29日施肥灌溉条件下芦苇生长动态和氮动态

ELSEVIER
Field Crops Research 55 (1998) 153-164
Field Crops Research
Growth and nitrogen dynamics of reed canarygrass ( Phalaris arundinacea L.) subjected to daily fertilization and irrigation in the fieldThomas K~itterer a,*, Olof Andr~n a, Roger PetterssonReceived 5 April 1997; revised 18 June 1997; accepted 19 June 1997
b
a Department of Soil Sciences, SLU, Box 7014, Uppsala S-75007, Sweden b Department of Ecology and Environmental Research, SLU, Box 7072, Uppsala S-75007, Sweden
Abstract A field experiment including daily fertilization and daily irrigation was conducted to study the growth and nitrogen dynamics of Phalaris arundinacea subjected to various moisture and fertilizer regimes. There were five treatments: Control 0 (Co), neither fertilized nor irrigated; Control 1 (C1), not irrigated and supplied with 15 g N m -2 yr-1 applied in a single dose in early spring; Irrigated (I1), fertilized as described for C 1, but with daily irrigation; Irrigated/Fertilized 1 (IF1), received the same amount of nitrogen as C 1 and I l, but supplied daily through a drip-tube system according to the predicted N demand of the crop; Irrigated/Fertilized 2 (IF 2) was irrigated as 11 and IF 1, but received higher fertilizer doses, to eliminate any nutrient limitation. The above-ground crop was sampled 26 times between autumn 1992 and spring 1995. The in situ decomposition of crop residues was studied during 1993-1994. Nitrogen concentrations were measured in dry mass produced during the current year and decomposing residues from previous years' production. The loss of macronutrients from the crop between late autumn and the harvest time in early spring was calculated. The main differences in growth dynamics and yield between the treatments were due to irrigation. Growth was similar among the irrigated treatments. Crop biomass and nitrogen dynamics were similar in I 1 and IF 1. Thus, the fertilizer regime did not affect crop growth or N amounts in the crop. Although crop nitrogen concentrations were highest in IF2, this did not result in higher crop production. Differences in growth dynamics between C o and C 1 were due to N fertilization. Differences in dry mass dynamics between 1993 and 1994 were explained by differences in temperature. The relatively cold May in 1994 retarded growth and decreased the crop's efficiency at converting radiation into biomass, even though N concentrations were almost twice as high in spring 1994 as in spring 1993. During the first year of establishment, about 0.4 kg dry mass m -2 was produced. Under favourable conditions (fertilized and no water stress), 1.5 kg dry mass m -2 was produced. However, during one winter, dry mass decreased by about 40% until harvest in spring. The main factor limiting production was suggested to be low straw stiffness, which caused lodging, mainly in the irrigated treatments. Unless plant breeding can produce stiffer straw, the one-cut system will probably not be able to compete with other en

ergy crops under intensive management at lower latitudes. The decomposition rates of crop residues were higher in the irrigated treatments than in C O and C 1. Nitrogen fertilization increased concentrations of N, P, K and S at harvest, and thus reduced the crop's quality as a biofuel. Changes in element
* Corresponding author. Tel.: +46-18-67-24-25; fax: +46-18-67-34-30; e-mail: thomas.katterer@mv.slu.se. 0378-4290/98/$19.00 ? 1998 Elsevier Science B.V. All rights reserved.PHS0378-4290(97)00075-0
154
T. Kiitterer et al. / Field Crops Research 55 (1998) 153-164
concentrations during winter were minor and not consistent. However, concentrations of K were about 2.5 to 5 times lower in early spring than in the preceding autumns. ? 1998 Elsevier Science B.V.Keywords: Biomass production; Daily irrigation and fertilization; Growth dynamics; Light interception; Nitrogen dynamics; Phalaris arundinacea
1. Introduction
Reed canarygrass occurs naturally on moist sites in the temperate zones (SjiSdin, 1990). Although it was cultivated for forage as early as the 17th Century (Alway, 1931), there is no such cultivation in Sweden today. However, interest in cultivating this species as an energy a n d / o r fiber crop has increased recently, especially in northern Europe. With about 4000 ha of canary grass currently cultivated in Scandinavia (Ericson et al., 1995). Compared with other tested grasses, production seems to be smaller during the first year (Andersson, 1977), but increasing during subsequent years (Aase et al., 1977; Ostrem and ~)yen, 1985; Andersson, 1987; Tuvesson, 1989; Lomakka, 1992). From an economical viewpoint, a major advantage is that existing farming machines and implements can be used for harvesting. A one-cut harvesting system, in which the crop is harvested during late winter or early spring, seems to be most profitable (LandstriSm et al., 1996). At that time, the crop water content is usually low, i.e., 10-15% (Lomakka, 1993; Hadders, 1994), and no drying is required before processing to combustibles, e.g., powder, pellets, wafers, cobbs (Ericson et al., 1995). Two areas are of major concern regarding the cultivation of Phalaris arundinaciea: (1) high nutrient concentration in the biomass reduces its quality as a biofuel (Burvall and Hedman, 1994), and (2) little is known regarding nutrient cycling within the cropping system (Landstrrm et al., 1996). The work presented in this paper was performed within the framework of a larger project including investigations of the soil (dynamics of water, heat and mineral nitrogen) and below-ground biomass and nitrogen dynamics (Andrrn et al., 1996). The data presented here will be included in comprehensive agroecosystem budgets. The objective of this study was (1) to study the growth dynamics of reed canarygrass in central Sweden under optimal water and nutrient conditions and under suboptimal conditions, i.e., under water a n d / o r
