Citation
Nutrient Dynamics in Bahiagrass Swards Impacted by Cattle Excreta

Material Information

Title:
Nutrient Dynamics in Bahiagrass Swards Impacted by Cattle Excreta
Creator:
White, Una Renee
Place of Publication:
[Gainesville, Fla.]
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (139 p.)

Thesis/Dissertation Information

Degree:
Master's ( M.S.)
Degree Grantor:
University of Florida
Degree Disciplines:
Agronomy
Committee Chair:
Sollenberger, Lynn E.
Committee Members:
Newman, Yoana C.
Vendramini, Joao M.B.
Graetz, Donald A.
Graduation Date:
5/1/2008

Subjects

Subjects / Keywords:
Cattle ( jstor )
Excreta ( jstor )
Feces ( jstor )
Forage ( jstor )
Grasslands ( jstor )
Nitrogen ( jstor )
Nutrients ( jstor )
Pastures ( jstor )
Soils ( jstor )
Urine ( jstor )
Agronomy -- Dissertations, Academic -- UF
City of Pensacola ( local )
Genre:
Electronic Thesis or Dissertation
bibliography ( marcgt )
theses ( marcgt )
Agronomy thesis, M.S.

Notes

Abstract:
Most nutrients consumed by grazing livestock are returned to pasture in excreta, but excreta effects on forage responses, plant nutrient recovery, and shallow soil water quality are not well defined. The objective was to determine the effect of management intensity, excreta type, and number of excreta applications on bahiagrass (Paspalum notatum Flugge) herbage dry matter (DM) harvested and nutritive value, excreta nutrient recovery in harvested herbage, and nutrient concentration in shallow soil water. Treatments were the factorial combinations of two management intensities (Average and High, 60 and 120 kg N ha-1, respectively), two excreta types (dung and urine), and three application frequencies (1, 2, or 3 season-1). Three control treatments received no excreta and 0, 60, or 120 kg N ha-1. Dung and urine were collected from animals grazing bahiagrass pasture and applied to ungrazed bahiagrass plots. A urine application was 2 L distributed to a 60-cm diameter circle, and a dung application was 2 kg fresh weight applied to a 30-cm diameter circle. Concentric rings (radii of 0-15, 15-30, 30-45, and for urine plots, 45-60 cm from the center of excreta application) were clipped monthly from July through October 2006 to measure forage responses, and lysimeters were placed 1.3 m below soil level directly under the excreta application to measure nutrient concentrations in soil water. Herbage data are reported for a circle of radius 45 cm from the center of the excreta deposit and also by ring to show spatial patterns of response. Herbage response for the circle of 45-cm radius was affected to the greatest extent by application frequency X excreta type interaction. Interaction occurred because dung application had no effect on most herbage responses, whereas responses to urine were consistently significant. In urine-treated plots, herbage DM harvested increased from 2760 at zero applications to 4670 kg ha-1 with three, while over the same application frequency herbage N concentration increased from 13.3 to 16.2 g kg-1, P concentration from 3.45 to 3.87 g P kg-1, in vitro digestible organic matter concentration from 569 to 588 g kg-1, N harvested from 37 to 75 kg ha-1, and P harvested from 9 to 18 kg ha-1. Excreta N recovery was greater from urine than dung and decreased as application frequency increased from one to three (28 to 18% for urine and 4 to < 1% for dung). Spatial characteristics of response were assessed within excreta type and were consistently affected by application frequency X ring number interaction. In urine-treated plots, herbage response generally was greatest near the center of the urine deposit and decreased as distance from center of excreta application increased. In dung-treated plots, physical interference by dung resulted in decreased herbage DM, N, and P harvested in the area under the dung deposit. Urine affected DM harvested and herbage P concentration up to 15 cm and N concentration and N harvested up to 30 cm beyond the edge of the urine application. Dung had no effect on any response outside the area physically impacted by dung. Shallow soil water was affected by treatment at only one sampling date, and NO3-N concentration was > 10 mg L-1 in only three samples during the entire study. In conclusion, urine has much greater impact on herbage response than dung, both in the area impacted physically by the excreta and beyond that area. Greater N recovery following single vs. multiple excreta events per site per year emphasizes the importance to sustainable grassland management of grazing practices that increase uniformity of excreta deposition. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis:
Thesis (M.S.)--University of Florida, 2008.
Local:
Adviser: Sollenberger, Lynn E.
Statement of Responsibility:
by Una Renee White

Record Information

Source Institution:
UFRGP
Rights Management:
Copyright by Una Renee White. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
7/11/2008
Classification:
LD1780 2008 ( lcc )

Downloads

This item has the following downloads:


Full Text







6500


6000

-5500

5000

m 4500

S4000

1 3500

o 3000

2500

2000


Ring


Figure 4-2. Ring number by urine application frequency interaction (P < 0.001) on herbage DM
harvested during 2006. Ring number effects were linear for three (P < 0.001) and
two applications (P < 0.001), linear (P < 0.001) and quadratic (P = 0.058) for one
application, and not significant (P > 0.352) for zero applications. Standard error of a
treatment mean was 76 kg ha1. Ring Numbers 1 through 4 refer to sampling areas 0
to 15, 15 to 30, 30 to 45, and 45 to 60 cm from the center of the urine application,
respectively.


-4-Urine-3
-40-Urine-2
-& Urine-1
--&Urine-0









CHAPTER 4
SPATIAL PATTERNS OF BAHIAGRASS HERBAGE ACCUMULATION AND NUTRIENT
CONCENTRATION RESPONSES TO TYPE AND FREQUENCY OF EXCRETA
DEPOSITION

Introduction

Planted grasslands and non-forested rangeland comprise nearly 30% of the USA land area

(Barnes and Nelson, 2003) and occupy more than 4 million ha in Florida (Dubeux et al., 2007).

Most grasslands in Florida are managed extensively with limited fertilizer input, but because

there is such a large area covered by grasslands, the fate of nutrients can have a major impact on

ecosystem function (Nair et al., 2007).

When livestock graze grassland, a large proportion of the nutrients consumed in forage are

returned to the sward in excreta (Sollenberger et al., 2002). A single urination from mature cattle

may provide the equivalent of 5 mm of rain and 400 to 500 kg N ha-1, while dung may supply the

equivalent of 110 kg P and 220 kg of K ha- along with other nutrients (Haynes and Williams,

1993). Nutrients recycling through animal excreta have long been considered beneficial to the

fertility of grazed pastures (Ball et al., 1979). For example, urine patches in a mixed grassland

-2 -
contained 112 g m-2 more above-ground biomass and 2.5 g m-2 more plant N than unaffected

areas (Day and Detling, 1990). In a Colorado study, urine patches affected only 2% of the

pasture surface, but they contributed 7 to 14% of consumed forage (Day and Detling, 1990).

