A REVIEW OF NITROGEN RESEARCH WITH IRRIGATED PASTURES IN NORTHERN VICTORIA

G. N. Mundy

Kyabram Dairy Centre, Institute of Sustainable Irrigated Agriculture, Agriculture Victoria

Department Natural Resources Environment

1999

Acknowledgments

The author thanks Peter Doyle, Richard Eckard, Kevin Kelly and Richard Stockdale for their comments on various drafts of the manuscript. This review was commissioned as part of the statewide nitrogen research project Best Management Practices for Nitrogen in Intensive Pasture Production Systems, with funding provided through the Victorian State Government Agriculture and Food Initiative.

Disclaimer

This publication may be of assistance to you but the State of Victoria and its officers do not guarantee that the publication is without flaw of any kind or is wholly appropriate for your particular purposes and therefore disclaim all liability for any error, loss or other consequences which may arise from you relying on any information in this publication.

Department of Natural Resources and Environment 1999

ISBN 0 7311 4319 1


CONTENTS

1. SUMMARY

2. INTRODUCTION

3. PASTURE GROWTH RESPONSES TO NITROGEN FERTILISER

Rate of application

Nitrogen efficiency

Late autumn/winter responses

Late winter/early spring response

Spring responses

Summer responses

Autumn responses

Annual application rate

4. RECOVERY OF FERTILISER IN PASTURE

5. FACTORS AFFECTING RESPONSE

Soil mineral nitrogen under irrigated pastures

Influence of other nutrients

Frequency of defoliation

Timing of N application within a pasture regrowth period

6. IMPACT ON CLOVER

Nitrogen fixation

Clover production

7. Fertiliser Types

8. Soil acidity

9. IMPLICATIONS FOR ANIMAL PRODUCTION

Response under mowing and grazing

Animal production

Pasture nutritive characteristics

10. ENVIRONMENTAL IMPACT OF NITROGEN FERTILISERS WITH IRRIGATED PASTURES

Denitrification

Ammonia volatilisation

Surface runoff

Leaching

11. CONCLUSIONS AND DIRECTIONS FOR FUTURE RESEARCH

12. REFERENCES


1. SUMMARY

This review examines nitrogen (N) research with irrigated pastures. It presents an assessment of our understanding of the N nutrition of these pastures, with particular emphasis on N fertiliser use on irrigated pastures. A companion review has been published on the use of N fertilisers on pastures for dairy production in rainfed areas of south-eastern Australia (Eckard 1998).

N fertilisers are commonly used on irrigated dairy farms to help fill expected feed gaps. Although the amount and frequency of use varies, most farmers use N strategically to overcome short-term pasture deficits. Pasture production can be increased by N fertiliser application at most times of the year. The shape of a pasture growth response function with increased amounts of applied N can be either linear or curvilinear and can vary from season to season. However, the upper limit for a single application of N, where pastures are defoliated about every 3 to 4 weeks, is 50 kg N/ha. This upper limit matches the growth requirements of rotationally grazed pastures, is unlikely to adversely affect clover production, and will only impact on N2 fixation temporarily.

In late autumn and winter, growth responses by perennial pastures to N fertiliser are limited by relatively cold temperatures and short daylengths, and the size of the responses can often be less than 10 kg DM/kg N applied. However, applications of N in mid to late winter frequently produce DM responses greater than 10 kg DM/kg N in 4 to 6 weeks, depending on the time of application. In spring and summer, pastures are responsive to N but the magnitude of the response depends on pasture type, ryegrass density and defoliation interval. During summer, paspalum dominant pastures are more responsive than ryegrass types. However, in autumn the ryegrass type remains responsive to applied N later into the season. The use of 15N-labelled fertilisers has shown that less than 50% of the applied N was present in the harvested herbage following the initial growth response. Therefore, plant uptake of fertiliser N may only contribute up to 40% of the total N in the harvested herbage. It is concluded that more research is needed to better understand the reasons for the variation in responses to improve the prediction of pasture growth responses to N.

The annual DM production of irrigated perennial pastures can be increased with repeated applications of up to a total of 900 kg N/ha/yr. With annual applications of up to this size, annual DM production is curvilinearly related to the total amount of N applied. However, when the application rate is 500 to 600 kg N/ha/yr, annual DM production can be linearly related to total N applied. Nevertheless, there are detrimental effects of using high N inputs, especially on clover production and N2 fixation in mixed pastures, and on the environment, which questions the sustainability of this practice. The experimental evidence indicates that clover production in irrigated pastures can decline where applications of 100 to 200 kg N/ha are applied over several months. Clover growth in ryegrass pasture can decrease when more than 100 kg N/ha is applied. Further work in this area may help improve our understanding of the mechanisms involved.

Previous research has found that frequency of defoliation and timing of N application within a pasture regrowth period can influence the growth response. The role of basal soil fertility on pasture DM response to N requires further investigation. Soil acidity can be increased with N application, but it appears that large amounts of N (except ammonium sulphate) can be applied before the pH of northern Victorian soils is significantly affected. Experiments that compared different types of N fertilisers have found that there is very little difference in their effectiveness at increasing pasture growth, if they are applied when the soil and environmental conditions favour their use.

There is some evidence that responses under mowing and grazing can be similar. However, areas of pasture affected by animal excreta would not be responsive to N fertiliser. Therefore, differences between mown and grazed pastures (eg. nutrient cycling and defoliation intensity) should be considered when extrapolating responses from mown to grazed pastures. Three stocking rate experiments with dairy cows have found that strategic N applications to irrigated pasture had little effect on total milk production.

The digestibility of grass is not altered by application of N fertiliser. However, the overall pasture digestibility may tend to be slightly lower with N, probably because of the higher content of grass, which is often of lower digestibility than clover. The N content of pasture is often unaffected by the application of N fertiliser, even though the N content of grass can be higher with N.

The loss of N fertiliser from irrigated pasture can have environmental and production implications. In northern Victoria, N balance data from 15N-labelled fertiliser experiments have indicated that there is a high potential for gaseous losses (denitrification or ammonia volatilisation) of N from fertiliser applied to irrigated pastures. Research has found that up to 35% of ammonium nitrate applied to flood irrigated pasture could be lost, presumably via denitrification. The N lost from urea was less than 14% under the same flood irrigation conditions.

Leaching of N to deeper layers of the heavier soil types commonly used for perennial pastures appears to be relatively low. However, some recent work suggested that leaching of fertiliser N could occur in these heavier soils, presumably via preferential pathways (macropore flow). Further work in this area is required to better understand the potential for N movement through the different pathways in soil profile.

Ammonia volatilisation loss from urea applied to pasture can be minimised by washing the urea into the soil with an irrigation or rainfall. However, without irrigation or rain, the loss of urea-N could be up to 20% in the cooler months of the year, presumably via ammonia volatilisation. Research into these loss processes is needed to quantify the likely annual loss of N from irrigated pastures grazed by dairy cows.

Surface runoff of irrigation water from pasture bays can contribute to the loss of applied N. Recent work has shown that up to 14% of applied urea-N can be lost in drainage water from irrigation. The majority is lost with the first irrigation after urea application, but this loss can be greatly reduced by using unfertilised buffer strips at the end of pasture bays.


2. INTRODUCTION

In south eastern Australia, irrigated pastures used for dairying are located in southern NSW, northern Victoria (including the north east region), the Macalister irrigation district in Gippsland, and the Murray lakes and swamps in South Australia. Milk production in these areas amounts to about 30% of total production in Australia. Most of this production occurs in northern Victoria.

In the northern irrigation region of Victoria, irrigated pastures are sown to perennial ryegrass (Lolium perenne L.) and white clover (Trifolium repens L.), but over time, they become dominated by paspalum (Paspalum dilatatum Poiret.) in summer. The white clover content can vary from less than 10% to above 50% on a dry matter (DM) basis. Therefore, in this review, irrigated perennial pastures are referred to as a ryegrass or a paspalum type depending on the dominant grass species. Some dairy production systems also incorporate irrigated annual pastures which are predominantly based on annual clovers, such as subterranean (Trifolium subterreaneum L.) or Persian (Trifolium resupinatum L.) clover. Perennial pastures can be irrigated from August to May, whereas annual pastures only receive irrigations in autumn and early spring. The environmental conditions, which apply in northern Victoria in different months, are given in Table 2. Clearly, pasture growth rates in northern Victoria will be limited by temperature and daylength during winter. In addition, the irrigation season finishes in May and the frequency and intensity of follow up rain can markedly affect pasture growth by either waterlogging the soil or limiting the availability of soil water to the plants in winter.

Nitrogen (N) fertiliser use on dairy farms has increased through the 1980’s and 90’s (see Eckard 1998). The use of N fertiliser is now promoted and is an accepted management option to help fill expected feed gaps on many irrigated dairy farms. However, the amount used generally does not exceed 200 kg N/ha/yr, which is below the recommended annual N applications for dairy farms in Western Europe (Whitehead 1995; Crosse and Dillon 1996).

 

Table 1. The mean monthly environmental conditions for Kyabram in northern Victoria (Source: Kyabram Dairy Centre).