nitrogen stress, (2) to assess the effects of the fe

rtilizer regime on growth and nitrogen uptake dynamics under non-limiting moisture conditions, and (3) to evaluate the potential for using reed canarygrass managed in a one-cut system as a raw material for fibre and energy production in central Sweden. The following hypotheses were tested: (1) daily N fertilization corresponding to the difference between the assumed crop N uptake and net mineralization will increase crop growth compared with application of the same amount of N applied at the beginning of the growing season; (2) irrigation will enhance crop growth; (3) N fertilization will increase crop nutrient concentrations at harvest; (4) nutrient concentrations, especially those of potassium, will decrease during winter.
2. Material and methods
2.1. Site description and treatments The field experiments were carded out at Ultuna, about 5 km south of Uppsala, in central Sweden, 59°48'N and 17°38'E. The climate is cold temperate, with a mean annual temperature of +5.4°C. The mean annual precipitation is 520 mm, with a maximum in late summer and a minimum in late winter. The soil is a Fluventic Eutrochrept with a clay content of about 53%. Other site properties were described in detail by Andrrn et al. (1990). The experimental layout was a randomized block design with four blocks. The 20 plots (5 treatments X 4 blocks) were separated by 2.5-m-wide untreated areas, and each plot was 8 X 11 m, including a 0.5-m protective zone around each plot, where no samples were taken. After 2 years of unfertilized grass ley the experimental field was ploughed in October 1991, harrowed twice in May and June 1992 and sown then on 20 June 1992 with 2 g m -2 reed canarygrass (P. arundenacea L., cv. Venture). Seeds were placed at about 1 cm depth. The row spacing was 12.5 cm.
T. Kiitterer et aL /Field Crops Research 55 (1998) 153-164
155
On 11 May 1993, five treatments were established: Control 0 (Co), neither fertilized nor irrigated; Control 1 (C1), not irrigated and supplied with 15 g N m -2 yr -1 as solid Ca(NO3) 2 applied in single doses in early spring, on 11 May in 1993 and on 17 May in 1994; Irrigated (II), fertilized as described for C i, but with water supplied daily by irrigation along each row; Irrigated/Fertilized 1 (IF l), received the same amount of nitrogen as C 1 and I l, but supplied as 3 g m - 2 of solid N fertilizer on 11 May 1993 and the remaining amount (12 g in 1993 and 15 g in 1994) supplied in daily doses in the irrigation water according to a logistic function fitted in advance to the predicted nutrient demand by the crop (Ingestad and Lund, 1986); Irrigated/Fertilized 2 (IF 2) was irrigated as described for 11 and I F i , but received higher nutrient doses, i.e., 6 g m -2 as solid N fertilizer and 36 g N m -2 pided into daily nutrient doses during 1993 and a total of 45 g N m -2 as daily nutrient doses during 1994. Irrigation and fertilization were started on 11 May in 1993 and on 17 May in 1994 (Fig. 1). Irrigation/fertilizati

on was conducted until 16 September in 1993 and until 29 July in 1994. Amounts of irrigation water applied in treatments I~, IF l and IF 2 (calculated to keep the soil close to field capacity) were 522 m m during 1993 and 401 m m during 1994. For every 100 g of N applied in IF 1 and IF 2, 14 g of P and 72 g of K were also supplied, together with micronutrients (Ingestad and Lund, 1986).1200 1000
2.2. M e a s u r e m e n t s
Before sowing, K and P concentrations in the topsoil ( 0 - 3 0 cm) were measured on 26 May 1992 after bulking 10 subsamples in each block. Concentrations of P and K in soil were measured in soil extracts using 2 M HC1 and a buffered solution of ammonium lactate, i.e., 0.10 M ammonium lactate and 0.40 M acetic acid, respectively. On the day of sowing, the whole field was fertilized with 6 g N, 15 g P and 47 g K m -2. Total above-ground dry mass was sampled in one randomly chosen plot in each block on 21 October 1992 and 21 April 1993, and in two randomly chosen plots in each block on 5 May 1993, i.e., before the commencement of irrigation and fertilizer treatments. Thereafter, the plots were sampled nine times during 1993, 13 times during 1994 and once in 1995 (5 April, the day of the final harvest). One circular area of crop of 0.25 m 2 was cut at the soil surface. The dry mass produced during 1992 was not removed from the field. The dry mass produced during 1993 (excluding stubble and harvest spillage) was removed on 28 March 1994. Stubble height was highly variable due to lodging. In the field, the samples (circular areas of 0.25 m 2) were pided into two fractions: (1) crop dry mass, i.e., biomass and necromass produced during the current year and (2) necromass, here defined as remaining plant material produced during the previous year (including standing dead mass, i.e., stubble). During 1994, it became~20
Average precip.,/, . . . . . . -:=--~'E
8oo~ 600 .? Irrigation p . . . . ~ . . . . ~ ~ ' ~ / ' 1 '-" "" " P r e c i p i t a t i ° n / ..... " / ~ ...... " # 4 - - ",'-
8o%0. 60 c "-
.,'/. ..........