The areas covered by a single dung or urine application by cattle have been estimated at ~

0.1 and 0.4-m2, respectively (Haynes and Williams, 1993). To effectively define the impact of

dung and urine application on grasslands, the spatial pattern of plant responses around an excreta

deposit needs to be described. Lotero et al. (1966) observed cattle grazing tall fescue (Festuca

arundinacea Schreb.) and reported that urine affected plant response in an area of- 1.02 m2

With dairy cows, Lantinga et al. (1987) reported that urine affected plant growth in an area of










There are limited data describing digestibility of herbage in studies of excreta deposition.

Greenhalgh and Reid (1968) stated in a general way that digestibility of perennial ryegrass

herbage near a dung deposit may be greater than similar herbage growing in mechanically

harvested fields.


595

590

585

580

o 575

570

O 565

560

555 -Urine
S Dung
550

545
0 1 2 3
Application Frequency (Number yr1)


Figure 3-5. Excreta type X excreta application frequency interaction (P = 0.099) for bahiagrass
herbage in vitro digestible organic matter (IVDOM) concentration during the 2006
growing season. There was no effect of dung application frequency on IVDOM (P >
0.300), but there was a linear (P = 0.004) effect of urine application frequency on the
response. Standard error of a treatment mean was 3.1 g kg-1. Herbage IVDOM
concentration is different for Frequency 3 (P = 0.006) but not different for
Frequencies 2 (P = 0.132) and 1 (P = 0.445).

Total Nitrogen Harvested

Total N harvested was affected by management intensity (P < 0.001), excreta application

frequency (P = 0.003), excreta type (P < 0.001), and application frequency X excreta type

interaction (P < 0.001) (Table A-i). High management intensity plots had greater N harvested









Excreta Source Pastures

There were two replicates of each source treatment (Average and High management

intensities) arranged in a completely randomized design. Pastures were stocked continuously and

pasture size was 1 ha for Average and 0.5 ha for High. Each pasture was grazed with two

crossbred yearling heifers (Angus x Brahman) with average initial weight of 408 kg. Grazing

was initiated on 26 May 2006. Details of animal management, pasture fertilization, and pasture

sampling were provided in Chapter 3.

Excreta Collection and Analyses

Animal selection and training were described in Chapter 3. During the trial, animals were

brought to stanchions for collection of urine and dung. Urine and dung were collected during

three periods (12-16 June; 26-28 July; and 7-8 Sept. 2006) leading up to the three dates (20 June,

1 Aug., and 13 Sept. 2006) when excreta was applied to the plots. Excreta application dates were

at 6-wk intervals.

The process used for excreta collection is described in Chapter 3. After urine and dung

were collected separately, they were immediately put into sealed containers and taken to a

refrigerator for storage at 40C. Prior to application, all excreta events of a given type (dung or

urine) from one intensity treatment (four animals) were composite across field replicates. Thus,

at time of application there were two types of dung and two types of urine (one of each excreta

type from both Average and High treatments). Three subsamples were taken from the

composite samples of urine and dung and analyzed to determine excreta chemical composition

so that amount of nutrients applied to plots could be calculated.

During the June excreta collection, samples were taken to ensure that chemical

composition did not change markedly during the 8 d of excreta storage prior to application, i.e.,

to ascertain that what was applied to the plots was similar in composition to what the animal









application frequency on N harvested (P > 0.404), but there was a linear (P < 0.001)
effect of urine application frequency on the response. Standard error of a treatment
mean was 1.5 kg ha-1. Herbage N harvested was greater for urine- than dung- treated
plots for Frequencies 1 through 3 (P < 0.001). ........................................ ............... 75

3-7 Excreta type X excreta application frequency interaction (P < 0.001) for P harvested
in bahiagrass herbage during the 2006 growing season. There was no effect of dung
application frequency on P harvested (P > 0.570), but there were linear (P < 0.001)
and quadratic effects (P = 0.052) of urine application frequency on the response.
Standard error of a treatment mean was 0.35 kg ha-1. Herbage P harvested was
greater for urine- than dung-treated plots for Application Frequencies 2 (P = 0.007)
and 3 (P < 0.001) and tended to be greater for Frequency 1 (P = 0.115). .........................77

3-8 Excreta type main effect (P < 0.001) for excreta N recovery in harvested bahiagrass
herbage during the 2006 growing season. There was no excreta type X excreta
application frequency interaction (P = 0.543), but there was a linear effect (P =
0.064) of excreta application frequency on N recovery. Standard error of a treatment
m ean w as 1.8% ......................................................... ................. 80

3-9 Excreta type X application frequency interaction for N03-N concentration in shallow
soil water on 15 Sept. 2006. There was a quadratic effect of urine application
frequency on N03-N concentration (P = 0.057), and there was a linear effect (P =
0.0394) of dung application frequency on the response. There was no effect of
excreta type at any level of application frequency (P > 0.181). Standard error of a
treatm ent m ean w as 0.03 m g L1 ........... .................. ......... ...............................83

4-1 Diagram of harvested rings to quantify spatial pattern of response to dung and urine
application (R1, circle of 15-cm radius; R2, 15- to 30-cm radius; R3, 30- to 45-cm
radius; R4, 45- to 60-cm radius). ............................................. .............................. 93

4-2 Ring number by urine application frequency interaction (P < 0.001) on herbage DM
harvested during 2006. Ring number effects were linear for three (P < 0.001) and
two applications (P < 0.001), linear (P < 0.001) and quadratic (P = 0.058) for one
application, and not significant (P > 0.352) for zero applications. Standard error of a
treatment mean was 76 kg ha-1. Ring Numbers 1 through 4 refer to sampling areas 0
to 15, 15 to 30, 30 to 45, and 45 to 60 cm from the center of the urine application,
resp ectiv ely ............................................................ ................ 9 8

4-3 Ring number by dung application frequency interaction (P < 0.027) on herbage DM
harvested during 2006. Ring number effects were quadratic for three (P = 0.025)
applications, linear (P < 0.001) and quadratic (P = 0.008) for two applications, and
not significant for one (P > 0.370) or for zero applications (P > 0.100). Standard
error of a treatment mean was 106 kg ha-1. Ring Numbers 1 through 3 refer to
sampling areas 0 to 15, 15 to 30, and 30 to 45 cm from the center of the dung
application, respectively. ..................... .................... ..................... ........ 99









CHAPTER 3
BAHIAGRASS HERBAGE DRY MATTER HARVESTED, NUTRIENT CONCENTRATION,
AND NITROGEN RECOVERY FOLLOWING EXCRETA DEPOSITION

Introduction

Grazed bahiagrass (Paspalum notatum Flugge) pastures comprise approximately 1.1

million hectares and 75% of the planted perennial pasture resource in Florida (Gates et al., 2004;

Mislevy et al., 2005). Bahiagrass is adapted to low soil nutrient levels, and limited amounts of

fertilizer are applied to grazed swards (Chambliss and Adjei, 2006). Despite this, these areas are

heavily scrutinized for their potential environmental impact. This is due to the large land area

planted to this forage and to the sensitivity of Florida ecosystems, especially in terms of nutrient

impacts on water quality (Nair et al., 2007). On agricultural lands, the lateral flow of P to surface

water, particularly in Central and South Florida (Nair et al., 2007), and leaching of nitrates to

ground water, more commonly in North Florida, are critical concerns (Woodard et al., 2002).

