; Robinson and Alberts, 1989). Moreover, current physiological understanding of the banana plant is mainly based on studies at the leaf level which may not be representative of the whole plant because of the variability of leaf scale measurements, in both time and space, that is exacerbated by the plant’s extremely large leaf size. A reliable method for measuring whole-plant transpiration in the field would not only be of great benefit to the irrigation management of this crop, but would also improve understanding of the whole-plant physiology of bananas.
|
Sap flow methods provide direct and continuous measurement of whole-plant water use with high time resolution (Smith and Allen, 1996; Wullschleger et al., 2000). They have distinct advantages, especially in large plants such as bananas, over other methods such as leaf gas exchange, plant chamber, soil water balance, and micrometeorological techniques. However, the banana pseudostem is basically unsuitable for sap flow measurement because it expands rapidly each day and the overlapping, close-fitting leaf sheaths contain a series of longitudinal canals of large lacunae filled with air (Skutch, 1927; Aubert, 1973). In a previous study, sap flow measurements on the banana pseudostem were disrupted because the Granier sap flow probes (Granier, 1987) were displaced by the rapid upward growth of the immature sheaths in the pseudostem (P Lu, unpublished results). Another researcher fitted a circular stem heat balance gauge to a banana leaf petiole but it grossly underestimated water use of the whole leaf (D Thomas, personal communication).
By contrast to the pseudostem, the corm is a compact, ‘wood-like’ organ divided into a massive central cylinder, an outer cortex and an apical ‘cambium’ region (Skutch, 1932). According to Skutch (1932), the central cylinder is the major water transport system in the corm. It contains a ‘bewildering array’ of large primary and secondary vascular bundles. Only the upper region of the leaf trace bundles follows a straight course for any considerable distance, and short lengths of the other bundles are encountered running in every direction, apparently at random.
In this study, the central cylinder of the corm was used to determine plant water use in bananas with the Granier probes. The study included preliminary investigations to establish the major water transport system in the corm of the banana cultivar under study and then to identify areas suitable for the installation of the sap flow probes in the corm. Overall, the study had three main objectives: (1) to quantify the spatial variation in water flow within the corm; (2) to evaluate the accuracy and sensitivity of Granier method when compared with gravimetric measurement of water use; and (3) to examine the responses of sap flow measurements to changes in micrometeorological parameters and soil water status.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionConclusionsReferences |
|---|
Banana plants
Young tissue-cultured banana (Musa ‘Cavendish’ cv. Williams) plants were transplanted into 50 l plastic containers containing a commercial potting mixture and allowed to grow outdoors for 3 months in full sun at the CSIRO Plant Industry Darwin Laboratory, Darwin (12.4° S; 130° E), Australia, during the dry season (May to September) in 2000. The weather conditions throughout the season were sunny and constant during daytime. The plants were watered daily for a total of 1.5 h by drip irrigation (12 l h–1). Fertilizer (20 g; NPK: 18:2:10) was applied to each plant monthly.
Homogeneous plants were selected for the experiments: 2.5 m high with 11 leaves (total leaf area was about 3.8 m2), stem diameter of 10 cm at 30 cm above the corm, corm diameter of about 12.3 cm and ‘sapwood’ area of about 50 cm2. The ‘sapwood’ area was the cross-sectional area of the central cylinder at the position of the heated sap flow sensor probes. During the experiment, the plants were left out in full sun on a wooden bench and irrigated daily (unless stated otherwise) until water drained freely from the base of the pot.
Selected banana plants growing in an orchard at the same site were used for the field trial. During the experiment, these plants were irrigated by sprinklers to field capacity every two days.