:'__--7'_ .....
,'
z
I
i
7
"'"
20
Or Jan
,
i Mar
i
i i May
i Jul
i
J i Sep
i Nov
i
i Jan
i
i Mar
J
i i May
i Jul
i
i J Sep
i Nov
t
0
1993
1994
Fig. 1. Precipitation, average precipitation (based on 30 years of measurements), irrigation in reed canarygrass (treatments 11, IF l and IF2)
and fertilizer input? Cumulative values during 1993 and 1994. Treatments: C l, fertilized with 15 g m -2 yr -t solid N - fertilizer each spring; I1, irrigated daily, fertilized as C1; IF1, drip irrigated and fertilized daily (15 g m z yr 1); IF2, drip irrigated as IFl and fertilized daily (42gm 2 in 1993 and 45gin 1994).
156
T. Kiitterer et al. / Field Crops Research 55 (1998) 153-164
increasingly difficult to separate these two fractions from each other, and at the final harvest on 5 April 1995 it was not possible to do so. All samples were dried at 80°C for two days and then weighed. After

drying, all samples were cut into small fragments, and subsamples were ground in a highspeed mill. Total carbon and nitrogen contents were determined for each plot inpidually on 5 April 1994 and for bulked samples from the four blocks on the other sampling dates. Crop dry mass samples taken before and after the winters of 1993-1994 and 1994-1995 were also analyzed for P, K, Ca, Mg and S. Concentrations of these elements were determined on the last sampling dates in 1993 and 1994. The interception of photosynthetically active radiation (PAR) was calculated from measurement of photosynthetic photon flux density (PPFD) above the crop and at the soil surface in treatment IF 2 (LI190SA sensors, LI-COR, Lincoln, NE 68504, USA). The sensors were connected to a Campbell CR 10 logger (Campbell, Logan, UT, USA) and monitored every minute.
and blocks as main plots and sampling dates as split-plots, was used. LSD tests (95% confidence limit) were used for assessing differences between means (SAS Institute, 1982). The graphs were constructed using the SAS procedure GPLOT (SAS Institute, 1985).
3. Results3.1. A b i o t i c c o n d i t i o n s
2.3. S t a t i s t i c a l a n a l y s i s
The SAS procedure GLM was used for analysis of variance. Besides two-way ANOVAs (classes: block, treatment), a split-plot design, with treatments
Winter and spring were warmer in 1992-1993 than in 1993-1994 (Table 1). Mean air temperature in May was almost 4°C higher in 1993 than in 1994 and about 3°C higher than normal (1963-1988). Light conditions in May did not differ between the 2 years (Fig. 2). The mean value for PAR during May, measured above the crop, was about 40 m o l m -2 day-1 in 1993 and 1994, which corresponds to about 8.8 MJ m -2 day -1 (McCartney, 1978). However, the summer of 1993 was colder and wetter than summer 1994 (Fig. 1; Table 1), and incoming PAR was lower during the summer of 1993 than during the summer of 1994 (Fig. 2). Concentrations of P and K in soil extracts on 26 May 1992 were 19 and 41 r a g / 1 0 0 g air-dry soil measured in a m m o n i u m lactate extract and 90 and 703 m g / 1 0 0 g air-dry soil measured in HCL extract, respectively.