In low-input bahiagrass pastures in Florida, animal excreta and decaying plant matter

(litter) are the major transitory nutrient pools, and the importance of excreta vs. plant litter

increases as stocking rate increases (Thomas, 1992). Chemical transformations that occur during

digestion and excretion of plant nutrients by herbivores cause nutrients in animal excreta to be

more available for subsequent uptake by plants and more susceptible to loss to the environment

than those in plant litter (Bardgett and Wardle, 2003; Boddey et al., 2004). Efficient recovery of

these nutrients is hindered by the large quantity supplied in a single excreta event and because a

disproportionately large number of these events occur in small areas where cattle congregate,

e.g., near to shade, water, and supplemental feed sources (Mathews et al., 2004; Sollenberger et

al., 2002).

It has been suggested that rotational stocking with short grazing periods, i.e., many

paddocks per pasture, decreases the opportunity and tendency of animals to congregate in



















o) 35



S30
(0- Dung-0
--Dung-1
z / Dung-2
25 -Dung-3
25



20 ------------------------
20
1 2 3
Ring


Figure 4-7. Ring number X dung application frequency interaction (P = 0.019) for bahiagrass
herbage N harvested during 2006. The ring number effect was quadratic (P = 0.048)
for three applications and linear (P < 0.001) and quadratic (P = 0.007) for two dung
applications, but was not significant (P > 0.288) for one and no dung applications (P
> 0.167). Standard error of a treatment mean was 1.5 kg ha-1. Ring Numbers 1
through 3 refer to sampling areas 0 to 15, 15 to 30, and 30 to 45 cm from the center of
the dung application, respectively.

Herbage Phosphorus Harvested

For both dung- and urine-treated plots there was a ring number main effect (P < 0.003) on

herbage P harvested and an application frequency X ring number interaction (P = 0.0584 for

dung and P < 0.001 for urine; Tables A-2 and A-3).

The application X ring number interaction in urine-treated plots occurred because P

harvested decreased with increasing ring number for Application Frequencies 1 through 3, but

there was no effect of ring number for the zero application frequency (Fig. 4-8). This pattern of

response is nearly identical to DM harvested because there was no effect of ring number on









water quality may result in greater use of this practice in the future. Implementation of rotational

stocking may make important contributions to the overall economic and environmental

sustainability of bahiagrass-livestock systems in Florida.









NRC. 1996. Nutrient requirements of beef cattle; 6th ed. National Academy Press; Washington,
DC.

Nair, V.D., P.K.R. Nair, R.S. Kalmbacher, I.V. Ezenwa. 2007. Reducing nutrient loss
from farms through silvopastoral practices in coarse-textured soils of Florida,
USA. Ecological Engineering. 29:192-199.

National Agricultural Statistics Service. 2006.
http://www.nass.usda.gov:8080/QuickStats/PullDataUS. April 2006.

Newman, Y.C., M.B. Adjei, J.M. Scholberg, C.G. Chambliss, and J. Rechcigl. 2005.
Nitrogen rate effects on productivity and nutritive value of forages used on South
Florida dairies. In K. Cassida et al. (ed.), Proc. Am. Forage and Grassl. Counc.
Vol. 14. Springfield, IL.

Newman, Yoana C., Lynn E. Sollenberger, Kenneth J. Boote, L. Hartwell Allen, Jr., Jean
M. Thomas, and Ramon C. Little. 2006. Nitrogen fertilization affects bahiagrass
responses to elevated atmospheric carbon dioxide. Agron. J. 98:382-387.

Norman, M.J.T., and J.O. Green. 1958. The local influence of cattle dung and urine upon
the yield and botanical composition of permanent pasture. Journal of the British
Grassland Society. 13:39-45.

Oenema, O., G.L. Velthof, S. Yamulki, S.C. Jarvis. 1997. Nitrous oxide emissions from
grazed grassland. Soil Use and Management. 13:288-295.

Overman, A.R. and R.L. Stanley. 1998. Bahiagrass response to applied nitrogen and
harvest interval. Commun. Soil Sci. Plant Anal. 29:237-244.

Owens, L.B., W.M. Edwards, and R.W. Van Keuren. 1992. Nitrate levels in shallow
groundwater under pastures receiving ammonium nitrate or slow release nitrogen
fertilizer. J. Environ. Qual. 21:607-613.

Pakrou, N. and Peter Dillon. 1995. Preferential flow, nitrogen transformations and N15
balance under urine-affected areas of irrigated and non-irrigated clover based
pastures. Journal of Contaminant Hydrology. 20:329-347.

Pedreira, C.G.S. and R.H. Brown. 1996a. Physiology, morphology, and growth of
individual plants of selected and unselected bahiagrass populations. Crop
Sci. 36:138-142.

Pedreira, C.G.S. and R.H. Brown. 1996b. Yield of selected and unselected Bahiagrass
populations at two cutting heights. Crop Sci. 36:134-137.









APPENDIX
STATISTICAL TABLES

Table A-1. P values for interactions and main effects on response variables discussed in Chapter 3.