Determination of pathway of water transport in banana corm using a staining method
The leaves of two selected banana plants were initially sprayed with water to minimize water stress and xylem embolism before the plants were removed from the container early in the morning. All the roots were severed about 5 cm from the corm. The attached soil was carefully washed away with a jet of water and the soil-free bulb was immediately immersed upright into 10 l of 1% eosin Y (Sigma–Aldrich, USA) solution. The immersed roots were trimmed with a sharp knife and the plant was left in this solution in full sun for about 3 h until the leaves became visibly stained. Then the corm was cut into longitudinal and cross sections and photographed to reveal the water transport pathway within the organ and to identify region(s) with least structural variability suitable for the installation of sap flow sensor probes. The results identified the central cylinder region just below the cambium as most suitable for the installation of sensor probes.
Relationship between corm size and cortex thickness
To install the sensor probes correctly in situ into the corm requires knowledge of the relationship between the thickness of the cortex (or central cylinder) and the mean diameter of the corm under study. This relationship was determined empirically using 18 potted and field-grown banana plants of different sizes. The corms of these selected plants were sectioned and the diameter (cm) of the corm (Dc) and central cylinder (Dcc) determined. The relationship obtained was: Dcc=0.5973 Dc (r2=0.91, n=18). Thus, the diameter of the corm of the plant under study was initially determined and the sensor probes subsequently installed into the inner central cylinder and outer cortex using the above equation.
The veracity of the equation was evaluated by comparing sap flow in eight plants based on direct measurement of the cross-sectional area of the central cylinder with values determined indirectly using cross-sectional area of the central cylinder derived from the above equation. The indirect method was found to underestimate the values determined by the direct method by <3%.
Granier sap flow system
The Granier system for measuring sap flow uses two sensor probes (Granier, 1987). Each probe consists of a heating element spirally coiled around a steel needle containing a T-type thermocouple (copper-constantan). The constantan ends of the two thermocouples are connected together to measure the temperature difference between these two probes. The heating coil of each probe is then glued to the inside of a copper tube (2 cm long and 2 mm diameter) for homogeneous heat dissipation (Granier, 1987; Lu, 1997). In woody species the two probes are inserted into the tissue under study at 10–15 cm apart. The downstream probe is heated at constant power (using a constant current of 130 mA) while the upstream reference probe is left unheated to measure the ambient temperature of the tissue. This separation eliminates the effect of heat carried by sap flow from the heated probe to the reference probe.
The temperature difference between the two probes is influenced by the sap flux density in the vicinity of the heated probe. Sap flux density (Fd, kg cm–2 h–1) is estimated using an empirical relationship (Granier, 1985, 1987):
Fd=0.04284x[(
Tmax–T)/T]1.231
Tmax and T are temperature differences between the heated and reference probes, at zero flux and at time t, respectively. Total sap flow F (kg h–1) was estimated as the product of sap flux density (Fd) and total sapwood area (cross-sectional area) of the tissue examined.
Installation of the Granier sap flow probes in the corm
The heated probe was inserted into the upper region of the central cylinder beneath the cambium region while the unheated reference probe was installed in the outer cortex, about 1–2 cm vertically below the heated probes. According to Skutch (1932) the cortical region 1–2 cm vertically below a certain point in the central cylinder (i.e. where the heated probe is located) is certainly not downstream to that point and, thus, free from interference from the heat dissipated by the heated probe. Unless stated otherwise, the heated probe was installed at a depth of 1–3 cm within the inner central cylinder and the unheated reference probe at a depth of 0–2 cm within the outer cortex.
The probes were installed late at night after the banana plants were watered and their leaves sprayed with water just before dark to ensure that water potentials were as high as possible. The surface of the corm and the probes were initially disinfected by 70% alcohol. Two holes (2.0 mm diameter) were punched into the corm using a long, sharp-edged copper tube and a probe was inserted into each hole. As observed by Davis (1961), sap exuded immediately from the holes as a result of the high root pressure. Once the probes were in place, the holes were sealed with Vaseline and an adhesive pad (Bostik Blu-Tack®) to stop leaks and, most importantly, prevent air from entering the vascular system when negative pressure develops during the daytime (Davis, 1961). To minimize the effect of temperature fluctuations on sap flow measurement and water loss from the soil/potting mixture surface, the ensemble of sap flow sensors, the entire pseudostem and the container were covered and sealed with two layers each of bubble-wraps and aluminium foil. Furthermore, the containers were placed on a wooden bench. Sap flow signals were measured at 10 s intervals, and 30 min means were calculated and stored using a Campbell 21X data logger (Campbell Scientific Inc., Logan, USA).