Table 1 Air temperature (°C at 1.5 m height, monthly mean), global radiation (MJ m-2 month-1) and precipitation (mm month-1 ) during 1993-1994 and average values based on 25 years of measurements Month January February March April May June July August September October November December Air temperature Average - 4.2 - 4.6 - 1.3 3.8 9.7 14.6 16.0 14.8 10.8 6.1 1.2 - 2.1 1993 - 0.3 - 1.0 1.0 5.1 12.6 11.9 16.0 13.3 7.7 4.8 - 0.2 - 1.0 1994 - 3.0 - 9.0 - 0.3 6.1 8.9 13.2 20.1 15.6 11.0 5.3 1.7 1.1 Global radiation Average 46 102 243 377 553 603 563 433 260 129 50 25 1993 39 99 237 424 640 594 494 407 286 129 29 18 1994 35 112 236 395 648 575 756 462 244 142 51 19 Precipitation Average 35 26 24 28 33 46 8 65 53 48 48 43 1993 21 15 16 25 23 61 49 85 10 50 20 47 1994 35 2 43 30 16 55 13 78 105 32 28 42
T.

Kiitterer et a l . / Field Crops Research 55 (1998) 153-16460 50i "o
157
40
7 E 30
~ 2o10
Apr
Jul
Oct
Jan
Apr
Jul
Oct ) and at the soil surface ( - - ) in
1993
1994
Fig. 2. Photosynthetically active radiation (PAR, weekly means) measured above the crop ( treatment IF 2.
3.2. Crop growth dynamicsDuring the establishment year (1992) about 0.4 kg dry m a s s m -2 was produced. Under favourable conditions (fertilized and no water stress), 1.5 kg dry mass m -2 was produced during 1993 and 1994 (Fig. 3). However, during one winter (October 1993-April 1994), crop dry mass decreased by about 40%. During both years, above-ground crop dry mass increased up to a peak, whereafter it fluctuated (Fig. 3). In the irrigated treatments, peak dry mass values were similar in 1993 and 1994. Using time as a split-plot factor, all treatments, except IFl and IF 2, differed significantly from each other when tested
against the interaction between treatment and block, which was not significant. When testing treatment effects for each sampling date separately, crop dry mass differences between the irrigated treatments were significant on some dates, but these differences were not consistent over time. Crop dry mass was significantly higher in at least one of the irrigated treatments than in C O and C 1 at almost all dates. Differences between C O and C 1 were only significant at the final harvest in 1995. At harvest in March 1994, crop dry mass was significantly higher in IF 2 than in C O and C 1. In April 1995, crop dry mass was significantly lower in C 1 than in I 1 and IF 1, and significantly lower in C O than in the other four
1.8
?
q"E a,
a,,
,A',
0.9 ~ ~i
P
\
.....
_
............
~
0.6 0.3 0.0 Apr Jun Aug Oct Dee Feb Apr Jun Aug Oct Dec Feb Apr
1993
1994
1995
Fig. 3. Dynamics of above-ground total dry mass between spring 1993 and the final harvest on 5 April 1995: Pre-treatment means ( * ) ; C 0, unfertilized control ( O ;- - -); C 1, fertilized with 15 g m - 2 y r - 1 solid N-fertilizer each spring ( [] ; - - - - - - ) ; I1, irrigated daily, fertilized as C l ( A ; - - ) ; IF1, drip irrigated and fertilized daily (15 g m -2 yr 1) ( 0 ; - - ) ; IF 2, drip irrigated as IF 1 and fertilized daily (42 g m -2 in 1993 and 45 g in 1994) (O; . . . . . ).
158
T. Kiitterer et al. / Field Crops Research 55 (1998) 153-164
Table 2 Absolute (Adry m a s s / A t ; g dry mass m 2 d a y - 1) and relative (A log e (dry m a s s ) / A t ; % day l) growth rates (within parentheses) during the first two months of the growing seasons in 1993 and 1994. Treatments with the same letter are not significantly different ( P > 0.05). Treatments: Co, unfertilized control; C 1, fertilized with 15 g m -2 yr-1 solid N-fertilizer in spring; 11, irrigated daily, fertilized as C l; IFI, irrigated and fertilized daily (15 g m -2 y r - l) through a drip-tube system; IF 2, irrigated as I F 1 and fertilized daily (45 g m -2 y r - 1) through a drip-tube systemDa

te 1 Date 2
At 12 21 15 14
CO 7.9bc (6.1bc) 17.2a (a589312a647d27284b735103b) 16.2bc (2.5ab) 4.3ab (0.4a)
C1 7.3 c (5.7 c) 18.3a (5.4ab) 23.4ab (3.4a) - 3.1b (0.0a)
11 12.6ab (8.4ab) 19.1a (4.6ab) 31.7a (3.6a) 31.9a (2.5a)
IF l 12.4ab (a589312a647d27284b735103b) 25.0a (5.6a) 10.4c (1.2b) 20.5ab (1.7a)
IF 2 15.2a (9.4a) 17.1a (4.0b) 28.6ab (3.5a) 11.7ab (1.0a)
1993 05/05 05/17 06/07 06/22 1994 05/07 05,/23 06/06 06/21
05/17 06/07 06/22 07/06
05/23 06/06 06/21 07/05
16 14 15 14
8.6ab (7.0c) 6.2c (2.7c) 18.0a (4.4a) - 1.8a ( - 0.4a)