Month

Maximum temperature (oC)

Minimum temperature (oC)

Rainfall (mm)

Evaporation (mm)

Sunshine (h/day)

Terrestrial temperature (oC)

January

30

14

35

266

9.8

11

February

30

15

24

222

9.7

12

March

26

12

29

173

8.5

10

April

22

9

39

97

7.1

6

May

17

6

48

52

5.2

4

June

14

4

41

32

4.5

2

July

13

3

45

36

4.4

1

August

15

4

46

57

5.2

2

September

17

5

47

85

6.5

3

October

21

7

43

136

7.7

5

November

24

10

32

191

9.0

8

December

28

12

31

244

9.9

10

 

It is suggested that N fertiliser use on irrigated dairy farms should only be used where pasture shortages are foreseen. For convenience, irrigated dairy farming systems are categorised as:

1. No N use (low stocked farms)

2. Strategic applications to overcome short-term pasture shortages (low to medium stocked farms)

3. Repeated or multiple applications to fill feed gaps over several months (high stocked farms).

Anecdotal evidence suggests that most farms in northern Victoria are probably in category 2.

Previous reviews which consider N fertiliser use on pastures relevant to southern Australia include those by Ellington (1986), McGowan (1987), Simpson (1985, 1987), Whitehead (1970; 1995), and Eckard (1998). The purpose of this review is to examine N research with irrigated pastures and present a critical assessment our current understanding of the N nutrition of these pastures. It complements the review of Eckard (1998) which covered N research for rainfed pastures of south eastern Australia, and information from that paper will be used to supplement that presented here.


3. PASTURE GROWTH RESPONSES TO NITROGEN FERTILISER

The majority of the growth response to N fertiliser applied to recently defoliated irrigated pasture occurs by the next defoliation, often in 3 to 4 weeks (Mundy 1993). The carry over effect of the fertiliser on DM production in the second and subsequent pasture regrowth is usually relatively small compared with the initial growth response (Mundy 1993). The size of the residual DM response tends to be greater when the initial response is large.

 

Rate of application

In northern Victoria, Roufail (1978) has reported linear and curvilinear responses in total annual DM production to increasing fertiliser applications up to 112 kg N/ha per application. This and previous work (eg. Martin 1962) examined the effect of different rates of N applied a number of times over the year on either seasonal or annual DM production response. The effect of increasing N fertiliser as a single application on growth responses was not reported. However, Mason et al. (1987) showed that responses from a single application varied from season to season, and suggested that different factors had influenced the effects of N on pasture growth at different times of the year.

Factors affecting seasonal responses could include: soil N mineralisation (including nutrient recycling), pasture species, grass density, clover content, frequency and severity of defoliation, status of other nutrients, direct loss of fertiliser N, and other environmental factors such as water availability, temperature and levels of pests and diseases. Some of these factors are discussed later in this review.

The slope of the pasture growth response function to N fertiliser has implications for the economic optimum amount of N to apply. However, setting of an upper limit for a single N application on only economic grounds takes no account of any possible adverse effects on the pasture after an initial response. Mundy (1995a, 1996) suggested that where irrigated pastures are defoliated every 3-4 weeks, the upper limit for a single application should be 50 kg N/ha. With grass-clover pastures, this upper limit matched the N requirements of rotationally grazed pastures, was unlikely to have an adverse affect on clover production, and would not cause a decline in production after the initial response (Table 2). With an application of 100 kg N/ha and 3-4 week defoliation intervals (regrowth periods), there may be luxury uptake of N (including an elevated nitrate concentration in the herbage) with insufficient time to use the applied N to maximise the growth response (Mundy 1996). In addition, with 0, 50 and 100 kg N/ha applied, the reduction in clover growth with the highest application rate severely impacted on the amount of N2 fixed by clover over the 18 weeks after the initial application (Mundy 1991).

Table 2. The effect of rate of applied nitrogen (single application) on DM production and clover content in a ryegrass-white clover pasture in spring/summer (Source: Mundy 1996).

Nitrogen

Rate

DM production over 25

Days

Clover content after 25 days

DM production over 18 weeks (t DM/ha)

(kg N/ha)

(t DM/ha)

(kg DM/kg N)

(% DM basis)

Pasture

Clover

0

1.2

 

55

5.8

2.7

50

1.7

10

35

6.7

2.8

100

1.9

7

33

5.9

2.1

l.s.d. (P=0.05)

0.20

   

0.73

0.45

 

Twentyman (1970) and Roufail (1978) have also suggested that about 50 kg N/ha was usually the most efficient rate of application for irrigated pasture. While Eckard (1998) made similar conclusions for rainfed perennial pastures, it was based on the diminishing response to N at higher rates of application.

 

Nitrogen efficiency

While N2 fixation by clover often promotes more production in mixed grass-clover pastures than occurs with pure grass swards (without applied N), responses as kg DM/kg N by grass swards are higher than those of mixed swards (Whitehead 1970). Responses to applied N in irrigated perennial pastures which are white clover-based are likely to be lower and more variable than those in pure grass swards. This has important implications for predicting the likely response to N fertiliser applied to irrigated pastures. Where pastures are irrigated through summer, N can be applied at any time during the year and it is important to understand differences in response within and between seasons.

 

Late autumn/winter responses

Pasture growth responses have ranged from about 3 to 25 kg DM/kg N from split applications of N fertiliser (usually 4) between late autumn and late winter (Garland et al. 1962; Martin 1962). The application rates in these studies ranged between 0 to 200 kg N/ha. Similar work conducted in the Macalister irrigation district found pasture growth responses of up to 18 kg DM/kg N applied, but over 4 years the average response was 11 kg DM/kg N (Norman 1962). In these early studies, measurements of response in pasture production were usually made in September (eg. Martin 1962), with no sequential measurements taken through winter. Hence, it is possible that most of the response occurred from late winter onwards. Growth of annual pastures that contained a high proportion of grass species such as Wimmera ryegrass (Lolium rigidum L.) or barley grass (Hordeum spp.) also responded to N fertiliser applications in winter (Garland et al. 1961; Martin 1962). The magnitude of these responses varied markedly from very low to about 30 kg DM/kg N. This variation in response was due to the change in annual grass densities from year to year (Kelliher 1962; Martin 1962).

With more frequent measurements, Mundy (1993) found that pasture growth responses in both ryegrass and paspalum-based pastures were much less than 10 kg DM/kg N in winter when 25 and 50 kg N/ha had been applied in late autumn or early winter. It was apparent that 33 to 62 days of regrowth was not sufficient for the pasture to respond when winter pasture growth is limited by daylength and relatively low temperatures. This is in contrast to other dairying areas in southern Australia where air and soil temperatures are higher over winter and pastures may reach a suitable mass for grazing within 4 or 5 weeks (Eckard 1996; McKenzie and Jacobs 1997).

 

Late winter/early spring response

There are no published experiments that have investigated the timing of N applications in winter for an early spring response based on soil temperatures. However, N fertiliser applied in mid to late winter can markedly increase perennial pasture growth by early spring (Mundy 1993; Dickens 1994). In a number of experiments with different pastures, Mundy (1993) found that the DM production response was on average greater than 10 kg DM/kg N applied when the fertiliser was applied from June to August. With earlier applications, a response time of up to 60 to 70 days was needed, whereas with later applications about 4 weeks was required. The data show that applying N in mid-winter can increase pasture growth by early September while an August application can achieve a response by mid to late September. During winter, growth rates of perennial grasses are limited by low mineralisation rates of soil N and so are potentially highly responsive to applied N. The time taken to achieve the response is affected by the environmental conditions following the fertiliser application.

 

Spring responses

Perennial pasture growth rates peak in mid-spring (Stockdale 1983). Although N assimilation rates (kg N/ha/d) by pasture without added N are relatively high in spring (Mason et al. 1987), perennial pastures can respond relatively well to N fertiliser through this season (Kaddous 1968; Roufail 1978; Mason et al. 1987; Mundy 1993). The environmental conditions for plant growth in the region are near ideal for ryegrass and clover providing irrigation scheduling provides water in a way that matches plant requirements.

Growth responses in irrigated perennial pastures in northern Victoria during spring range from about 5 to greater than 20 kg DM/kg N. The lower responses (2 to 8 kg DM/kg N) have occurred with established ryegrass pasture (Mundy 1993), and may be due in part to low ryegrass plant and tiller densities. Lawson et al. (1998) have described the decline in ryegrass plant and tiller numbers with age of pastures in this region and found that it may be a key constraint to DM production. In contrast, Kaddous (1968) reported a 23 kg DM/kg N response for a spring sown ryegrass pasture over spring and summer in the first year after sowing. Further evidence that ryegrass plant density may be a critical factor in response efficiency was that repeated applications of N up to 1200 kg N/ha only reduced the clover content from 50% to 25% of DM (Mundy 1996). The same amount of N applied to a paspalum dominant pasture reduced the clover content from 27% to 1%.