Spatial variation in sap flux density in the corm
The effect of spatial variation in sap flux density within the corm was also evaluated using three potted (A1–A3) and two field plants (F1, F2). In this experiment, three heated probes were installed in three different areas: 0–2 cm in the outer cortex, and 0–2 cm and 2–4 cm in the central cylinder. All three unheated reference probes were installed in the cortex. To avoid heat interference between the two heated probes in the central cylinder, the probes were heated on alternate days.
Validation of sap flow measurement by gravimetric method
To validate the sap flow measurements determined in this study, three methods were employed to generate a range of daily sap flow and water loss data: (1) induced water stress by a series of wetting–drying cycles; (2) differential shading of canopy; and (3) the total leaf area reduced by halving or removing leaves.
In method 1, transpirational water loss in three potted plants (B1–B3) was determined simultaneously by sap flow and gravimetric methods. Each experimental plant/container was irrigated at the start of the experiment and then subjected to a series of wetting–drying cycles. After 2–3 d, watering was stopped in order to initiate a drying cycle. When sap flow declined to about zero, the plants were rewatered to initiate another wetting cycle. This treatment was maintained for up to 3 weeks or until the end of the experiment. Gravimetric water loss was measured either during the daytime when the plants were irrigated daily or over a 24 h period during wetting–drying cycles. The measurements were determined by weight loss using an electronic balance (maximum capacity 200±0.02 kg).
In method 2, shade levels of 0%, 40% and 70% were applied to six plants over a period of 4 d. In method 3, the total leaf area of bananas, used in the method 2, was halved by removing the top-half of the leaf at the end of the experiment.
Stomatal conductance and leaf folding
Stomatal conductance (gs) on the lower leaf surface (banana is hypostomatous) was determined by an AP4 cycling porometer (Delta-T Devices, Cambridge, UK). For each banana plant, five leaves (from the second to sixth youngest leaves) were used. Five measurements were made on each leaf at a position 2–5 cm from the leaf margin in the mid section of the leaf, every 2 h between 09.00–17.00 h.
Leaf folding may indicate plant water status of bananas (Thomas and Turner, 1998). Relative leaf folding, F, was measured near the centre of the leaf on the second to the fifth youngest leaves in full sun every 2 h according to Turner and Thomas (1998). Relative leaf folding, RLF=1–da/dx, where dx, is the maximum distance between lamina margins and da, the actual distance between lamina margins. This gives a value of 0.0 for a flat leaf and a value of 1.0 for a fully folded leaf when the lamina halves are vertical to the ground.
Micrometeorological factors and soil water conditions
An automatic weather station was mounted on a 2 m high stand at a free space approximately 2 m from the potted bananas and about 80 m from the field-grown plants. Solar radiation (Li-Cor L1200X silicon pyranometer, Li-Cor, USA), air temperature/relative humidity (CS500 Vaisala Humitter, Vaisala, Finland), wind speed (RM Young anemometer, USA), and rainfall (Hydrological Services tipping bucket rain gauge) were measured at 10 s intervals, 30 min averages were calculated and stored using a Campbell CR10X logger (Campbell Scientific Inc., Logan, USA). Reference evapotranspiration (RefET) was computed according to Smith et al. (1992).