6.1b (5.0c) 11.7bc (4.6bc) 21.5a (4.5a) 12.9a (1.8a)
7.7ab 24.5a 26.1a 12.9a
(8.0ab) (7.6a) (3.5a) (1.4a)
12.7a (14.0a) 15.2b (4.7bc) 31.5a (4.8a) 10.5a (1.1a)
12.2ab (13ab) 20.0ab (5.9ab) 36.4a (5.0a) 2.5a ( - 0.1a)
treatments. Differences between the other treatments were not significant at the final harvest. Crop dry mass in C O was significantly smaller during the summer of 1994 than during the summer of 1993, whereas no consistent differences between years were observed in the other four treatments. Dry mass changes during the winter were significant only during the winter 1993-1994, when crop dry mass decreased in all treatments. Growth rates during May and June were not always higher in the irrigated treatments (Table 2), but maximum rates were highest in the irrigated treatments. Rates exceeded 30 g m -2 day -1 in 11, IF 1 and IF 2. Above-ground crop dry mass was linearly related to intercepted PAR during May and June during both years (Fig. 4). For the irrigated treatments, the slope of a linear regression between crop mass and intercepted PAR, i.e., the difference between PAR measured above and below the crop (cf. Fig. 2), was 0.57 and 0.52 g m o l - l for 1993 and 1994, which corresponds to about 2.6 and 2.4 g MJ - l , respectively.
However, from the end of June onwards, the N concentration in IF 2 was higher than in 11 and IF 1. The initial N concentration in spring 1994 was about twice that measured in spring 1993. Estimates of the amount of N stored in the above-ground parts of the crop (Fig. 6) mainly reflect the biomass dynamics (Fig. 3). The decrease in crop N after a peak in summer was more pronounced in 1993 than in 1994. In 1993, about 50% of the N1.6 ? E 13E 1.0
A 1993
0.7
74yj"~ 0.4 '~ 0.1
/ I0.5 1.0 1,5PAR (kmol rn-2)
-0.20.0
20
2.5
3.0
3.3. Crop nitrogen dynamicsDuring both years, nitrogen concentrations in above-ground dry mass decreased rapidly from spring to autumn (Fig. 5). After May 1993, they were lowest in C 0. During early summer, the N concentration was lower in IF 1 and IF 2 than in C 1 and I 1.
Fig. 4. Relationship between above-ground crop dry mass (kg m 2) and PAR (kmol m - z ) intercepted from I May to 6 July in 1993 ( - - , large symbols) and 1994 ( - - ; small symbols) in the irrigated treatments. Lines denote linear regressions (R 2 = 0.99 and 0.96, respectively). Treatments were: I l, irrigated and fertilized with 15 g m -2 yr I solid N-fertilizer each spring (A); IFl, drip irrigated

and fertilized daily (15 g m -2 yr -1) ( 0 ) ; IF 2, drip irrigated as described for IF 1 and fertilized daily (42 g m - 2 in 1993 and 45 g in 1994) (C,).
T. Kiitterer et aL / Field Crops Research 55 (1998) 153-1646
159
g s4o
~2o 1 0 Apr Jun Aug Oct Dec Feb Apr Jun Aug Oct Dec Feb Apr ~ "~'-e-.e_'~_"~--""Z ........ "~ ~'e~'e-e--e. . . . . . . . . . . -o
1993
1994
1995
Fig. 5. Nitrogen concentrations in above-ground dry mass between spring 1993 and the final harvest on 5 April 1995: Pre-treatment means ( * ); C 0, unfertilized control ( O ;- - -); C 1, fertilized with 15 g m - 2 yr - l solid N-fertilizer each spring ( [] ; - - - - - - ) ; I i, irrigated daily, fertilized as Cj ( A ; - - ) ; IF1, drip irrigated and fertilized daily (15 g m -2 yr 1) ( O ; - ) ; IF 2, drip irrigated as IF 1 and fertilized daily (42 g m -2 in 1993 and 45 g in 1994) ( ~ ; . . . . ).
mass was lost from above-ground crop dry mass during late summer and early autumn.
3.4. Changes in nutrient content during winterGenerally, concentrations of nutrients were highest in C 1 and IF 2, intermediate in 11 and IF 1 and lowest in C O at both main harvests (Table 3). At the final harvest, when the chemical analysis was conducted on samples from each block, treatment effects could be tested, and differed significantly ( P < 0.05)
regarding all elements except for C and Ca. For example, the N concentration in C~ was about twice as high and that in IF 2 was about 2.4 times as high as in C O. Nutrient concentrations in each respective treatment were similar on the two main harvest occasions. During the winter no general changes (except for K) in nutrient concentrations were recorded. Concentrations of K decreased considerably in all treatments during the two winter periods, i.e., by 5 3 - 6 8 % during the winter of 1993-1994 and by 6 4 - 7 9 % during the winter of 1994-1995.