Response Source of Variation
variable Management Application Excreta MI x AF MI x E AF x E MIx AF xE
intensity (MI) frequency (AF) type (ET)
DM Harvested 0.0004 0.2224 <0.0001 0.7093 0.2288 0.0633 0.2996

Herbage N 0.0385 <0.0001 <0.0001 0.5707 0.7387 0.0001 0.9200

Herbage P 0.9235 0.0033 0.8355 0.7151 0.1743 0.0019 0.1776

Herbage IVDOM 0.0097 0.0045 0.0014 0.1892 0.9280 0.0987 0.1210

N Harvested <0.0001 0.0025 <0.0001 0.6270 0.4111 0.0008 0.3279

P Harvested 0.0006 0.0001 <0.0001 0.6256 0.0327 <0.0001 0.6859

N Recovery 0.5657 0.1249 <0.0001 0.5204 0.4631 0.5453 0.3771









Nitrogen Concentration

For dung-treated plots, there was no effect of ring (P > 0.160) on herbage N concentration

nor were there interactions of other treatments with ring (P > 0.686) (Table A-2). Means for

Rings 1 through 3 were 13.5, 13.3, and 13.6 g kg-1, respectively. There has been limited research

investigating the spatial impact of dung on herbage N concentration. Jorgensen and Jensen

(1997) applied 2.1 kg of sheep feces, the equivalent of 960 kg N ha-1, in July to a 25-cm diameter

circle in a mixture of perennial ryegrass cv. Sisu and white clover cv. Milkanova. The herbage

was harvested in concentric circles around the application area such that the first extended 0 to

15 cm beyond the edge of that area and the second extended 15 to 30 cm beyond the edge. In the

0- to 15-cm zone, neither grass nor clover herbage N concentrations were affected by feces in

October or the following June. By August, 13 mo after feces application, grass N concentration

was greater in the 0 to 15 cm zone than for control plots (25.2 vs. 20.5 g kg-1). These authors

noted that less than 2 g kg-1 of dung N was recovered in harvested herbage by October following

July application, and after 13 mo only 35 g kg-1 of dung N was recovered in harvested herbage.

The application frequency x ring number interaction (P < 0.001) on urine-treated plots

occurred because there were linear effects of ring for two and three urine applications, but there

was no ring effect for the control (means of 13.3, 13.1, 13.5, and 13.3 g kg-1 for R1 through R4,

respectively) or for one urine application (14.1, 13.6, 13.6, and 13.6 g kg-1 for R1 through R4,

respectively) (Fig. 4-4). Moving from the center of the urine deposit outward, N concentration

decreased linearly (P < 0.001) for two (17.8, 16.8, 15.3, and 14.4 g kg-1) and for three urine

applications (17.8, 16.4, 15.6, and 14.0 g kg-1, respectively).

Lotero et al. (1966) evaluated urine impacts on tall fescue pastures. They measured

herbage N concentration at increasing distances from a urine deposit. They found a marked

decrease in N concentration moving away from the point of impact, but it was apparent only at









N and P concentration using a micro-Kjeldahl technique (Gallaher et al., 1975) and for in vitro

digestibility using a two-stage procedure (Moore and Mott, 1974).

Lysimeter Placement, Water Sampling, and Water Analyses

One lysimeter, designed as described by Woodard et al. (2002), was positioned in each plot

such that the ceramic cup was directly below the center of each plot and 1.3 m below soil level.

The PVC pipe entered the soil at approximately 40 cm from the center of the excreta deposition

site (beyond the area to which either dung or urine were applied) and angled through the soil so

that the ceramic cup was properly placed beneath the center of the plot and the excreta

application site. This approach allowed the sod immediately around the excreta deposit to be

undisturbed. This depth for placement of the lysimeters cup was chosen because it is below

nearly all grass roots.

The intent was that lysimeters be sampled every 14 d or after major rainfall events (> 25

mm). During dry periods when 14-d sampling was not possible because of inadequate soil water,

re-initiation of sampling was triggered by major rainfall events (> 25 mm) that resulted in

sufficient soil water so that the lysimeters could hold a vacuum. Two days prior to sampling, any

water contained in the lysimeter was evacuated, then a suction (40 to 45 kPa) was placed on it.

At sampling, suctions were released. Samples were acidified (pH of <2) and placed in a cooler

within 15 min of extraction. Water sampling between 20 June 2006 and 2007 occurred on 23 and

28 June, 10 and 24 July, 8, 22, and 29 Aug., 15 and 27 Sept., 12 and 19 Oct., 19 Nov., and 29

Dec. 2006, and 5 and 26 Jan., 5 and 19 Feb., and 23 Apr. 2007. The water was analyzed for NO3-

N and total P. Occasionally throughout the growing season, samples were analyzed for NH4-N to

insure that there was no water moving down along-side the external wall of the lysimeters and

transferring fertilizer N directly to the collection cup. At no time were there measurable

concentrations of P or NH4-N, so these data are not presented.









fertilization increased from 0 to 224 kg ha-1. Herbage accumulation was not different when N

fertilization increased from 224 to 672 kg ha-1.

Mislevy et al. (2005) found that Pensacola produced 50 to 60% of its total seasonal yield

during long days of June, July, and August, and average annual yield was between 10 000 and

12 000 kg ha-1 when 112 kg N ha-1 yr- was applied. Total seasonal yield was 13 400 kg ha-1 in

the first year of the 3-yr study, while yields in the second and third years were 11 000 and

6300 kg ha-1. The reason for this drastic decrease in Year 3 was drought. Total rainfall for the 1st,

2nd, and 3rd years was 1740, 1270, and 813 mm, respectively.

Burton et al. (1997) fertilized Pensacola bahiagrass at differing rates of N, P, and K, the

lowest being 56 kg N, 24 kg P, and 46 kg K ha-l; the highest was 448 kg N, 49 kg P, and 278 kg

K ha-l. Average dry matter harvested over the 3-yr period was 10 620 kg ha-l, with the highest

yield of 15 070 kg ha-1 from the greatest fertilizer amount. In Louisiana, Twidwell et al. (1998)

found that a single application of 224 kg N ha-1 to bahigrass increased annual yield from 4040 kg

ha-l for the zero control to 11 900 kg ha-1.

Bahiagrass response to P fertilization has been less consistent than to N, due in part to

bahiagrass' ability to access soil P from below the depth sampled for soil analysis (Mylavarapu

et al., 2007), especially when growing in Spodosols. In a study conducted in South Central

Florida, bahiagrass herbage yield increased 12% above the zero control when fertilized with 15

kg P ha-l (Ibrikci et al., 1999). Burton et al. (1997) found bahiagrass P rates had no effect on DM

yield across a range of N and P rates. In that study, bahiagrass was fertilized with varying rates

of N, P, and K (from 56 kg N, 24 kg P, and 46 kg K ha-1 as the lowest fertilizer rate, to the

highest of 448 kg N, 49 kg P, 279 kg K ha-1). McCaleb et al. (1966) found that Pensacola

bahiagrass yield was affected by P fertilization only for an initial increment of 6 kg P ha-1 yr-1.









summer application increased yield only 10 to 15%, probably due to greater volatilization losses

in summer. Total N uptake for the spring and summer applications was greater than the fall.

Average urinary N uptake in the harvested herbage over the 2-yr period was approximately 40%;

for the spring harvest urinary N uptake ranged from 52 to 60% across soil types. The fall uptake

of urinary N was poor compared to the spring and ranged from 16 to 41%.