Relative available water content (RAWC) was measured by weighing the plant-container regularly. RAWC= (Wa–Wmin)/(Wmax–Wmin), where Wa is the actual weight of the plant-container, Wmin is the weight when sap flow is almost zero, and Wmax is the weight at the field capacity.
|
Results and discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionConclusionsReferences |
|---|
The water pathway in banana corm
The central cylinder of the corm was heavily stained by eosin Y indicating active water transport in this region (Fig. 2). By contrast, the cortex was poorly stained indicating that it is a region with relatively low water transport activity. But the boundary between the cortex and central cylinder in the lower half of the corm was heavily stained, presumably, because this region contained extensive longitudinal secondary vascular bundles (Skutch, 1932). The upper half of the central cylinder beneath the cambium region was relatively homogeneously stained. In this region, the stained vascular bundles were mostly principal leaf-trace bundles. By contrast, the apical cambium region, which did not contain any mature vascular bundles (Skutch, 1932), was barely stained. These results are consistent with the observation by Skutch (1932) on other banana cultivars.
|
Clearly, the cortex and central cylinder of the banana corm differ structurally and physiologically with respect to water transport activity. There are three important conclusions from these observations: (1) the corm has a distinctive sapwood-like structure that may be suitable for sap flow measurement; (2) the central cylinder is the primary water transport region in the banana corm; and (3) the upper region of the central cylinder beneath the cambium region is the most suitable region for installing sap flow sensor probes because it has less structural variability.
Spatial variation in sap flux density in the corm
Sap flux density measured in two regions (0–2 cm and 2–4 cm) within the upper region of the central cylinder in potted plants showed a typical bell-shaped diurnal pattern (Fig. 3A). The two diurnal variations obtained were essentially indistinguishable from each other indicating that sap flow activity within the central cylinder was intense but relatively uniform. On the other hand, sap flow activity in the cortex was very low and varied substantially from day to day. Essentially similar results were obtained in field-grown plants although the activity determined in the central cylinder of these plants was substantially greater than that determined in the potted plants (Fig. 3B). Sap flow activity was also low in the cortex of the field plants.
|
Significant night-time sap flow activity was observed in both the potted and field plants (Fig. 3A, B). The activity was probably associated with a high level of vapour pressure deficit (VPD) prevailing in the night. For the potted plant, the night VPD values at 00.00–04.00 h and 20.00–24.00 h on the day 1 (probe at 0–2 cm) were 0.1 and 0.3 kPa, respectively. On day 2 (probe at 2–4 cm) the respective values were 0.25 and 0.1 kPa. For the field plant, the respective VPD values on day 1 were 0.2–1.2 and 0.5 kPa, and on day 2, 0.8 and 0.3 kPa.
The results confirmed that the central cylinder is the primary water transport region in the banana corm and that sap flow in the upper region of the central cylinder was spatially homogeneous at least when measured by Granier probes which may largely integrate sap flow over the entire surface area of the probes (Lu, 1997; Clearwater et al., 1999). In all subsequent studies, only one single pair of sap flow sensors was installed at 1–3 cm beneath the cortex zone in the upper region of the central cylinder.
Validation of sap flow in banana corm using a gravimetric method
Wetting–drying cycles: Figure 4 shows that excellent agreement (r2=0.92–0.93) was obtained from three individual banana plants and the pooled data between daily sap flow and gravimetric water loss in a series of wetting–drying cycles. The maximum sap flow values obtained were very similar to the whole plant water loss of about 7 kg plant–1 d–1 reported for irrigated plants determined by the gravimetric method and by lysimeters (Turner, 1987). Evidently, sap flow in the central cylinder is a good estimate of whole-plant water use in container-grown banana plants.
|
Sap flow measurements are sensitive to changes in plant internal water storage. In plants such as bananas with soft, highly hydrated tissue and a pseudostem made up of about 90% water, these changes are likely to be large, and expected to have a substantial impact on sap flow determination. In another wetting–drying cycle experiment involving four potted plants, the weight of each plant was determined at field capacities and zero sap flow stages using a destructive method. The average plant weights at the well-watered stage and maximum stressed conditions were about 11 and 6.4 kg, respectively. Hence the total plant water content decreased about 42% during a drying cycle. This value is 5-fold greater than the 7% reduction reported for relative water content in banana leaves (Turner and Thomas, 1998). Despite the massive decrease in the plant water reserve, uncertainties remained with respect to: (a) when and at what stress level this stored water began to recess to the transpirational stream, thus affecting the comparison of the sap flow/gravimetric methods; and (b) the rate of recharging this reservoir after rewatering.