Although grass density may play a major role in any pasture growth response to N fertiliser, the transfer of fixed N in pastures high in clover content to the grasses may also influence responses (Ledgard 1991). To improve predictions of responses to N fertiliser, effects of grass tiller density and transfer of fixed N on responses to N fertiliser require further investigation. This is particularly so for flood irrigated pastures where the N cycling and N mineralisation processes are quite different to those in rainfed pastures, particularly during summer and early autumn.

 

Summer responses

A number of experiments have shown that both temperate and tropical grass species respond to applied N in summer (Roufail 1978; Mason et al. 1987; Mundy 1993). However, during hot northern Victorian summers (Table 1), paspalum dominant pastures respond better than temperate grass dominant pastures (Roufail 1978; Mundy 1993). With optimum temperatures for shoot DM production of ryegrass, white clover, and paspalum being 18 to 24, 24, and 29 oC, respectively (Mitchell and Lucanus 1962), summer temperatures above 30oC favour perennial ryegrass growth least during this season. Furthermore, the requirements for irrigation vary between the main pasture species. White clover requires more frequent irrigation than ryegrass which requires more frequent watering than paspalum (Blaikie et al. 1988).

The differences in response are undoubtably due to the adaptation of paspalum to high day temperatures. On paspalum dominant pastures, Mundy (1993) measured responses ranging from 3 to 30 kg DM/kg N applied, with response times of about 20 days. Such pastures with a 20 to 30% clover content consistently produced responses of 10 to 20 kg DM/kg N by the next defoliation. However, responses as low as 5 to 6 kg DM/kg N were measured with defoliation intervals of around 14 days.

Most pastures dominated by paspalum contain less than 10% clover. These pastures respond to N at the upper limit and, often during the second regrowth after N application, there can be a significant carry over response to the fertiliser (Mundy 1993). Presumably, the higher initial response and the carry over effect of the N are due to the lower contribution of fixed N to available soil N. While there is little doubt that N fertiliser increases DM production from paspalum dominant pastures, the digestibility of this material can be low compared with that of clover and ryegrass (Stockdale 1999). Mundy (1993) also found that while N fertiliser can increase the N content of paspalum, it had no effect on its digestibility.

Irrigated ryegrass pasture responses to N vary between 2 to 20 kg DM/kg N in summer (Mundy 1993). Short defoliation intervals were again associated with lower responses, compared with longer defoliation intervals. The slower growth of ryegrass in summer relative to paspalum suggests that longer regrowth times are needed by ryegrass to maximise the response. As suggested previously, ryegrass plant density could be limiting response.

 

Autumn responses

In autumn, pasture (annual and perennial) growth responses to N are often lower than at other times of the year (Martin 1962; Roufail 1978; Mason et al. 1987; Mundy 1993). A similar observation has been made in other countries (Feyter et al. 1985; Frame and Boyd 1987), where it has been suggested that the lower response may be due, in part to higher soil mineralisation rates in autumn under rainfed conditions (Fetyer et al. 1985). Mason et al. (1987) found that apparent uptake of soil N by perennial pasture was 1.6 kg N/ha/d in early autumn, whereas it was 0.3, 1.3, and 1.3 kg N/ha/d for winter, spring, and summer, respectively. Presumably, lower growth responses by annual pastures to N would be due to the relatively high concentrations of soil mineral N that build up under annual pastures over summer because they are not irrigated.

In early autumn, both ryegrass and pasplaum-based irrigated pastures produce similar responses to N, ranging from 6 to 12 kg DM/kg N applied (Mundy 1993). However, as the season progresses, decreasing temperature and daylength cause paspalum growth rates to decline, and the response by these pastures can decline to less than 6 kg DM/kg N by April (Mundy 1993). Although ryegrass growth rate also declines during autumn (Stockdale 1983), 2 years of production data with N showed that these pastures were capable of reasonable DM responses (5 to 12 kg DM/kg N) well into the autumn (Mundy 1993). The variation in responses were possibly due to the same factors that affect ryegrass response at other times of the year, plant density, soil water relations and defoliation interval. A better understanding of the contribution of, and interaction between, these factors is required to better predict pasture growth responses to fertiliser.

 

Total annual applications

Research in the UK has shown that pastures can respond to applications of up to 800 to 900 kg N/ha/yr (Whitehead 1970), especially when cut relatively frequently (Holliday and Wilman 1965). Both pure grass and grass-clover swards can have similar maximum DM production when a total of 300 to 400 kg N/ha/yr is applied (Whitehead 1995). Annual DM production responses to N fertiliser are essentially linear with applications of this magnitude. In northern Victoria, A.K. Stubbs (unpublished data) measured increased production of irrigated perennial grass-clover pasture with up to 430 kg N/ha/yr, whereas Roufail (1978) found that N applications of up to 896 kg N/ha/yr increased DM production of paspalum dominant pasture. Roufail (1978) reported a strong non-linear relationship between DM production (t/ha/yr) and rate and frequency of application (Fig. 1). Furthermore, annual growth of pasture showed a significant curvilinear relationship with total N applied per year (Fig. 2).

Fig. 1. The annual DM production of paspalum-white clover pasture with 5 rates of N fertiliser applied at 2 (l), 4 (m), or 8 (t) times per year (Source: Roufail 1978).

Fig. 2. The curvilinear relationship between annual DM production of paspalum-white clover pasture and the amount of N fertiliser applied per year (Source: Roufail 1978).

Mundy (1993) applied 25 or 50 kg N/ha at various frequencies to a perennial ryegrass based and a paspalum based pasture over 2 years. The highest cumulative DM production of each pasture type occurred when 50 kg N/ha was applied 24 times, after each of 24 defoliations. The relationship between total cumulative DM production and amount of applied N was linear, presumably as the maximum rate of a single N application was only 50 kg N/ha (Fig. 3).

Fig. 3. Relationships between cumulative DM production over 2 years and the total amount of N applied to ryegrass-white clover (l) or paspalum-white clover (m) pastures (Source: Mundy 1993).

Over the 2 years, the average total pasture response to the applied N was 6 and 8 kg DM/kg N for the ryegrass and paspalum pastures, respectively (Mundy 1993). In contrast, the average response by the non-legume component of the pastures was 12 kg DM/kg N. The loss of clover production with continued use of N reduces the potential pasture growth response (Table 3) unless the grasses compensate for the lost clover production (Mundy 1996). When this occurs, the pasture becomes totally reliant on N fertiliser to maintain its high level of production. Once N applications are stopped, DM production declines because there is no clover to support a lower plane of N supply (Mundy 1993). It would appear that farmers who adopt a high N fertiliser strategy may be advised to only apply it to pure grass swards.

 

Table 3. Total pasture and clover DM production of two pasture types fertilised with 0, 150, 300 or 600 kg N/ha over 12 defoliations in each of two years (Source: Mundy 1996).

a. Ryegrass-white clover pasture

Nitrogen applied

(kg N/ha)

Cumulative pasture DM production

(t/ha)

Cumulative clover DM production

(t/ha)

 

Harvest 1 to 12

Harvest 13 to 24

Harvest 1 to 12

Harvest 13 to 24

0

15.8

15.5

8.5

8.5

150

17.0

16.7

7.7

8.1

300

17.2

16.0

7.1

6.0

600

20.2

18.7

5.3

5.0

l.s.d. (P=0.05)

2.161

1.762

1.48

1.21

1.21

0.99

1.26

1.03

Harvest 1 to 12 = 393 days production. Harvest 13 to 24 = 332 days production.

b. Paspalum-white clover pasture

Nitrogen applied

(kg N/ha)

Cumulative pasture DM production

(t/ha)

Cumulative clover DM production

(t/ha)

 

Harvest 1 to 12

Harvest 13 to 24

Harvest 1 to 12

Harvest 13 to 24

0

17.5

13.3

5.2

4.7

150

19.0

14.3

4.5

3.9

300

20.5

15.6

4.8

3.0

600

22.6

17.7

3.8

1.6

l.s.d. (P=0.05)

1.311

1.072

0.79

0.64

0.87

0.71

0.68

0.55

Harvest 1 to 12 = 377 days production. Harvest 13 to 24 = 321 days production.

1 Within a column, to compare the difference between N fertilised and no N.

2 Within a column, to compare a difference between any 2 N fertilised means.


4. RECOVERY OF FERTILISER N

The uptake of fertiliser N by pasture is often assessed as the percentage apparent recovery of fertiliser in the harvested herbage. The apparent recovery is estimated as the difference in total N content between fertilised and unfertilised pastures. This assumes that the amount of N from the soil and N2 fixation is the same for fertilised and unfertilised pastures, which may not be correct. The apparent recovery of fertiliser N is often reported to be between 50 and 80% (Whitehead 1995). However, apparent fertiliser N recoveries often overestimate the actual recovery in herbage, which can be determined using 15N-labelled fertiliser. In northern Victoria, the actual uptake and recovery of applied N in irrigated pastures has been measured with labelled fertilisers applied to pastures enclosed in microplots that were located within the main plots of fertiliser experiments.