Silva et al. (2005) applied dairy cow urine to perennial ryegrass-white clover plots. Plots

received urine+urea (1000 + 4000 kg N ha-1 yr1), urine alone (1000 kg N ha-1 yr1), or a control

with no N applied. Dry matter production over a 2-yr period was 10 700 kg ha-1 yr1 (control),

19 415 kg ha-1 yr- (urine), and 20 000 kg ha-1 yr1 (urine+urea). Uptake of N by the control

treatment was approximately half that (364 kg ha-1 yr1) of the urine (610 kg ha-1 yr1) and

urine+urea (705 kg ha-1 yr1) treatments.

In another study, perennial ryegrass cv. Concord and white clover cv. Grasslands Huia

were used in short-term pasture rotations (Williams and Haynes, 2000). The pasture had been

planted previously to cereal and grain crops for 20 yr and then planted to grass-clover 2 yr

before start of the experiment. Nitrogen concentration in herbage was measured 14, 33, 77, and

133 d after application of sheep urine. Resulting concentrations were 46, 31, 29, and 22 g N kg-',

respectively. In the long-term pasture, a pasture that was sown -25 yr prior to the experiment and

rotationally stocked by sheep, N concentration in herbage decreased slightly compared to the

short-term pasture and was 33, 28, 24, and 24 g kg-1, respectively, for the four time periods after

urine application. Clearly the impact of urine is greatest soon after application and diminishes

with time.









Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

NUTRIENT DYNAMICS IN BAHIAGRASS SWARDS IMPACTED BY CATTLE EXCRETA


By

Una Renee White

May 2008

Chair: Lynn E. Sollenberger
Major: Agronomy

Most nutrients consumed by grazing livestock are returned to pasture in excreta, but

excreta effects on forage responses, plant nutrient recovery, and shallow soil water quality are

not well defined. The objective was to determine the effect of management intensity, excreta

type, and number of excreta applications on bahiagrass (Paspalum notatum Flugge) herbage dry

matter (DM) harvested and nutritive value, excreta nutrient recovery in harvested herbage, and

nutrient concentration in shallow soil water. Treatments were the factorial combinations of two

management intensities (Average and High, 60 and 120 kg N ha-1, respectively), two excreta

types (dung and urine), and three application frequencies (1, 2, or 3 season-'). Three control

treatments received no excreta and 0, 60, or 120 kg N ha-1. Dung and urine were collected from

animals grazing bahiagrass pasture and applied to ungrazed bahiagrass plots. A urine application

was 2 L distributed to a 60-cm diameter circle, and a dung application was 2 kg fresh weight

applied to a 30-cm diameter circle. Concentric rings (radii of 0-15, 15-30, 30-45, and for urine

plots, 45-60 cm from the center of excreta application) were clipped monthly from July through

October 2006 to measure forage responses, and lysimeters were placed 1.3 m below soil level

directly under the excreta application to measure nutrient concentrations in soil water. Herbage

data are reported for a circle of radius 45 cm from the center of the excreta deposit and also by









Table A-3. P values for main effects, interactions, and polynomial contrasts for ring number for urine-treated plots as discussed in
Chapter 4.

Excreta Response Source of Variation
Type variable Management Application Ring # (R) MI x AF MI x R AF x R MI x AF x R
Intensity (MI) Frequency
(AF)
Urine DM Harvested <0.0001 <0.0001 tL-<0.0001 0.1053 0.5152 <0.0001 0.4439
Q-0.5930
C-0.0188
Herbage N <0.0001 <0.0001 L-<0.0001 0.0431 0.0009 <0.0001 0.2391
Q-0.7160
C-0.3752
Herbage P 0.8317 <0.0001 L-0.3421 0.9125 0.1458 0.1839 0.4481
Q-0.5378
C-0.3637
Herbage IVDOM 0.0329 0.0028 L-0.0828 0.0661 0.7085 0.0539 0.7354
Q-0.9581
C-0.5357
N Harvested <0.0001 <0.0001 L-<0.0001 0.0796 0.4811 <0.0001 0.5471
Q-0.7771
C-0.0357
P Harvested 0.0024 <0.0001 L-<0.0001 0.3842 0.3090 <0.0001 0.1828
Q-0.7856
C-0.0097


T L = linear, Q = quadratic, and C


cubic effect of ring number.









White (1988) reported that NO3 leaching losses from intensively managed grazed pastures

can be in the range of 100 to 200 kg N ha-1 yr-. Jabro et al. (1997) found during spring, summer,


2-
and fall applications of urine leached more NO3-N below 1 m (19, 15, and 25 g m-2, respectively)

than the untreated control (1.7 g m-2). In comparison, when dung was applied in the summer of

the first and second year, 2.2 g m-2 and 1.8 g m-2 were leached, respectively.

Fate of Phosphorus from Excreta Applied to Grazed Grass Swards

Phosphorus Cycle in Grazed Grasslands

The P cycle is much more complex than the N cycle because the availability of P depends

not only on biologically mediated processes of organic P but on the chemistry of inorganic P

(Dubeux et al., 2007). Phosphorus exists in various chemical forms, inorganic P and organic P,

which vary widely in their behavior and fate in soils (Fuentes et al., 2006). The main sources of

organic P include manure, crop residues, and sludge. Organic P can be degraded into a long

availability form of orthophosphate. Recovery of P in a growing season by plants can be very

low; more than 80% may become immobile and unavailable for uptake because of adsorption,

precipitation, or conversion to the organic form (Schachtman et al., 1998).

Phosphorus in manure is in various forms but is mostly inorganic, indicating that P

availability following application should be high because this P fraction in manure converts to

plant-available P in a short period after application (Eghball et al., 2005). Phosphorus is

accumulated at low pH mainly into the alkali-extractable Al-P and Fe-P fractions which

represent phosphate adsorbed to soil colloids (Haynes and Williams, 1993).

Pathway of Phosphorus Loss from the Agroecosystem

Leaching losses and runoff

Phosphorus loss is much greater in surface runoff than subsurface flow and is dependent

on the rate, time, and method of P application; form of fertilizer or manure applied; amount and












Woodard, K.R., E.C. French, L.A. Sweat, D.A. Graetz, L.E. Sollenberger, B. Macoon,
K.M. Portier, B.L. Wade, S.J. Rymph, G.M. Prine, and H.H. Van Horn. 2002.
Nitrogen removal and nitrate-N leaching for forage systems receiving dairy
effluent. J. Environ. Qual. 31:1980-1992.