In this study the heated probes were installed in the corm. Hence only changes in the downstream water reservoir (i.e. in the leaves and pseudostem) would affect the relationship between sap flow and gravimetric water loss. Furthermore, the uncertainties in water reserve changes might not have affected accumulated sap flow and gravimetric water loss over a complete wet–dry–wet cycle. The relationship could also be affected by the amount of water incorporated into the tissues for new growth. Over a 24 h period, the new leaf may grow by 10–30 cm depending on plant size (Turner, 1994). However, as most new growth occurs at night, this factor may not be significant in this study because the comparison between gravimetric and sap flow measurements was made during daytime.
Differential shading of canopy: Varying the shade levels had little effect on transpiration (data not shown) presumably because, in bananas, transpiration was governed by the prevailing high VPD (2.5–5.0 kPa) of the environment (P Lu, unpublished results).
Leaf removal: Leaf removal also had little effect on plant water loss (data not shown) because of the high exudation rate produced at the cut leaf edges due to damaged laticifers (Milburn et al., 1990) and/or high root pressure (Davis, 1961). In this study, the weight (1.40 kg) of exudate collected over a 24 h period from a leafless stump of about 1.2 m high was found to be comparable to the estimate (1.25 kg) determined from sap flow measurements.
Response of sap flow to plant/soil water status and micrometeorological factors
This experiment was conducted on three potted plants (B1–B3 which were also used in the validation of the sap flow measurements) and two field-grown plants (F3 and F4). Diurnal variations of sap flow and gravimetric water loss in these plants were determined with respect to diurnal changes in solar radiation (Rg), RefET, RLF, RAWC, and gs. Results from all five plants were similar, so only those from the potted B2 and field-grown F3 plants are discussed.
Figure 5 shows the diurnal and/or daily patterns of Rg, RefET, sap flow, RLF, RAWC, and gravimetric water loss over two wetting–drying cycles between 29 August and 11 September (DOY 242-255) in the potted B2 plant. Diurnal Rg and RefET were fairly constant during the study period (Fig. 5A). RAWC decreased during the drying cycle (DOY241-245 and DOY249-252) and increased again when the plant was rewatered (Fig. 5B).
|
Maximum gs followed the same trend as daily RAWC (Fig. 5B). Daily maximum gs values declined sharply in the drying cycle and regained their high values in the wetting cycle. The average maximum gs values obtained at high RAWC were about 300 mmol m–2 s–1 which were similar to those reported for potted Cavendish cv. Grand Nain bananas in a heated glasshouse (Turner and Thomas, 1998). However, individual gs values of about 800– 1000 mmol m–2 s–1 that were similar to values reported by Eckstein and Robinson (1995) for the same cultivar in the subtropics were frequently observed.
Diurnal changes in RLF (Fig. 5C) followed the same trend as diurnal sap flow (Fig. 5D). The leaves folded progressively in the morning and then unfolded gradually after midday. However, the diurnal changes in RLF did not appear to be influenced by changes in soil water availability. The RLF pattern obtained at the start of the drying cycle was similar in nature and magnitude to that obtained at the end of the drying cycle when the RAWC value had reached a minimum. Evidently, banana leaves can regain full turgor over the night even under very dry soil conditions. These results are consistent with those of Turner and Thomas (1998) and indicate that RLF is not a good indicator of plant water stress in bananas.
Diurnal and daily sap flow declined during the drying period and increased during the wetting period indicating that both were greatly affected by soil water availability (Fig. 5D, E). Their values were similar to the respective values determined for diurnal and daily gravimetric water loss. However, the correlation between hourly sap flow and gravimetric measurements was found to vary between days; good agreement was observed on DOY 248 but not DOY 249 (Fig. 5D). Such variable results may be attributed to changes in internal water storage and water use by new growth.