30
~-
~°~o Apr
4?\.. ~,,, *'~'\ ii ', ,'~', \/
~\ t"~,\ /,----_______ ---o ,,I A:..~ , ---
° '~'~'~-~-~-~-----~-~-7.~V ￠~'~'o~.................Jun Aug Oct Dec Feb Apr Jun Aug Oct Dec Feb Apr
1993
1994
1995
Fig. 6. Nitrogen mass in above-ground crop between spring 1993 and the final harvest on 5 April 1995: Pre-treatment means ( * ) ; C 0, unfertilized control ( O ; - - -); C1, fertilized with 15 g m -2 yr i solid N-fertilizer each spring ((:3 ; - - - - - - ) ; Ii, irrigated daily, fertilized as C l ( A ; - - ) ; IF1, drip irrigated and fertilized daily (15 g m 2 y r - 1 ) through a drip-tube system ( 0 ; - ) ; IF2, drip irrigated as IFj and fertilized daily (42 g m -2 in 1993 and 45 g in 1994) (C~;. . . . ).
160
T. Kiitterer et a l . / Field Crops Research 55 (1998) 153-164
Table 3 Carbon and nutrient concentrations (% of dry mass) in above-ground dry mass at the main harvests in spring 1994 and 1995, and (values within parenthesis) relative to those measured the preceding autumns in 1993 and 1994 (100). For 1995, replicates were available (n = 4) and treatment differences were

tested C1994 Co C1 11 IF l IF 2 1995 CO C1 I~ IF~ IF 2
N (97) (99) (101) (101) (99) 0.77 1.49 1.20 1.18 1.41 (128) (164) (167) (126) (114)
P 0.10 0.16 0.15 0.14 0.16 (101) (155) (114) (88) (91)
K 0.22 0.25 0.30 0.24 0.28 (40) (41) (47) (32) (33)
Ca 0.29 0.42 0.30 0.26 0.36 (121) (127) (107) (100) (100)
Mg 0.05 0.07 0.05 0.05 0.07 (115) (110) (93) (84) (89)
S 0.08 0.13 0.11 0.10 0.12 (88) (118) (115) (91) (87)
44.3 45.2 44.4 44.8 44.9
44. la 43.9a 43.6a 44.4a 43.8a
(101) (103) (103) (101) (103)
0.62d (84) 1.30ab (113) 1.07c (107) 1.18bc (104) 1.49a (95)
0.09c (63) 0.15ab (106) 0.13b (81) 0.14ab (96) 0.16a (88)
0.21c 0.47a 0.33b 0.32b 0.33b
(21) (36) (32) (29) (24)
0.37a 0.57a 0.44a 0.45a 0.54a
(109) (188) (152) (167) (164)
0.08b (98) 0.12a (148) 0.10ab (143) 0.10ab (141) 0.1 lab (133)
0.07c 0.14a 0.12b 0.12b 0.15a
(64) (99) (76) (75) (79)
Treatments with the same letter are not significantly different ( P > 0.05). Treatments: Co, unfertilized control; C 1, fertilized with 15 g m -2 yr 1 solid N fertilizer in spring; 11, irrigated daily, fertilized as C1; IF 1, irrigated and fertilized daily (15 g m -2 y r - 1) through a drip-tube system; IF2, irrigated as IF I and fertilized daily (45 g m -2 y r - 1) through a drip-tube system.
3.5. Necromass dynamicsThe dry mass produced in 1992, which was not harvested, was reduced by about 50% between October 1992 and April 1993 (Fig. 7). Thereafter, the reduction continued and was faster in the irrigated treatments than in C O and C 1 (significant differences between 11, IF l, IF 2 and C O and C 1 before and after
August, respectively). During 1994, stubble and other necromass (produced during 1993) decreased during the course of the growing season. Although significant treatment differences were recorded on some dates, the overall dynamics was not significantly different between treatments. Difficulties in separating abscissed dead leaves produced during the current year (1994) from the necromass originating
0.4-
E
0.3 "
,~
E ￡
0.2-
:'~,~.~k
o I
Oct1992
Dec
Feb
Apr
Jun
Aug
Oct
Dee
Feb
Apr1994
Jun
Aug
Oct
1998
Fig. 7. Above-ground necromass between 21 October 1992 and 23 October 1994: Pre-treatment means (*); C o, unfertilized control (? ;- -); C 1, fertilized with 15 g m -2 yr-1 solid N-fertilizer each spring ( n ; - - - - - - ) ; Ii, irrigated daily, fertilized as C 1 ( A ; - - ) ; IF1, drip irrigated and fertilized daily (15 g m 2 y r - 1 ) ( 0 ; - - ) ; IF 2, irrigated as IF l and fertilized daily (42 g m - e in 1993 and 45 g in 1994) through a drip-tube system ( ~ ; -).