Woodard, Kenneth R., Edwin C. French, Lewin A. Sweat, Donald A. Graetz, Lynn E.
Sollenberger, Bisoondat Macoon, Kenneth M. Portier, Stuart J. Rymph, Brett L.
Wade, Gordon M. Prine, and Harold H. Van Horn. 2003. Nitrogen removal and nitrate
leaching for two perennial, sod-based forage systems receiving dairy effluent. J.
Environ. Qual. 32:996-1007.

Wrage, N., G.L. Velthof, M.L. van Beusichem, and 0. Oenema. 2004. Role of nitrifier
denitrification in the production of nitrous oxide. Soil Biol. Biochem. 33:1723-1732.

Yokoyama, Kazuhira, Hideaki Kai, and Hirofumi Tsuchiyama. 1991. Paracoprid dung
Beetles and gaseous loss of nitrogen from cow dung. Soil Biol. Biochem. 23:
643-647.









Herbage Phosphorus Concentration

Total-season P concentration was not affected by ring number in either the dung- (P >

0.498) or urine-treated plots (P >0.342) (Tables A-2 and A-3, respectively). There were no

interactions involving ring number for either dung (P > 0.376) or urine (P > 0.146). Herbage P

concentrations were 3.4, 3.3, and 3.4 g kg-1 in Rings 1 through 3, respectively, in dung-treated

plots. In spite of large amounts of total P applied in dung, there was little apparent uptake of P

from dung (Chapter 3) and as such there was little impact on spatial characteristics of herbage P.

In urine-treated plots, P concentrations were 3.3, 3.4, 3.3, and 3.6 g kg-1 for Rings 1 through 4,

respectively. The primary effect of urine on P concentration was associated with application

frequency (Chapter 3) and likely due to greater soil exploration by roots of more vigorous plants

that had received more urine. Spatial variability (i.e., ring effect) in P concentration was minimal

across the area sampled regardless of application frequency or management intensity.

Adeli and Varco (2001) evaluated the use of swine lagoon effluent on bermudagrass

[Cynodon dactylon (L.) Pers.] and johnsongrass [Sorghum halepense (L.) Pers.]. They suggested

that total P accumulation by forage grasses is more closely related to DM production rather than

tissue P concentration, which varied little. Newman et al. (2005) assessed the effect of N

fertilization on tissue P concentration and removal. Bahiagrass tissue P concentration decreased

from 3.7 for the no-N control to 3.1 g kg-1 when N was applied at 45 kg ha-1 per harvest, but

there was no change in the response as N fertilization increased to 60 and 90 kg ha-1 per harvest.

Herbage DM harvested more than doubled across this range of treatments. In urine-treated plots

in the current study, DM harvested and herbage N concentration were greatest closest to the

center of urine applications, and for most application frequencies decreased as distance from the

center increased. However, greater DM harvested and herbage N concentration were not

accompanied by decreasing herbage P concentration as occurred in the Newman et al. (2005)









Perez, T., S.E. Trumbore, S.C. Tyler, P.A. Matson, I. Ortiz-Monasterio, T. Rahn, and
D.W.T. Griffith. 2003. Identifying the agricultural imprint on the global N20
budget using stable isotopes. J. Geophys. Res. Atmos. 106:9869-9878.

Petersen, R. G., H.L. Lucas, and W.W. Woodhouse, Jr. 1956. The distribution of excretaof
freely grazing cattle and its effect on pasture fertility. I. Excretal distribution. Agron.
Journal. 48:440-444.

Peterson, P.R. and J.R. Gerrish. 1996. Grazing systems and spatial distribution of
nutrients in pastures: Livestock management considerations. p. 203-212.
Nutrient cycling in Forage Systems. 7-8 March 1996.

Pineiro, Gervasio, Jose Maria Paruelo and Martin Oesterheld. 2006. Potential long-term
impacts of livestock introduction on carbon and nitrogen cycling in grasslands of
Southern South America. Global Change Biology. 12:1267-1284.

Powell, J.M., F.N. Ikpe, and Z.C. Somda. 1999. Crop yield and the fate of nitrogen and
phosphorus following application of plant material and feces to soil. Nutrient
Cycling in Agroecosystems. 54:215-226.

Powell, J.M., F.N. Ikpe, Z.C. Somda, and S. Fernandez-Rivera. 1998. Urine effects on
soil chemical properties and the impact of urine and dung on pearl millet yield.
Expl. Agric. 34:259-276.

Powell, J.M., M.A. Wattiaux, G.A. Broderick, V.R. Moreira, and M.D. Casler. 2006.
Dairy diet impacts on fecal chemical properties and nitrogen cycling in soils.
Soil Sci. Soc. Am. J. 70:786-794.

Rhoads, F.M., R.L. Stanley, Jr., and E.A. Hanlon. 1997. Response to bahiagrass to N, P,
K on an Ultisol in North Florida. Soil Crop Sci. Florida Proc. 56:79-83.

Robinson, Donald L. 1996. Fertilization and nutrient utilization in harvested forage
systems Southern forage crops. p. 65-92. In R.E. Joost and C.A. Roberts (ed.) Nutrient
cycling in forage systems. Potash and Phosphate Institute and The Foundation for
Agronomic Research, Manhattan, Kansas.

Rockmann, T., J. Kaiser and C.A.M Brennikmeijer. 2003. The isotopic fingerprint of the
pre-industrial and anthropogenic N20 source. Atmos. Chem. Phys. 3:315-323.

Rotz, C.A., F. Taube, M.P. Russelle, J. Oenema, M.A. Sanderson, and M. Wachendorf.
2005. Whole-farm perspectives of nutrient flows in grassland agriculture.
Crop Sci. 45:2139-2159.









Herbage In Vitro Digestible Organic Matter

Herbage IVDOM was affected by management intensity (P = 0.001), application frequency

(P = 0.005), excreta type (P = 0.001), and application frequency X excreta type interaction (P =

0.099). Herbage in High management intensity plots was greater in IVDOM than in Average

plots (577 vs. 567 g kg-1). In no-excreta plots, IVDOM increased linearly (P = 0.007) from 559

to 572 g kg-1 as N rate increased from zero to 120 kg ha-1 (Table 3-7). Coleman et al. (2004)

suggested that N fertilization has shown no consistent effect on herbage digestibility, however a

number of recent studies have shown greater IVDOM of bahiagrass herbage grown at greater N

fertilization (Newman et al., 2006; Stewart et al., 2007). Stewart reported IVDOM of

Pensacola bahiagrass increased from 459 to 479 to 505 g kg-1 as N rate increased from 40 to 120

to 360 kg ha-1. Similarly, Newman et al. (2006) reported increases in IVDOM from 443 to 487 g

kg-1 when N fertilizer amount increased from 80 to 320 kg ha-1. One factor that works against

increasing IVDOM with greater N fertilization is increasing stem development at greater N

amounts (Coleman et al., 2004). Bahiagrass, however, has limited stem elongation, thus its

digestibility is less likely than most C4 grasses to be affected negatively by N fertilization.