Both daily sap flow and daily water loss were closely correlated with RAWC (Fig. 6) indicating that sap flow responded sensitively to changes in soil/plant water deficit during the wetting–drying cycles in potted banana plants.
|
For the field-grown bananas, daily micrometeorological conditions and soil water status remained relatively constant throughout the study period (data not shown). Daily whole-plant water use (sap flow) of field-grown bananas was estimated to be about 9–10 l in this study. This value is much less than the current recommended irrigation rate of 42 l d–1 plant–1 in the Darwin region, Northern Territory, Australia (Diczbalis and Toohill, 1993). The latter would include plant water use as well as water losses from deep drainage, runoff and evaporation of water from the soil and between the water delivering points (e.g. sprinklers) and the soil surface. Nonetheless, the great disparity between whole-plant water use and current industry practice indicates a potential for improving the irrigation efficiency of banana plantation in the Darwin region. Furthermore, the hourly sap flow measured was highly correlated with Rg, VPD and RefET (Fig. 7) suggesting that the relationship established between sap flow and micrometeorological factors might be used to estimate water use of banana plantation directly from micrometeorological data.
|
Banana has a high leaf gas exchange rate (Turner, 1994; Eckstein and Robinson, 1995), but the maximum sap flux density determined in this study was only 0.015 kg cm–2 h–1. By contrast, the maximum sap flux density measured with the same method in spruce (Lu et al., 1995), mango (Lu et al., 2000) and tropical rain forest species (Granier et al., 1996) were 0.019, 0.035 and 0.040 kg cm–2 h–1, respectively. Banana also has a very low leaf area to ‘sapwood’ area ratio (<0.08 m2 cm–2) that is 10–150 times less than that determined in the above woody species. Thus the low sap flux density value in banana is probably attributed to its low leaf area/sapwood ratio.
A recurring problem with the sap flow measurement in the corm
In the above experiment about half of the probes produced, immediately after the installation, the normal patterns of sap flux density as those shown in Fig. 3. The rest produced an abnormal but consistent pattern of diurnal sap flux density. Figure 8A and B show, respectively, the normal and abnormal diurnal patterns of the temperature difference between the heated and reference probes (T) obtained in banana plants. In a normal diurnal cycle (Fig. 8A), T increased immediately after installation and heating (a, b). It reached a maximum before declining slowly during the night (b, c). This nocturnal decline was probably due to tissue rehydration, which enhanced conductive heat loss from the heated probe. T decreased rapidly after sunrise, reaching a minimum at midday (c, d). This substantial decline was caused by convective heat loss due to sap flow. A decrease in sap flow in the afternoon caused a corresponding rise in T (d, e). This cycle was repeated in the following 24 h period. This is the pattern normally observed on actively transpiring woody plants. T values from these sensors were utilized for calculation of sap flux density.
|
Figure 8B shows an abnormal
T diurnal pattern. The initial pattern obtained during the night (a', c') was similar to the normal pattern. But at sunrise, T decreased initially (c', d') and then increased rapidly throughout the day (d', e'). After sunset, T decreased rapidly reaching a steady state within 1–2 h (e', f’). This stable T value was higher than that obtained the previous night. This cycle was repeated the following day.
The d', e' phase was abnormal and had no physiological basis. It was not consistent with gravimetric measurements obtained during this phase, which clearly showed that the plant was actively transpiring. Such active sap flow would be expected to substantially decrease T. These abnormal patterns were observed with both potted and field-grown plants.