The determination of the actual recovery of fertiliser N in harvested herbage from irrigated pastures showed that about 70% of the total amount of fertiliser N in herbage from 3 consecutive harvests was present at the first harvest, with 24 and 8% of the total fertiliser N in herbage being present at harvests 2 and 3, respectively (Mundy 1993). Clearly, the majority of plant uptake of fertiliser N occurs rapidly after the N application to recently defoliated pasture.

The actual recovery of applied 15N in herbage after 1 regrowth of fertilised pasture has ranged between 13 and 40% (Mason et al. 1987; Mundy 1993). An additional 8 to 12% of the fertiliser N may be present in the pasture stubble. After 2 or 3 defoliations of pasture, Mason et al. (1987) and Mundy (1993) found that the cumulative recovery of fertiliser N in the above ground herbage ranged from 22 to 54%, with a further 7 to 12% in plant roots. The reasons for the low recovery of fertiliser N in pasture could include losses of N by denitrification or ammonia volatilisation, and unfavourable environmental conditions (eg. soil water) affecting plant growth and N uptake. However, the actual recoveries of fertiliser in herbage of irrigated pastures are similar to those reported elsewhere for other pastures (eg. Dowdell and Webster 1984; Dawson and Ryden 1985).

Although the initial growth response accounts for the majority of the applied fertiliser, only 20 to 40% of the N in grass may have come from this source (Mason et al. 1987; Mundy 1993). However, the contribution of fertiliser N increases with the amount of fertiliser applied. The balance of N in herbage comes from N mineralised in soil and possibly some remobilised from the roots and unharvested plant material. It is not only the grass that utilised fertiliser N. Mason et al. (1987) found that white clover could derive between 14 and 22% of its N from fertiliser during the first regrowth after an application.

The contribution of fertiliser to the N in pasture, after the initial growth response, may decrease to about 10 to 20% during a second regrowth and to less than 10% by a third (Mundy 1993). The fertiliser N recovered in herbage in these later harvests has come from the soil and probably translocation of stubble and root N.

Fig. 4. The percentage of pasture N derived from fertiliser N with the 4 repeated application strategies (50 kg N/ha applied after every 4th defoliation (l), 50 kg N/ha applied after every 2nd defoliation (n), 50 kg N/ha applied after each defoliation (t) or 25 kg N/ha applied after each defoliation (s))(Source: Mundy 1993).

The contribution of fertiliser N to herbage N is influenced by the frequency of application, and the rate of N applied. Fig. 4 shows the effect of 4 fertiliser application strategies with ryegrass-white clover on the proportion of N in the pasture derived from the applied urea. With regular applications of 25 or 50 kg N/ha as urea, the fertiliser became a relatively larger contributor to pasture N than with less frequent application of fertiliser. With less frequent applications, the contribution by the fertiliser declines between applications (Fig. 4). Similar findings were obtained with paspalum–white clover (Mundy 1993).


5. FACTORS AFFECTING RESPONSES

Soil mineral nitrogen under irrigated pastures

Ammonium (NH4) is the dominant form of indigenous inorganic N in the soil (0 to 300 mm) under flood irrigated pastures with levels of 9, 9, 8 and 17 kg N/ha in winter, spring, summer and autumn, respectively (Mason et al. 1987). In comparison indigenous nitrate (NO3) levels are usually 3 kg N/ha or less.

The relatively stable and low levels of indigenous inorganic N under irrigated perennial pasture contrasts with dry land pastures in which large accumulations of NO3, in particular, can occur over summer and autumn in response to wetting and drying after rainfall (Simpson 1962). Presumably pasture uptake is largely responsible for the stability of soil inorganic N concentrations in an irrigated system.

Soil mineral N concentrations under irrigated pastures that have had up to 50 kg N/ha applied as frequently as 24 times over 2 years showed no indication of mineral N accumulation (Mundy 1993). The soil N pool data gave no indication of the turnover of mineral N under these pastures but did suggest that responses to N fertiliser are unlikely to be influenced by the size of the soil mineral N pool.

 

Influence of other nutrients

Responses to N fertiliser can be limited by deficiencies of other nutrients, such as phosphorus (P), potassium (K) and sulphur (S) (Whitehead 1995). In northern Victoria, experiments have shown that lime or K applications do not improve pasture growth with or without N (Kelliher 1962; Martin 1962) unless K deficiency has been induced by continual herbage removal (Roufail 1978). Low levels of available S in soils may occur during winter especially in waterlogged soils. However, Mundy (1993) found that the addition of S with N fertilisers applied in mid-winter to a temporarily waterlogged perennial pasture had no effect on the growth response to the applied N.

The soils in the main irrigated pasture areas in south-eastern Australia rely on regular annual applications of P to sustain pasture production. Consequently, in most N response experiments, P has been applied at least annually across all treatments. This practice should have reduced the likelihood of interactions with soil P availabilities and applied N.

Roufail (1978) conducted a 4 year study in which N (0 to 112 kg N/ha per defoliation) was applied to pastures that received 0, 47 and 188 kg P/ha/yr. This work found that pasture production was dependent on both P and N application rates. However, it did not demonstrate a consistent relationship between soil P status and pasture growth response to N. Perhaps growth responses to N may not be sensitive to P availability in soil until P limits grass growth. In this study, there was some evidence that P applications may have helped support clover persistence where multiple N applications were used. For flood irrigated pastures on red-brown earth soils, there is a need to clarify the influence of basal soil fertility and applications of P on responses to N fertiliser.

In northern Victoria, groundwater pumping is part of the regional salinity control strategy. Saline groundwater is diluted with good quality channel water before it is used to irrigate pastures. White clover is more sensitive than ryegrass to salinity, thus clover production in pasture declines before ryegrass when irrigated with saline water (Mehanni 1987). Mehanni and West (1992) suggested that production of ryegrass-white clover pastures could be restricted by both salinity and N deficiency when irrigated with saline water. They found that the application of N fertiliser had no effect on relative salt tolerance of perennial pasture, but that it did increase the production of the grass in salt affected pasture. Therefore, application of urea at 150 kg N/ha to pasture irrigated with saline water up to 3000 mg TDS/L could be justifiable. At these relatively high saline levels, DM responses can vary from less than 10 to 26 kg DM/kg N (Anon 1992). The reasons for these differences in response are not known.

 

Frequency of defoliation

On northern Victorian dairy farms, irrigated perennial pasture paddocks are grazed on a rotation of around 3 weeks during warmer months of the year. This defoliation interval is based on the time it takes for 3 new leaves to theoretically develop on ryegrass tillers. This takes into account DM production (Bartholomew and Chestnut 1977; Chestnut et al. 1977) and pasture digestibility (Chestnut et al. 1977; Frame et al. 1989). Mundy (1993) reported that, with a defoliation interval of 3 weeks in ryegrass-white clover pastures, growth responses to N fertiliser were 10 kg DM/kg N applied, on average. However, Mundy (1993) also found that frequency of defoliation can affect the cumulative growth response of a paspalum dominant pasture to applied nitrogen in late summer. A non-linear model predicted that the overall response of the pasture to nitrogen after 60 days was 13.8, 16.3, 18.8, and 23.8 kg DM/kg N applied, with 10, 15, 20, and 30 day defoliation intervals, respectively (Fig. 5).

Fig. 5. The cumulative DM production of paspalum pasture over 60 days with a single N application of 0 (l), 30 (n), or 60 (s) kg N/ha at each of 4 defoliation frequencies (Source: Mundy 1993).

The increase in apparent response to N with decreased frequency of defoliation indicted that the regrowth period after N was applied was a factor that affects pasture production and responses to N fertiliser. Clearly, the interaction between frequency of defoliation and pasture growth response to N fertiliser will vary for different pasture types and different times of the year. Some farmers have adopted a grazing management of frequent defoliation with regular applications of N fertiliser over the summer. This practice may not be optimum for pasture production, and could increase the non-protein N content of the herbage. Perhaps these pastures can adjust to that management.

 

Timing of N application within a pasture regrowth period

With both paspalum and perennial ryegrass dominant pastures, the time of application of N fertiliser within a regrowth cycle affected not only the production of pasture during that cycle, but also during the subsequent regrowth cycle (Mundy 1993). The optimum time to apply the N fertiliser was within 7 days after defoliation, or a few days before defoliation. A loss of 8 to 15 % of the production response could occur when the fertiliser was applied mid way through a regrowth period. Fertiliser uptake by the grass indicated that unless there was enough time for the N to be used by the pasture, some loss of response was likely. Extrapolating these effects to pastures that are frequently defoliated (eg. at 10 day intervals) would suggest that some loss of potential production response to N would occur.


6. IMPACT ON CLOVER

Nitrogen fixation

In northern Victoria, estimates of N2 fixation by white clover without N fertiliser by the 15N dilution technique have indicated that the proportion of clover N derived from fixation ranged from 70 to 90%, and the amount of N2 fixed by white clover estimated as 150 to 300 kg N/ha/yr in above ground herbage (Mason et al. 1987; Mundy 1993; Smith et al. 1993). This is consistent with the findings of other researchers (eg. Ledgard et al. 1990), but may be considerably higher than Riffkin et al. 1999 reported for western Victoria. However, Mundy et al. (1988) found that within irrigation cycles, fixation activity by white clover was significantly depressed by waterlogging, soil dryness, and N fertiliser application. This work also found that peak nodule activity only occurred towards the end of a drying cycle. It may be possible to overcome the detrimental effects of flood irrigation on N2 fixation by clover by using other irrigation technologies.