Excreta application frequency X excreta type interaction occurred because there was no

effect of dung application frequency on bahiagrass IVDOM while urine application frequency

did have an effect (Fig. 3-5). Herbage IVDOM changed little following a single urine

application, but subsequent applications increased IVDOM from 565 to 588 g kg-1 (Fig. 3-5).

With the exception of the lack of IVDOM response to the first application of urine, these

IVDOM responses to urine (added N) correspond to those already described in the current study

for management intensity treatments and the no-excreta control treatments. Excreta type affected

IVDOM only for Application Frequency 3 when herbage from urine-treated plots had greater

IVDOM than from dung-treated plots (P = 0.006).









Data Presentation and Statistical Analyses

Herbage dry matter (DM) harvested is expressed on a total-season basis, i.e., the sum of

four harvests. Tissue N, P, and in vitro digestible organic matter (IVDOM) concentrations are the

weighted averages across four harvests. Nitrogen and P harvested were calculated for each

harvest date (product of DM harvested and nutrient concentration in harvested herbage) and

summed across the four harvests for a total-season measure. Excreta N recovery was calculated

by summing N harvested across the four harvests for a given excreta treatment and subtracting

from this sum the total seasonal N harvested from the appropriate (Average or High) no-excreta

control. This number was then divided by excreta N applied to that plot and the result expressed

as a percentage.

Data were analyzed using PROC MIXED in SAS (SAS Institute, Inc. 2007). For herbage

mass and nutritive value from the excreta source pastures, management intensity was a fixed

effect and replicate was a random effect. Date was considered a repeated measure.

For herbage data from the excreta application experiment, an initial analysis was

conducted using the 12 factorial treatment combinations (two management intensities, two

excreta types, and three application frequencies). These factors and their interactions were

considered to be fixed effects and replicate a random effect. Polynomial contrasts were used to

determine the nature of the response to application frequency.

The excreta type X application frequency interaction was significant for six of seven

herbage response variables for the factorial data set. To more completely explore this interaction

and to integrate the no-excreta controls, a follow-up analysis was conducted. This approach

considered the no-excreta control treatments for the Average and High management intensities

(60 and 120 kg N ha-1 yr-) as a zero level of excreta application frequency, providing a fourth

level of application frequency (0, 1, 2, and 3). Because of the presence of excreta type X












Table 3-5. Nutrients applied to dung treatments in bahiagrass swards. Calculations are based on
chemical analyses of fresh dung (Table 3-3) and a 2-kg fresh weight dung application
to a circle of 30-cm diameter. Data are expressed as kg of N, P, and K applied ha-1.
Date
Source Frequency 20 June 1 August 15 September Total
kg N-P-K ha-1
Average 1 557-182-389 0-0-0 0-0-0 557-182-389
2 557-182-389 574-201-373 0-0-0 1131-383-762
3 557-182-389 574-201-373 636-204-415 1767-587-1177
High 1 594-209-422 0-0-0 0-0-0 594-209-422
2 594-209-422 603-190-373 0-0-0 1197-399-795
3 594-209-422 603-190-373 523-164-405 1720-563-1200

Table 3-6. Nutrients applied to urine treatments in bahiagrass swards. Calculations are based on
the chemical analyses of urine (Table 3-4) and a 2-L volume of urine applied to a
circle of 60-cm diameter. Data are expressed as kg of N, P, and K applied ha-1.
Date
Source Frequency 20 June 1 August 15 September Total
kg N-P-K ha-1
Average 1 113-0.23-267 0-0-0 0-0-0 113-0.23-267
2 113-0.23-267 286-0.35-267 0-0-0 399-0.58-534
3 113-0.23-267 286-0.35-267 69-0.58-98 468-1.16-632
High 1 112-0.30-169 0-0-0 0-0-0 112-0.30-169
2 112-0.30-169 210-0.33-195 0-0-0 322-0.63-364
3 112-0.30-169 210-0.33-195 133-0.35-160 455-0.98-524

Herbage Dry Matter Harvested from Excreta-Treated Plots

Total-season herbage DM harvested was affected by management intensity (P < 0.001) and

excreta type (P < 0.001) main effects and by the excreta type X application frequency interaction

(P = 0.063) (Table A-i). For plots receiving excreta, as management intensity increased from

Average to High, forage DM harvested increased from 3230 to 3850 kg ha-1. Deenen and

Middelkoop (1992) found the extent to which urine affected herbage growth of perennial

ryegrass was dependent upon the level of N fertilizer applied. There was no effect of urine in

swards receiving 400 kg N ha-1, but positive effects on yield were noted in swards receiving 250

kg N ha-1. In the current study, there were no interactions involving management intensity (i.e.,









applications (for the two and three applications per year treatments) occurred at the same

location.

Forage Harvest and Laboratory Analyses

Forage was harvested every 28 d following initial excreta application (20 June), and

harvest dates were 18 July, 15 Aug., 12 Sept., and 10 Oct. 2006. To characterize the spatial

response of herbage DM harvested and nutritive value, herbage was harvested beginning with a

circle of radius 15 cm that was centered on the midpoint of the excreta application. Thereafter,

concentric rings were harvested sequentially. Rings are defined based on their radius from the

center of the excreta application (Fig. 4-1).

Rings 1 through 3 were harvested for dung-treated plots and Rings 1 through 4 for urine-

treated plots. For the urine-treated plots, R4 was harvested because of the greater area of urine

vs. dung applications and because urine typically affects an area -2.3 times the area of the urine

patch (de Klein, 2001). Throughout this chapter, the rings will be referred to as R1 through R4,

as defined above.

Following harvest, the herbage from each ring was dried separately at 600C, weighed, and

ground to pass 1-mm screen. Herbage analyses for N and P were conducted at the Forage

Evaluation Support Laboratory using a micro-Kjeldahl technique followed by semi-automated

colorimetric analysis of the digestate (Gallaher et al., 1975; Hambleton, 1977). Nitrogen and P

harvested in each ring were calculated by multiplying DM harvested times nutrient

concentration. In vitro digestible organic matter concentration (IVDOM) was determined using a

two-stage procedure (Moore and Mott, 1974).









Focusing on individual rings, there was an effect of number of dung applications on DM

harvested only in Ring 1. This response had significant linear (P = 0.047) and cubic (P = 0.041)

terms. The linear effect reflects greater depression in DM harvested with increasing dung

application frequency. The cubic effect is due to the greater negative impact of two vs. three

dung applications, a response for which there is no apparent explanation. In Rings 2 (P > 0.401)

and 3 (P > 0.137), there was no effect of dung application frequency on DM harvested. This

indicates that effect of dung on this response was limited to the circle of 15-cm radius that was

the initial area of application.