Corms sectioned at the end of the experiment showed the presence of discoloured and dry tissues and gaps between the soft tissues and the probes that produced the abnormal patterns. By contrast, gaps were not observed with probes that produced the normal patterns. Evidently, the increase in T might be caused by the disruption in heat conductivity related to gap formation. This problem could also be caused by the toxic exudates from damaged laticifers that affect the vascular tissues directly. The size of the gap formed may be related to the severity of damage caused during hole puncturing and also to the number of laticifers damaged. The results in Fig. 8B suggest that contact between probe and tissue was apparently disrupted soon after sunrise (at d') when tension increased in the vascular system because of falling water potential. Consequently, T increased rapidly before a new but higher steady-state (f', g') was established when the corm tissue was partially rehydrated and the intervening gaps partially refilled.
A potted plant with abnormal results was monitored continuously for 3 weeks. Surprisingly, the abnormal pattern reverted to normal towards the end of this period (data not shown). The gaps responsible for the abnormal results had apparently resealed suggesting that it is worthwhile to persist with sensor probes that generate abnormal measurements initially. Finally, this problem could be alleviated by ensuring a tight-fit between the hole and sensor probe and by minimizing tissue damage when the hole was created.
Sap flow measurements may also be perturbed by high ambient thermal gradient between the heated and reference probes (Goulden and Field, 1994). In this study, excellent insulation and isolation of the entire experimental system of potted plants from the ground had kept ambient gradients to less than 0.05 °C between the heated and reference probes. These low gradients were insignificant to perturb sap flow determinations in the experiments reported in this study.
|
Conclusions |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionConclusionsReferences |
|---|
Water transport occurred mainly in the central cylinder of banana corm (Fig. 1). The upper region of the central cylinder, which was located above-ground, had less structural complexity and was found to be suitable for sap flow measurement using the Granier sensor probes. However, in some cases, it was necessary to wait 2–3 weeks after probe installation before the normal pattern of sap flow signals (
T) could be obtained. Sap flow values obtained in the central cylinder were highly consistent with those of plant water loss determined by the gravimetric method. Furthermore, sap flow in the central cylinder varied diurnally in response to diurnal variations in micrometeorological factors and to soil water availability during wetting–drying cycles. The results show that the Granier sap flow system is an accurate and sensitive method for determining whole-plant water use in both potted and field-grown banana plants. In the field, sap flow measurements were highly correlated with micrometeorological parameters indicating that water use in banana plantations could possibly be estimated directly from micrometeorological data. In situ measurements of whole-plant water use in banana by sap flow methods are useful for understanding whole plant physiology and potentially invaluable to irrigation management in banana plantations. Finally, under dry tropical conditions, leaf gas exchange (Turner and Thomas, 1998) and canopy transpiration (sap flow) are superior to the traditional measures of leaf water status, such as relative leaf folding, for determining the response of banana plants to water deficit. |
Acknowledgements |
|---|
The authors thank L Zhong for construction of 10 cm long sap flow sensor probes used in the study, D Turner, D Thomas and R Foo for their useful comments on the original experimental plan and members of the CSIRO Plant Industry Editorial Panel (B Loveys, T Condon and J Passioura).
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionConclusionsReferences |
|---|
Aubert B. 1973. Particularités anatomiques liées au comportement hydrique des bananiers. Fruits 28, 589–604.
Clearwater MJ, Meinzer FC, Andrade JL, Goldstein G, Holbrook NM. 1999. Potential errors in measurement of non-uniform sap flow using heat dissipation probes. Tree Physiology 19, 681–687.[Abstract]
Davis TA. 1961. High root-pressure in palms. Nature 192, 277–278.
Diczbalis Y, Toohill B. 1993. Irrigation management of bananas in the top end. Agnote No. 567. Northern Territory Department of Primary Industry & Fisheries (Darwin, Australia)
Eckstein K, Robinson JC. 1995. Physiological responses of banana (Musa AAA; Cavendish subgroup) in the subtropics. II. Influence of climatic conditions on seasonal and diurnal variations in gas exchange of banana leaves. Journal of Horticultural Science 70, 157–167.[Web of Science]
Goulden ML, Field CB. 1994. Three methods for monitoring the gas exchange of individual tree canopies: ventilated-chamber, sap-flow and Penman–Monteith measurements on evergreen oaks. Functional Ecology 8, 125–135.