T. K~itterer et al. / Field Crops Research 55 (1998) 153-1643,0-
161
E.5 2.5c
8 2.O ~
§i
~ 1.51.0-
~
"~'~"
"~"~ -~.
~a
~, / ~
4>
Ez
0.50.0
?
Jun
Aug 1993
Od
Dec
Feb
Apr 1994
Jun
Aug
Oct
Fig. 8. Nitrogen concentrations in above-ground necromass between 22 June 1993 and 23 October 1994: C 0, unfertilized control (C);- - -); C1, fertilized with 15 g m -2 y r -

1 solid N-fertilizer each spring (I-q; - - - - - - ) ; Ii, irrigated daily, fertilized as C l (A;--); IF1, drip irrigated and fertilized daily (15 g m -2 yr-1) ( 0 ; - - ) ; IF2, drip irrigated as IF l and fertilized daily (42 g m - 2 in 1993 and 45 g in 1994)
(0;
). 3). However, those measured on 23 October 1994 were similar to those in the crop. In 1994, all nutrient concentrations were higher in IF 2 and lower in C O than in the other treatments.
from the previous year (1993) caused a high temporal variation in this fraction. N concentrations in the necromass fraction fluctuated considerably during the growing seasons, and no consistent differences between treatments were observed (Fig. 8). However, N concentrations were considerably higher during 1993 than during 1994. Concentrations of N and other nutrients in the necromass on 22 October 1993 (Table 4) were considerably higher than those in crop dry mass (Table
4. Discussion 4.1. Growth and nutrient dynamics Differences in growth between the treatments were mainly due to irrigation (Fig. 3). Growth was similar in the irrigated treatments. Although crop nitrogen concentrations were highest in IF z (Fig. 5), this did not result in greater dry mass production (Fig. 3). Differences between C O and C 1 were due to N fertilization, which is reflected by the relatively low crop N concentrations in C O (Fig. 5). General weather patterns (Fig. 1; Table 1) probably induced the differences in biomass dynamics between 1993 and 1994, and the relatively cold May in 1994 retarded growth slightly (Fig. 3). Growth rates (on average over all treatments) were slightly lower during May 1994 compared with May 1993 although the difference was not significant (Table 2). This conclusion is supported by the observation that the crop's efficiency in converting intercepted PAR to biomass (i.e. light-use efficiency, e) was lower during spring 1994 than during spring 1993 (Fig. 4). A function describing the linear dependence of pro-
Table 4 Carbon and nutrient concentrations (% of dry mass in pooled samples) in above-ground necromass at the last sampling date in October 1993 (no necromass remained in the irrigated treatments) and October 1994 C N 2.19 2.43 P 0.26 0.31 K 0.62 0.68 Ca 1.17 1.58 Mg 0.29 0.33 S 0.20 0.23
1993 CoC1
37.5 37.5
1994 Co Cl I~ IF 1 IF2
39.0 35.9 35.8 37.4 34.7
0.78 1.35 1.45 1.26 1.58
0.11 0.17 0.21 0.18 0.23
0.54 0.77 0.86 0.80 0.92
0.87 1.52 1.35 1.20 1.57
0.21 0.32 0.29 0.28 0.35
0.08 0.13 0.16 0.14 0.16
Treatments: Co, unfertilized control; C~, fertilized with 15 g m 2 yr - I solid N-fertilizer in spring; I l, irrigated daily, fertilized as C~; IF 1, irrigated and fertilized daily (15 g m -2 yr 1) through a drip-tube system; IF2, irrigated as IF l and fertilized daily (45 g m-2 yr ~) through a drip-tube system.