Table 5. The estimated amount of N from fixation in the tops of two pastures that were treated with four amounts of N fertiliser over 12 defoliations (Source: Mundy 1993).

N applied (kg N ha)

Ryegrass-white clover1

Paspalum-white clover2

0

278

177

150

204

150

300

206

145

600

118

108

1393 days, 2377 days

 

N fertiliser can reduce the proportion of clover N derived from fixation to as low as 30 to 40% in the initial regrowth following application (Mason et al. 1987; Mundy 1993). The recovery of fixation is either partial or complete in the second regrowth cycle after a single application of N. Hence, the longer term effects of a single application of N on the amount of N2 fixed by white clover will be small unless clover production declines (Mundy 1991). However, the amount of N fixed in pastures treated with multiple applications of N will be severely reduced (Table 5). This is because clover growth (Table 3) and fixation activity (G.N. Mundy unpublished data) were lower in pastures with high rates of N applied. The different amounts of N fixed by the two pasture types (Table 5) are mainly a reflection of their different clover contents.

 

Clover production

Pasture growth responses to N fertiliser usually change the botanical composition in favour of grass (Table 2). This increase in growth of grass does not necessarily reduce clover production unless high rates of N have been applied (Mundy 1993) or there has been an excessively long regrowth time (Eckard 1998). Mundy (1993) found that single applications of 50 kg N/ha applied at different times of the year to ryegrass-white clover and paspalum-white clover did not reduce clover production when these irrigated pastures were defoliated at a pasture mass of 4 t DM/ha or less. These results agree with the conclusion of Eckard (1998) that N should not affect clover growth when pastures are defoliated in the timeframe commonly used by dairy farmers.

In northern Victoria, some dairy farmers use repeated applications of N to boost pasture growth over several months. In contrast to the effects of single N applications, experimental results show that this practice can reduce clover production in irrigated pastures (Roufail 1978; Mundy 1993, 1996). The effect of increased inputs of N fertiliser on clover DM production in irrigated perennial pastures is illustrated in Table 3, with the decline in clover growth inversely related to the amount of N fertiliser applied over time (Mundy 1993).

Fig. 6. The effect on clover DM production in ryegrass-white clover pasture with 6 repeated applications of 100 (l) or 200 (n) kg N/ha applied over 4 defoliations. The effect on clover DM production for each data point on the figure was calculated as the cumulative clover growth after 4 defoliations of N fertilised pasture as a of the cumulative clover growth of unfertilised pasture (Source: Mundy 1993).

Mundy (1993) also showed that the decline in clover production (or clover persistence) with repeated N fertiliser applications was different for ryegrass-white clover and paspalum-white clover pastures. In the case of ryegrass-white clover (Fig. 6), clover DM production with repeated N applications declined, as a proportion of unfertilised clover DM production, to about 0.75 or 0.6 with repeated applications of 100 or 200 kg N/ha applied over 4 defoliations, respectively. However, the decline in clover DM production in paspalum-white clover (Fig. 7) continued to show an increased loss (as a proportion of unfertilised clover DM production) as more N was applied. Mundy (1993) suggested that these different effects were due to differences in the growth habit of the two grass species. With ryegrass, low plant density and an erect growth habit helped reduce the competitive effects on clover, compared with paspalum where lateral spread crowded out the clover.

Fig. 7. The effect on clover DM production in paspalum-white clover pasture with 6 repeated applications of 100 (l) or 200 (n) kg N/ha applied over 4 defoliations. The effect on clover DM production for each data point on the figure was calculated as the cumulative clover growth after 4 defoliations of N fertilised pasture as a proportion of the cumulative clover growth of unfertilised pasture (Source: Mundy 1993).

By limiting the total amount of N applied to a pasture, the detrimental effects on clover are reduced, with evidence to suggest that between 100 and 200 kg N/ha can be applied over several months to pasture without having a major impact on clover persistence (Mundy 1996). The effect of 4 repeated applications of 0, 25 or 50 kg N/ha during autumn, spring or summer on total clover production (mean of the 3 seasons production) in ryegrass and paspalum pasture types is shown in Fig. 8. For the ryegrass pasture, up to 100 kg N/ha could be applied over a season before total clover production declined.

Fig. 8. The effect of N fertiliser applied 4 times at 25 or 50 kg N/ha on clover DM production in ryegrass-white clover (l) or paspalum-white clover (m). Clover DM production is the mean cumulative DM production of 3 seasons (vertical bars represent l.s.d. P=0.05) (Source: Mundy 1996).

Surprisingly, the clover in the paspalum pasture was not affected after 200 kg N/ha had been applied (Fig. 8) when the 3 seasons were averaged together. However, during the summer, when paspalum growth is high, 200 kg N/ha reduced clover DM production (Fig. 9). The different effect on clover growth between the two pasture types reflects the differences in clover content and in competitiveness of the grasses at different times of the year.

wpe4.jpg (9962 bytes)

Fig.9. The clover DM production of paspalum-white clover pasture with no N fertiliser (l) and 50 kg N/ha (m) applied 4 times during summer (vertical bar represents l.s.d. P=0.05) (Source: Mundy 1996).


7. FERTILISER TYPES

Urea is the cheapest source of N fertiliser and is the most commonly used form of N by dairy farmers on irrigated pastures. During the irrigation season, most farmers use flood irrigation to wash N fertiliser into the soil. This practice reduces the potential for ammonia volatilisation, but may lead to losses through denitrification of nitrate. Mundy and Mason (1989) and Mundy (1993) confirmed that urea was an effective form of N to increase pasture production during the irrigation season and that N losses from urea were relatively small if urea was watered in. In an earlier report, Roufail (1978) found that ammonium nitrate applications to irrigated pastures were effective at increasing pasture growth during the irrigation season. However, there is evidence that using flood irrigation to wash ammonium nitrate into the soil can lead to relatively large losses (up to 35%) of fertiliser (presumably through denitrification of nitrate), which can reduce its effectiveness for increasing pasture growth (Mason et al. 1987; Mundy and Mason 1989).

Experiments that have compared a number of different forms of N fertiliser, including foliar sprays with perennial pastures over the autumn to late winter period, have shown little difference in effectiveness at comparable rates of application (Kelliher 1962; Martin 1962; Mehanni 1974; Mehanni and West 1992; Mundy 1993). This is in general agreement with the findings of McKenzie and Jacobs (1997) for southern Victoria. The important aspect of using any form of N is to apply it to pasture when the soil and weather conditions favour its use.


8. SOIL ACIDITY

It is well known that N fertilisers acidify soils (Whitehead 1995). In northern Victoria, N fertilisers have usually had only minimal effects on soil pH in pastures (Roufail 1978; Mundy 1993). Roufail (1978) found soil pH was only lowered by 0.4 units after 784 kg N/ha/yr as ammonium nitrate had been applied for 4 years. Similarly, Mundy (1993) reported that 1200 kg N/ha applied over 2 years as urea to 2 pasture types had no significant effects on soil pH, although a downward trend in soil pH occurred at 1 study site. However, the application of the strongly acidifying ammonium sulphate at about 125 kg N/ha/yr reduced soil pH by 1.0 unit after 4 years of application to soils near Kerang, but only by 0.5 units on clay soils at Swan Hill or red brown earth soils in the Goulburn Valley (Twentyman 1970). This compound is used in some fertilisers and acidification could be a concern if it were to be used on pastures each year.


9. IMPLICATIONS FOR ANIMAL PRODUCTION

Response under mowing and grazing

As was the case for rainfed N research (Eckard 1998), the influence of nitrogen fertiliser on DM production of irrigated pastures has been largely determined with cutting experiments. In a few experiments, cows were used to defoliate pasture after plant production measurements had taken place (Roufail 1978; Mundy 1993). The idea was to better simulate a grazed pasture system that had nutrients returned by the animals.

Roufail (1978) conducted a 4 year experiment in which the growth response of irrigated pasture to N fertiliser was determined under mowing and grazing. The effect of N rate on annual DM production of grazed pasture was more consistent over the 4 years than with mown pasture. However, the effect of N rate on annual production was very similar for both defoliation methods in 2 of the 4 years. Mundy (1993) found that for 2 pasture types, N growth responses were linear under grazing and mowing. However, the responses were on average lower under grazing for the paspalum pasture, whereas with the ryegrass pasture they were similar. It appears that grazed pastures often respond to N fertiliser in a similar manner to mown pastures. Hence, it is likely that the results for pasture growth from relatively short term mowing experiments are reasonably applicable to grazed pastures. This is in agreement with Eckard (1998), who presented a similar view for rainfed pastures. Nevertheless, some important differences between mown and grazed pastures (eg. the amount of N cycling under grazing, the amount of damage by grazing, the defoliation intensity) should be considered when extrapolating pasture growth response under mowing to that under grazing.