Granier A. 1985. Une nouvelle méthode pour la mesure du flux de sève brute dans le tronc des arbres. Annales des Sciences Forestières 42, 193–200.[Web of Science]
Granier A. 1987. Evaluation of transpiration in a Douglas fir stand by means of sap flow measurements. Tree Physiology 3, 309–320.[Abstract]
Granier A, Huc R, Barigah ST. 1996. Transpiration of natural rain forest and its dependence on climatic factors. Agricultural and Forest Meteorology 78, 19–29.[Web of Science]
INIBAP(International Network for the Improvement of Banana and Plantain). 2001. http://www.inibap.org/promusa/bitw.htlm, Montpellier, France
Lu P. 1997. A direct method for estimating the average sap flux density using a modified Granier measuring system. Australian Journal of Plant Physiology 24, 701–705.[Web of Science]
Lu P, Biron P, Breda N, Granier A. 1995. Water relations of Norway spruce (Picea abies (L) Karst) under soil drought in the Vosges mountains: water potential, stomatal conductance and transpiration. Annals of Forest Science 52, 117–129.
Lu P, Muller W, Chacko EK. 2000. Spatial variations in xylem sap flux density in the trunk of orchard-grown, mature mango trees under changing soil water conditions. Tree Physiology 20, 683–692.[Abstract]
Milburn JA, Kallarackal J, Baker DA. 1990. Water relations of banana. I. Predicting the water relations of the field-grown banana using the exuding latex. Australian Journal of Plant Physiology 17, 57–68.
Robinson JC, Alberts AJ. 1989. Seasonal variations in the crop water use coefficient of banana (cultivar ‘Williams’) in the subtropics. Scientia Horticulturae 40, 215–225.
Skutch AF. 1927. Anatomy of leaf of banana, Musa sapientum L. var. hort. Gros Michel. Botanical Gazette 84, 337–391.
Skutch AF. 1932. Anatomy of the axis of the banana. Botanical Gazette 93, 233–258.
Smith DM, Allen SJ. 1996. Measurement of sap flow in plant stems. Journal of Experimental Botany 47, 1833–1844.
Smith M, Allen RG, Monteith JL, Perrier A, Pereira L, Segeren A. 1992. Report of the expert consultation on procedures for revision of FAO guidelines for prediction of crop water requirements. Rome, Italy: UN-FAO.
Stover RH, Simmonds NW. 1987. Bananas. London: Longman Group.
Thomas DS, Turner DW. 1998. Leaf gas exchange of droughted and irrigated banana cv. Williams (Musa spp.) growing in hot, arid conditions. Journal of Horticultural Science and Biotechnology 73, 419–429.
Turner DW. 1987. Nutrient supply and water use of bananas in a subtropical environment. Fruits 42, 89–93.
Turner DW. 1994. Bananas and plantains. In: Schaffer B, Andersen PC, eds. Handbook of environmental physiology of fruit crops, Vol. 2. Subtropical and tropical crops. Boca Raton: CRC Press, 37–64.
Turner DW, Thomas DS. 1998. Measurement of plant and soil water status and their association with leaf gas exchange in banana (Musa spp.): a laticiferous plant. Scientia Horticulturae 77, 177–193.
Wullschleger SD, Wilson KB, Hanson PJ. 2000. Environmental control of whole-plant transpiration, canopy conductance and estimates of the decoupling coefficient for large red maple trees. Agricultural and Forest Meteorology 104, 157–168.[Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  S. I. N. Ayutthaya, F.C. Do, K. Pannengpetch, J. Junjittakarn, J.-L. Maeght, A. Rocheteau, and H. Cochard Transient thermal dissipation method of xylem sap flow measurement: multi-species calibration and field evaluation Tree Physiol, October 27, 2009; (2009) tpp092v1. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||