162
T. Kiitterer et al. / Field Crops Research 55 (1998) 153-164
duction on intercepted light was first introduced by Monteith (1977), and changes in e may be att

ributed to variation in leaf photosynthetic properties, structure of the canopy and relationships with environmental factors (and respiration). Monteith (1977), comparing a range of crops, found that the average seasonal value of e was 1.4 g MJ-1 based on total solar radiation, or 2.8 g MJ-1 based on PAR (values based on PAR will be about double those based on total solar radiation (Gallagher and Biscoe, 1978)). Cannell (1989), in a review, reported values for a range of crops to be in the range of 1-1.5 g M J - ~ (or 2 - 3 g M J - 1 based on PAR), and Eckersten and Nilsson (1990) calculated to be 1.96 for Salix viminalis in southern Sweden. Hence, our data for 1993 (2.6) and 1994 (2.4) are well within the range of values reported in the literature. Variations in e may also be attributed to other factors, such as nutrient status. However, this was probably not the case in our experiment since N concentrations were almost twice as high in May 1994 as compared with May 1993 (Fig. 5), and reserves of carbohydrates in below-ground plant organs should have been higher in 1994 than in 1993. In our experiment, the combination of relatively high temperatures and low precipitation in June and July 1994 caused water stress in C o and C~ and retarded growth to a greater extent than in 1993 (Fig. 3; Table 1). Temporal variation in crop dry mass estimates during summer and autumn were higher during 1994 than during 1993. The drop in dry mass during September 1994 coincided with lodging and a heavy aphid infestation, especially in the irrigated treatments. Later, during autumn, the crop recovered, and dry mass increased again. Crop dry mass was greater at harvest in spring 1995 than in October 1994 in four of the five treatments, but this difference was not significant. The fact that necromass (originating from 1993) could not be distinguished from standing dead mass (originating from 1994) at that time, and thus was included in the 'crop dry mass' fraction, could partly explain the dry mass increase during winter 1994-1995. The total above-ground crop mass in C~ on 16 August (1.2 and 0.9 kg m -2 in 1993 and 1994, respectively) was similar to the mean biomass mea-
sured on the same date in 18 similarly N-fertilized field experiments in central and southern Sweden (0.92 kg m -2 excluding stubble, i.e., about 1.0 kg if stubble would have been included; Lomakka, 1993). The dry mass measured during summer in C 1 was also similar to the mean crop dry mass (1.1 kg m -2) of about 100 Norwegian field trials (Ostrem and Oyen, 1985) and to yields in Minnesota (1.1 kg m - 2 ; Sheaffer et al., 1990). The dry mass produced during the establishment year, i.e., 0.4 kg m -2 (Fig. 7), is also very similar to that reported for 100 field trials (Ostrem and ~yen, 1985). Above-ground crop dry mass in the irrigated treatments in summer was similar to the mean yield (various treatments) reported for 24 German experiments in areas with a relatively high groundwater table (K~iding an

d Kreil, 1990), i.e., 1.4 kg m -2 (two cuts per year differing in stubble height). In one experiment in Sweden in 1983, production in a two-cut system was 2 kg m -2 (Tuvesson, 1989). This high amount probably cannot be reached in a single-cut system, at least not with the cultivar used here (Venture). However, about 1.5 kg m -2 may be produced under favorable conditions (see main harvest in 1995; Fig. 3), which would correspond to an exported yield of about 1.2 kg m 2 (standing crop mass at harvest minus stubble and harvest residues), assuming 0.15 kg stubble and 0.15 kg waste in conventional harvesting (Hadders, 1994). Under less favourable winter conditions, the exportable yield is probably well below 1 kg m -2 (see main harvest in 1994; Fig. 3). These yields are much lower than those that can be produced by, for example, winter wheat. In a corresponding experiment (daily irrigation/fertilization), winter wheat produced 2.1 kg dry mass m -2, whereof grain mass accounted for about 40% (Flink et al., 1995). Assuming that 80% of the straw can be exported, this winter wheat crop would have yielded about 1 kg m -2 of grain (15% water content) and 1 k g m - 2 of straw dry mass for combustion. A major factor explaining why reed canarygrass did not reach the production level measured in winter wheat is probably the lack of straw stiffness. After a peak in summer, the crop lodged, whereafter new shoots emerged from the lower nodes of the straw. This was probably the main reason for the large fluctuations in crop dry mass during summer
T. Kiitterer et al. / Field Crops Research 55 (1998) 153-164
163
and autumn, which were more pronounced in the irrigated treatments, especially during 1994 when the crop also was heavily attacked by aphids. Most cultivars used in Sweden originate from areas where reed canarygrass is used for forage production. Plant breeding has concentrated on reducing alkaloid concentrations and increasing seed production (Ostrem, 1987, 1988). Unless plant breeding can result in stiffer straw, the one-cut system will probably not be able to compete with winter wheat straw as an energy crop in central and southern Sweden. However, in northern Sweden, where winter wheat cannot be grown and where the use of short-rotation forest (Salix etc.) is limited by the climate, reed canarygrass may offer a viable alternative. Corresponding to our results in 1993, Lomakka (1993) reported a decrease in biomass in experimental crops harvested later during the season; i.e., compared with mid-August, biomass had decreased by about 20% by the end of September and by about 40% by the following spring. However, this trend was not observed in our experiment during 1994, when crop dry mass in C 1 did not change significantly between August or September and the following April. Differences in mass losses between October 1993-1994 and the following main harvests, i.e., about a 40% decrease in 1993 and no decrease in 1994, were probably not due to weather con

本文来源：https://www.bwwdw.com/article/p3pl.html