 

Animal production

In 3 stocking rate experiments with dairy cows, there was no increase in total milk production from irrigated pasture when N fertiliser was used to boost pasture growth (Anon 1974a,b, King and Stockdale 1980). In the work of King and Stockdale (1980), pasture was fertilised 4 times per year with 56 kg N/ha per application with no effect on total milk or butterfat production per cow. However, the fertiliser consistently increased protein yield per cow, from 2 to 7 kg protein per cow as stocking rate increased from 4.4 to 8.6 cows/ha. Butterfat production was increased with N in early spring at the higher stocking rates. Eckard (1998) also cited results that suggested that animal response to N was better in early lactation.

In addition this research showed that as stocking rate increased from 4.4 to 8.6 cows/ha, the pasture production response declined from 17 to 3 kg DM/kg N (Stockdale and King 1980). This was associated with a decline in residual pasture mass from over 2 t DM/ha at 4.4 cows/ha to only about 1 t DM/ha at 8.6 cows/ha. This raises interesting questions: was the low response at the high stocking rates due to the low residual pasture mass; was the poor effect of N on animal production due to the low response at the high stocking rates?

 

Pasture nutritive characteristics

Stockdale and King (1980) found that pasture treated with N tended to have a lower digestibility than unfertilised pasture. This was probably due to the higher content of less digestible grass where N fertiliser was used. Some unpublished data (G.N. Mundy) clearly showed that N fertiliser had no appreciable effect on the digestibility of the grass or clover in ryegrass-white clover or paspalum-white clover pastures. Mundy (1993) also found that N had no effect on the digestibility of paspalum. These results are in general agreement with Whitehead (1995). N fertiliser can reduce pasture digestibility but principally when marked changes in botanical composition occur with a growth response to N fertiliser. However, at different times of the year (eg. in early to mid spring), the digestibility of ryegrass can be similar to that of white clover (G.N. Mundy unpublished data). At those times, growth responses to N fertiliser would not alter the pasture digestibility.

The N content of irrigated perennial pastures varies with the time of year and the dominant pasture species (Doyle et al. 1996). Generally, N fertiliser increases the N content of grass in mixed pastures and has no effect on clover N content (Mason et al. 1987; Mundy 1993). In late spring, N content of ryegrass tops can be increased from 3.3% with no N fertiliser, to 4.3% with 100 kg N/ha applied to perennial ryegrass-white clover pasture (Mundy 1993). These very high levels of plant N result from the uptake of fertiliser N over a relatively short duration and may increase the plant non-protein N content (including nitrate). Recent research suggests that the N content of perennial pastures is often excessive in relation to the requirements for efficient fermentation in the rumen and milk production (Stockdale et al. 1997). Therefore, higher N levels in these pastures may in fact have a cost in terms of milk production because of the need to expend of energy to excrete excess N from the body.

Low rates (less than 50 kg N/ha) of applied N may only marginally increase the N content of grass (Mason et al. 1887). However, in late winter/early spring, the N content of grass has been increased by up to 0.5% with 50 kg N/ha (Mundy 1993). This increase in N content of grass is usually offset by the lower contribution of clover N to the N content of the total pasture. Therefore, a single application of fertiliser to pasture with a reasonable clover percentage often has no effect on the N content of the mixed herbage.

Mundy (1993) reported that both frequency of defoliation and timing of a fertiliser application within a regrowth cycle may influence the N content of pasture grasses. Frequent defoliation (10 to 15 day interval) increased the N content of unfertilised paspalum herbage by up to 0.3% when compared with 20 to 30 days of regrowth of paspalum. However, defoliation of the paspalum pasture 10 to 15 days after 30 or 60 kg N/ha had been applied, increased its N content by up to 0.6 and 0.8%, respectively when compared with herbage of 20 to 30 days of regrowth after fertiliser application. The larger effect of N fertiliser on N content of relatively young herbage was most likely due to an insufficient amount of time for the N, that was assimilated in the grass from the fertiliser to be diluted by DM production that occurred with 20 to 30 days of regrowth. For a similar reason, N fertiliser applied midway through a regrowth cycle increased the N content of ryegrass and paspalum by about 0.2 % above that of grass fertilised immediately after defoliation.

The detrimental effect of repeated applications of N on clover content and production in irrigated pastures may result in a reduction in the N content of irrigated pastures, presumably because of the smaller contribution of clover N to total herbage N (G.N. Mundy unpublished data). For example, the N content of ryegrass-white clover pasture treated with 50 kg N/ha after each defoliation was found to be 0.3 to 0.7 % lower than the unfertilised pasture. In contrast, the herbage of a similarly fertilised paspalum dominant pasture had a lower N content than the ryegrass pasture, but its N content was usually unaffected by repeated N applications (G.N. Mundy unpublished data).


10. ENVIRONMENTAL IMPACT OF NITROGEN FERTILISERS

N from fertiliser may be lost to the environment from irrigated pastures through 4 processes: denitrification, ammonia volatilisation, runoff from irrigation bays, and leaching. These losses can have important environmental consequences and can have adverse economic implications when the amounts of N lost affect pasture growth.

 

Denitrification

Denitrification in soils occurs when soil oxygen concentrations become very low following heavy rainfall or flood irrigation. Under these conditions, microorganisms can denitrify soil nitrate into gaseous forms of N (nitrous oxide and N2) which can diffuse into the atmosphere. Substrate concentration, soil oxygen, soil temperature, pH, and available carbon all directly influence denitrification rate (Firestone 1982).

Mundy and Mason (1989) found that after application of ammonium nitrate to pasture, soil nitrate concentrations declined rapidly following a flood irrigation, possibly through denitrification. The apparent loss of up to 35% of applied N from ammonium nitrate, determined by 15N mass balance under waterlogged soil conditions (Mason et al. 1987; Mundy and Mason 1989), provides quantitative evidence of the likely potential for denitrification in flood irrigated pastures. Under similar waterlogged conditions, losses of N from urea were less than 14% (Mundy and Mason 1989). As urea does not contain nitrate, it is initially protected from loss by denitrification, and hence, it may be a more efficient form of N for application to flood irrigated pastures in spring, summer and autumn.

Clearly, the denitrification potential of irrigated pastures is high, but there are no data as to the rates of the process. These are needed for determining N fluxes in a dynamic N cycle system.

 

Ammonia volatilisation

Ammonia volatilisation from urea is potentially an important pathway for loss of N from irrigated pastures. Working with alkaline sewage wastewater applied to pasture Smith et al. (1996) found that up to 24% of ammoniacal-N was lost by volatilisation. Volatilisation rate was directly related to water evaporation from the soil, which is largely determined by wind speed and temperature.

In northern Victoria, several 15N balance experiments have shown that apparent losses of urea-N from pastures ranged from less than 10 to 32% of the N applied (Mundy 1993). These apparent losses could be due to volatilisation, but also to denitrification and leaching. Mundy (1995b) showed that initial soil water content and simulated rainfall (10 mm) had an effect on recovery of urea applied to pasture in late autumn. Recovery of applied N was greater than 90% when urea was applied to moist and wet soils, and 79% when applied to a relatively dry soil. The differences in losses were attributed to ammonia volatilisation. Further work is required to better quantify the effects of different soil water contents when N is applied to irrigated pastures as urea or di-ammonium phosphate (DAP). This information would assist in developing better fertiliser management practices.

The amount of N consumed by dairy cows that is excreted on pasture can be up to 80%, the majority as urine-N (Haynes and Williams 1993). Ammonia volatilisation from these urine patches in grazed pastures is an important loss that affects the N economy of pastures. The annual amount of N volatilised from grazed pastures has been estimated as about 15% of N excreted in urine for temperate areas (Whitehead 1995). However, the reported amounts of ammonia volatilised from N in urine patches varies from about 3 to nearly 40% (Whitehead 1995). The amount of ammonia lost from pastures grazed by cows where the pastures were treated with 0, 200 or 400 kg N/ha was 15, 45 and 63 kg N/ha/yr, respectively (Ledgard et al. 1996). The increase in ammonia loss with N rate was largely from the direct loss of N fertiliser after application (Ledgard et al. 1996). Mundy (1993) also found that N losses determined from 15N balances were higher with increased amounts of N applied to irrigated pastures. However, work is required to better quantify these gaseous N losses for grazed pastures under the conditions that occur in northern Victoria.

 

Surface runoff

When perennial pastures are flood irrigated, relatively large volumes of water may runoff the end of bays (up to 20% or more of applied water) (Nexhip and Austin 1998). This runoff water contains nutrients that may end up in inland waterways.

Following the application of N fertiliser, N concentrations in runoff after flood irrigation are elevated (Mundy 1993; Bush and Austin 1999; Nexhip et al. 1997; Nexhip and Austin 1998). These studies have found that, when urea was applied to pasture the day prior to an irrigation, between 5 and 14% of the applied N was lost in runoff water. The loss of N (kg N/ha) in runoff is proportional to the rate of application and the loss is mostly as urea (Bush and Austin 1999). Nexhip and Austin (1998) have shown that when applied urea is withheld from the bottom end (buffer strip) of irrigation bays, N concentrations in runoff declined exponentially with the length of the buffer strip. Thus, the loss of N declined from 13% of applied N with no buffer strip, to 2.5% with a 60 m buffer strip. Their data confirmed that most urea infiltrates with the wetting front during irrigation and that lateral and over-land flow of urea-N is relatively small.

 

Leaching

In northern Victoria, irrigated pastures occur mainly on duplex soil types. These soils have a low hydraulic conductivity, which should reduce the potential for leaching of nitrate-N. Studies with 15N labelled fertiliser have consistently shown that the amount of applied fertiliser recovered from the soil profile decreases with depth (eg. Mason et al. 1987; Mundy 1993), with usually up to 5% of the fertiliser being found at 30 to 40 cm . This suggests that N can move through these soils, albeit in small amounts. Leaching losses could be greater on the better draining soils that occur in some parts of the irrigated regions of southern Australia. Further studies that determine nitrification of applied urea and soil N may shed some light on why small amounts of N are present deep in the soil profile well after application.

Bush and Austin (1999) measured soil water concentrations of different forms of N in the soil profile after pasture was fertilised with urea and flood irrigated. The very elevated concentrations of urea and total N in soil water extracted to depths up to 60 cm in the profile clearly showed movement of N down through the soil with irrigation water. The mechanism of the N movement is not well understood and may be through cracks and macropores, as well as matrix flow. This work has also indicated that nitrate concentrations in soil water increased after follow up irrigations. It is not known whether the nitrate was leached from the upper soil depths or originated from urea that leached to down through the soil with the first irrigation. Hence, a better understanding of leaching processes (including water movements) is needed for flood irrigated pastures in northern Victoria.

Recent research has indicated that, in intensively grazed pasture systems, nitrate leaching is largely due to recycling of N in excreta from dairy cows (Ledgard et al. 1996, Hooda et al. 1998). However, Ledgard et al. (1996) also suggested that the apparent increased leaching losses with larger amounts of applied fertiliser N were partly due to direct leaching of applied fertiliser. Irrigated pastures in northern Victoria are grown on relatively poorly drained soils unlike the better internally drained soils of New Zealand. However, Ridley et al. (1999) have reported that high concentrations of nitrate-N occur at 100 cm depth in slowly draining red-brown earth soils, albeit under rainfed conditions. As both N fertiliser and recycling of N in dairy cow excreta (mainly urine) are involved in leaching losses from pastures, research is needed to determine their importance in irrigated pastures.


11. CONCLUSIONS AND DIRECTIONS FOR FUTURE RESEARCH

  1. The most important aspect of using N fertiliser on irrigated pastures is knowing how much extra feed will be grown and then utilising that feed for milk production.
  2. Pasture growth responses to increasing the amount of N fertiliser can vary in magnitude. This variation occurs between pasture types, and within and between seasons. However, the upper limit for a single application to grass-clover pasture should be 50 kg N/ha. Defoliation interval and timing of a N application within pasture regrowth cycles can affect the pasture response. Shorter defoliation intervals (rotations) are being adopted by farmers, and further work is needed to better understand the implications of this practice, especially with repeated N applications.
  3. Pastures are capable of producing a response of 10 kg DM/kg N at most times of the year, although there is considerable variability. However, pasture growth responses to N fertiliser are highest with late winter/early spring applications, and in summer with paspalum dominant pastures.
  4. Current knowledge enables farmers to make reasonable decisions on whether or not to apply N fertiliser. Future research needs to target understanding the factors that affect the size of the pasture growth response, and quantification of the losses of applied N with intensive grazing.
  5. Pasture growth responses to N fertiliser are influenced by the density of grass in the sward. There is evidence that perennial ryegrass density (plants and tillers) can decline with the age of irrigated pasture in northern Victoria. This decline may reduce the DM response to N fertiliser. The relationship between ryegrass density and DM response to N is needed for better prediction of growth responses to N fertiliser. This would be useful information for inclusion in a Decision Support System for N fertiliser with irrigated pastures.
  6. Farmers find that timing of grazing, spreading the fertiliser, and irrigating are all difficult to coordinate and manage, and often compromising best management practices. An assessment of alternative, effective options for applying N fertilisers to make life easier for farmers could be worthwhile.
  7. Environmental aspects of N fertiliser are important to the outcome of its use. There has been no direct measurement of the common gaseous losses from irrigated pastures. Leaching of N may be an issue for irrigated pastures with a better understanding of the process being required. This research should be conducted within a grazed pasture system.

12. REFERENCES

Anon (1974a) Stocking rate experiments. Field day notes March 1974. Irrigation Research Station, Kyabram. pp.3-5.

Anon (1974b) Nitrogen and stocking rate experiment. Field day notes March 1974. Irrigation Research Station, Kyabram. pp. 6-16.

Anon (1992). The Tongala groundwater pumping/reuse project. Volume 1 Monitoring and agronomic investigations. Final Report of project R147, Federal Water Resources Assistance Program.

Bartholomew, P.W., and Chestnut, D.M.B. (1977). The effect of a wide range of fertiliser nitrogen application rates and defoliation intervals on the dry-matter production, seasonal response to nitrogen, persistence and aspects of chemical composition of perennial ryegrass (Lolium perenne cv. S.24). Journal of Agricultural Science, Cambridge 88, 711-21.

Blaikie, S.J., Martin, F.M., Mason, W.K., and Connor, D.J. (1988). A basis for improved soil and water management for irrigated pastures in northern Victoria. Australian Journal of Experimental Agriculture 28, 315-19.

Bush, B.J., and Austin N.R. (1999). Environmentally sound on farm management of fertilisers for irrigated pasture. Final report for Murray Darling Basin Commission and Pivot Agriculture.

Chestnut, D.M.B., Murdoch, J.C., and Binnie, R.C. (1977). The effect of cutting frequency and applied nitrogen on production and digestibility of perennial ryegrass. Journal of the British Grassland Society 32, 177-83.

Crosse, S., and Dillon, P. (1996). Nitrogen use on dairy farms. In ‘Proceedings of the Large Herds Australia Conference’. Launceston, February. (Ed. M. Rawlings). pp. 158-72.

Dawson, K.P., and Ryden, J.C. (1985). Uptake of fertiliser and soil nitrogen by ryegrass swards during spring and mid-season. Fertiliser Research 6,189-92.

Dickens. C.R. (1994). Response of irrigated permanent pasture to nitrogen fertiliser application on laser-levelled soils in spring. Project report. Bachelor of Applied Agriculture. VACH Dookie.

Dowdell, R.J., and Webster, C.P. (1984). Effect of drought and irrigation on the fate of nitrogen applied to cut permanent grass swards in lysimeters: Experimental design and crop uptake. Journal of Science Food and Agriculture 35, 1092-104.

Doyle, P.T., Stockdale, C.R., and Lawson, A.R. (1996). Pastures for Dairy Production in Victoria. Agriculture Victoria, Kyabram Dairy Centre, Institute of Sustainable Agriculture.

Eckard, R.J. (1996). The window of opportunity for nitrogen fertiliser use on dairy farms. In ‘Proceedings of the Large Herds Australia Conference’. Launceston, February. (Ed. M. Rawlings). pp. 173-81.

Eckard, R.J. (1998). A critical review of research on the nitrogen nutrition of dairy pastures in Victoria. University of Melbourne and Department of Natural Resources and Environment.

Ellington, A. (1986). Nitrogen inputs and utilisation in leguminous pasture: A review of recent Australian literature. Technical Report No. 128. Department of Agriculture and Rural Affairs.

Feyter, C., O'Connor, M.B., and Addison, B. (1985). Effects of rates and times of nitrogen application on the production and composition of dairy pastures in Waikato district, New Zealand. New Zealand Journal of Experimental Agriculture 13, 247-52.

Firestone, M.K. (1982). Biological denitrification. In ‘Nitrogen in Agricultural soils’. Agronomy No. 22. (Ed. F.J. Stevenson). Published by American Society of Agronomy, Inc., Crop Science Society of America, Inc., Soil Science Society of America, Inc. Madison, Wisconsin USA. pp. 289-326.

Frame, J., and Boyd, A.G. (1987). The effect of strategic use of fertilizer nitrogen in spring and/or autumn on the productivity of a perennial ryegrass/white clover sward. Grass and Forage Science 42, 429-38.

Frame, J., Harkness, R.D., and Talbot, M. (1989). The effect of cutting frequency and fertiliser nitrogen rate on herbage productivity from perennial ryegrass. Research and Development in Agriculture 6, 99-105.

Garland, K.R., Jones, G.O., and Barkla, F.J. (1961). Fertilizer trials in the Loddon region. Irrigated annual pastures. Journal of Agriculture (Victoria) 59, 341-343.

Garland, K.R., Jones, G.O., and Barkla, F.J. (1962). Fertilizer trials at Kerang and Swan Hill. Irrigated perennial pastures. Journal of Agriculture (Victoria) 60, 44-46.

Haynes, J.R., and Williams P.H. (1993). Nutrient cycling and soil fertility in the grazed pasture ecosystem. Advances in agronomy 49, 119-199.

 

Holliday. R., and Wilman. D. (1965). The effect of fertiliser nitrogen and frequency of defoliation on yield of grassland herbage. Journal of British Grassland Society. 20, 32-40.

Hooda, P.S., Moynagh, M., Svoboda, I.F., and Anderson, H.A. (1998). A comparative study of nitrate leaching from intensively grazed monoculture grass and grass-clover pastures. Journal of Agricultural Science (Cambridge) 131, 267-75.

Kaddous, F.G. (1968). The effect of nitrogen, phosphorus and potassium fertilizers on the dry matter production of new irrigated perennial pasture at Kerang. Internal report. Department of Agriculture, Victoria.

Kelliher, P.D. (1962). Annual Report of Experimental Work. Irrigation Research Station Kyabram.

King, K.R., and Stockdale, C.R. (1980). The effects of stocking rate and nitrogen fertiliser on the productivity of irrigated perennial pasture grazed by dairy cows. 2. Animal production. Australian Journal of Experimental Agriculture and Animal Husbandry 20, 537-42.

Lawson, A., Kelly, K., Kelly, K., Byrne, E., and Phyland, G. (1998). Management techniques to improve the nutritive value and persistence of white clover and perennial ryegrass in irrigated pastures. Final Report to Dairy Research and Development Corporation, September 1998.

Ledgard, S.F. (1991). Transfer of fixed nitrogen from white clover to associated grasses in swards grazed by dairy cows, estimated by 15N methods. Plant and Soil. 131, 215-23.

Ledgard, S.F., Brier, G.J., and Upsdell, M.P. (1990). Effect of clover cultivar on production and nitrogen fixation in clover-ryegrass swards under dairy cow grazing. New Zealand Journal of Agricultural Research 33, 243-49.

Ledgard, S.F., Clark, D.A., Sporsen, M.S., Brier, G.J., and Nemaia, E.K.K. (1996). Nitrogen losses from grazed dairy pasture, as affected by nitrogen fertiliser application. Proceedings of the New Zealand Grasslands Association 57, 21-25.

Martin, F.M. (1962). Nitrogenous Fertilizer - cutting experiments in the Goulburn Region. Internal Report. Department of Agriculture.

Mason, W.K., Trigg, T.E., Mundy, G.N., Lemerle, C., Etheridge, M.O., and Jones, H. (1987). Production and Utilisation of Feed from Irrigation. Final Report to Australian Meat and Livestock Research and Development Corporation. DAV Project 36.

McGowan, A.A. (1987). Review of experiments with nitrogen fertiliser in Victoria. Research Report Series No. 54. Department of Rural Affairs and Agriculture.

McKenzie, F.R., and Jacobs, J. (1997). Strategic use of nitrogen on grazed pastures. In ‘Proceedings of Large Herds Australia Conference’ Warrnambool, April. pp. 98-111.

Mehanni, A.H. (1974). Foliar fertilisation on irrigated perennial pasture. Journal of the Australian Institute of Agricultural Science 40, 300-3.

Mehanni, A.H. (1987). Reclamation of saline/sodic soils by aquifer pumping, application of tillage and gypsum and reuse of saline groundwater. Australian Journal of Experimental Agriculture 27, 381-7.

Mehanni, A.H., and West, D.W. (1992). Effect of nitrogen fertiliser on yield of perennial pastures irrigated with saline water. Australian Journal of Experimental Agriculture 32, 1087-93.

Mitchell, K.J., and Lucanus, R. (1962). Growth of pasture species under controlled conditions. III. Growth at various levels of constant temperature with 8 and 16 hours of uniform light per day. New Zealand Journal of Agricultural Research 5, 135-44.

Mundy, G.N. (1991). Nitrogen fixation in pasture following a nitrogen fertiliser application. In ‘The Interaction of Pastures into Sustainable Pasture Systems’. Proceedings of the Pasture Specialist’s Conference Series, No. 42, p. 108.

Mundy, G.N. (1993). Guidelines for the economic use of nitrogen fertiliser on irrigated dairy farms. Final Report of DAV 140/241 to Dairy Research and Development Corporation.

Mundy, G.N. (1995a). Nitrogen fertiliser - Can we really go higher? In ‘Thinking About Tomorrow Today’. Target 10 Conference Proceedings (Shepparton) pp. 72-9.

Mundy, G.N. (1995b). Effect of initial soil water content and application of water on urea applied to pasture. Australian Journal of Agricultural Research 46, 821-30.

Mundy, G.N. (1996). Nitrogen fertiliser – What are the upper limits? In ‘Proceedings of the Large Herds Australia Conference’. Launceston, February. (Ed. M. Rawlings). pp. 182-93.

Mundy, G.N., and Mason, W.K. (1989). The effect of flood irrigation and nitrogen source on the fate of nitrogen fertiliser applied to pasture. Australian Journal of Agricultural Research 40, 107-16.

Mundy, G.N., Jones, H.R., and Mason, W.K. (1988). Nitrogen fixation activity by white clover pastures during flood irrigation cycles. Australian Journal of Agricultural Research 39, 409-14.

Norman, I.W. (1962). Fertiliser trials on irrigated pastures. Macalister irrigation district. Journal of Agriculture (Victoria) 60, 403-407.

Nexhip, K.J., and Austin N.R. (1998). Buffer zones as nutrient traps on irrigated pastures. Final Report for the Victorian State Government's Agriculture and Food Initiative & the Dairy Research and Development Corporation.

Nexhip, K.J., Mundy, G.N., Collins, M.D., and Austin, N.R. (1997). Development of nutrient water quality targets for irrigated pasture sub-catchments. Final Report to Murray Darling Basin Commission's Natural Resource Management Strategy and Dairy Research and Development Corporation.

Ridley, A.M., Simpson, R.J., and White, R.E. (1999). Nitrate leaching under phalaris, cocksfoot and annual ryegrass pastures and implications for soil acidification. Australian Journal of Agricultural Research 50, 55-63.

Riffkin, P.A., Quigley, P.E., Cameron, F.J., Peoples, M.B., and Thies, J.E. (1999). Annual nitrogen fixation in grazed dairy pastures in south-western Victoria. Australian Journal of Agricultural Research 50, 273-81.

Roufail, A. (1978). The effect of applied N and P under mowing and grazing on the yield and botanical and chemical composition of irrigated pasture and on the nutrient status of a red-brown earth at Kyabram. Master of Agricultural Science Thesis. University of Melbourne.

Simpson, J.R. (1962). Mineral nitrogen transformations in soils under improved pasture in southern New South Wales. Australian Journal of Agricultural Research 13, 1059-72.

Simpson, J.R. (1985). Nitrogen nutrition of pastures. Review 6. Australian Wool Corporation/CSIRO Symposium, Leura, NSW.

Simpson, J.R. (1987). Nitrogen nutrition of pastures. In ‘Temperate Pastures Their Production, Use and Management’. J.L. Wheeler, C.J. Pearson, G.E. Robards (Eds). CSIRO Australia. pp. 143-152.

Smith, C.J., Chalk, P.M., Noble, C.L., Prendegast, J.B., and Robertson, F. (1993). Nitrogen fixation in a white clover-grass pasture irrigated with saline ground water. Irrigation Science 13, 189-94.

Smith, C.J., Freney, J.R., and Bond, W.J. (1996). Ammonia volatilisation from soil irrigated with urban sewage effluent. Australian Journal of Soil Research 34, 789-802.

Stockdale, C.R. (1983). Irrigated pasture productivity and its variability in the Shepparton Region of northern Victoria. Australian Journal of Experimental Agriculture and Animal Husbandry 23, 131-39.

Stockdale, C.R. (1999). Effects of season and time since defoliation on the nutritive characteristics of three irrigated perennial pasture species in northern Victoria. Australian Journal of Experimental Agriculture 39, (In press).

Stockdale, C.R., and King, K.R. (1980). The effects of stocking rate and nitrogen fertiliser on the productivity of irrigated perennial pasture grazed by dairy cows. I. Pasture production, utilisation and composition. Australian Journal of Experimental Agriculture and Animal Husbandry 20, 529-536.

Stockdale, C.R., Dellow, D.W., Grainger, C., Dalley, D., and Moate, P.J. (1997). Supplements for Dairy Production in Victoria. Department of Natural Resources and Environment published by Dairy Research and Development Corporation.

Twentyman, J.D. (1970). Fertilizers. In ‘Irrigated Pastures’. Victorian Department of Agriculture. pp 141-56.

Whitehead. D.C. (1970). The role of nitrogen in grassland productivity. A review of information from temperate regions. Bulletin 48, Commonwealth Bureau of Pastures and Field Crops Hurley, Berkshire, England.

Whitehead., D.C. (1995). Grassland Nitrogen. CAB International. Wallingford, UK.

Go To Home Page

Last modified:30 October, 2007                         Please Note: Disclaimer             Authorised and maintained by:Richard Eckard
Copyright 1997-1999, The University of Melbourne.