In a New Zealand pasture that was primarily ryegrass and white clover, dung pats of 1.8 kg

fresh weight were applied to a circle of 15-cm radius (MacDiarmid and Watkin, 1971). They

reported that 75% of grass tillers and rooted nodes of clover stolons under the dung pat were

dead within 15 d of application. This resulted in a significant reduction in yield from the area of

the pat as was observed in the current study. They noted that in one experiment, yield increased

in response to dung to a radius of 61 cm from the center of application, but in a second study the

increase was limited to a radius of 30 cm. In related research, Deenen and Middelkoop (1992)

applied a single dung pat (2.5 kg fresh weight) to a circle with radius of 15 cm in perennial

ryegrass plots. They measured DM harvested in five concentric rings or bands around the dung

pat. The radii of the areas harvested were 0 to 15, 15 to 30, 30 to 45, 45 to 60, and 60 to 75 cm.

They found that the dung-affected area was confined to 15 cm from the edge of the dung patch.

Thus, results of the current study differ somewhat from previous work. In this experiment, there

were no measurable effects of dung on DM harvested outside the area of dung application, while

several studies in the literature report effects extending at least 15 cm beyond the dung pat.









ACKNOWLEDGMENTS

Sincere appreciation and thanks are extended to Dr. Lynn E. Sollenberger, advisor and

mentor, for his time and effort to make my time here a great experience. Also, I would like to

thank the advisory committee, Dr. Donald A. Graetz, Dr. Joao M.B. Vendramini, and Dr. Yoana

C. Newman for serving on the committee, providing assistance, and for reviewing the thesis.

Special thanks are given to Dr. Kenneth R. Woodard for his help in planning and carrying out the

research, including his insightful input and assistance constructing lysimeters.

My thanks go to those who assisted in both field and lab work. Sindy Interrante, Kesi Liu,

Miguel Castillo, and Tiberio Saraiva collected excreta and harvested. Dwight Thomas and Sid

Jones were essential in helping with the project. Thanks are due to Richard Fethiere and the

group of the Forage Evaluation Support Laboratory for conducting sample analysis. I am

grateful to Dr. Newman for her assistance with statistical analyses.

I would like to thank my family for their support, love, and prayers. I want to thank the

Leech family for their support. Last but not least I want to thank W. Jacob Leech for his support

and love. I am grateful for his time collecting dung beetles.









than Average management intensity (10.9 vs. 8.0 kg ha-1 in dung-treated plots, respectively, and

13.7 vs. 12.4 kg ha-1 in urine-treated plots, respectively).

There was excreta application frequency X excreta type interaction for most herbage

responses in Chapter 3. Interaction occurred because dung application had no effect on most

herbage responses, whereas responses to urine were significant consistently. In urine-treated

plots, herbage DM harvested increased from 2760 at zero urine applications to 4670 kg ha-1 with

three applications. Over the same urine application frequencies, herbage N concentration

increased from 13.3 to 16.2 g kg-', P concentration from 3.45 to 3.87 g P kg-', in vitro digestible

organic matter concentration from 569 to 588 g kg-', N harvested from 37 to 75 kg ha-1, and P

harvested from 9 to 18 kg ha-1. Greater herbage response to urine- than dung-treated plots was

expected because of the high proportion of dung N that is in an organic form and the greater

availability to plants of nutrients in urine. The general absence of response to dung was not

expected and could be attributed to a number of factors including physical interference of the

dung pat, the high concentration of organic N as a proportion of total N, limited apparent activity

of dung beetles, and a drier than normal year leading to rapid drying and crusting of dung.

Excreta N recovery in harvested herbage was greater for urine than dung, averaging 22 and

2%, respectively. Recovery was also affected by excreta application frequency, decreasing from

28 to 18% as urine application frequency increased from one to three yr-f, and from 4 to less than

1% as dung application frequency increased from one to three. These recoveries are in the lower

part of the range reported in the literature, but it does not appear that leaching losses explain this

response. Significant responses of shallow soil water NO3-N concentrations to treatment

occurred at only one sampling date (15 Sept. 2006), and these values were less than 0.15 mg L-.

Greater concentrations occurred in individual wells at two other dates in August, but these were

































Figure 4-1. Diagram of harvested rings to quantify spatial pattern of response to dung and urine
application (R1, circle of 15-cm radius; R2, 15- to 30-cm radius; R3, 30- to 45-cm
radius; R4, 45- to 60-cm radius).

Statistical Analysis

Herbage DM harvested, herbage N, P, and IVDOM concentrations, and herbage N and P

harvested were analyzed using analysis of variance in PROC MIXED of SAS (SAS Institute,

Inc., 2007). Herbage DM, N, and P harvested are the sums of four harvests, and nutrient

concentrations and digestibility are weighted averages across the four harvests. Data were

analyzed by excreta type because of the greater number of rings harvested for urine- than dung-

treated plots. The no excreta, 60 kg N ha-1 treatment served as a zero excreta application

frequency for the Average management intensity treatments (both dung and urine), and the no

excreta, 120 kg N ha-1 treatment served as a zero excreta application frequency for the High

management intensity treatments (both dung and urine). Fixed effects in the models were

management intensity, excreta application frequency, ring number, and their interactions.









time of rainfall after application; and vegetation cover of the land (Shigaki et al., 2006). Long-

term application of manures and biosolids typically results in soil-P levels in excess of crop

needs (Elliot et al., 2002). Phosphorus typically does not leach unless the concentration in the

soil is very high or the soil has very low retention capability.

Graetz et al. (1999) measured P accumulation in manure-impacted soils in Florida. Sites

were chosen to reflect a wide range of impact and included active dairies, dairies abandoned for

years, beef cattle pastures, and areas not significantly impacted by human activities (native

areas). Pastures consisted of bahiagrass and bermudagrass. Soil samples were taken of each

horizon to a depth of at least 120 cm using a bucket auger. These soil samples were analyzed for

water soluble P (WSP), double-acid extractable P (DAP), and total P (TP). Native areas had less

WSP in all horizons (1 mg kg-1) compared to the other areas. The high cattle densities in the

abandoned dairies had WSP concentrations averaging 45, 13, 16, and 8 mg kg-1 for the A, E, Bh,

and Bw horizons, respectively. This indicated P movement vertically down the soil profile of

high cattle density areas, and part of the P existed in a form readily removed by water. The DAP

concentrations of native areas averaged 4 mg kg-1 compared to high cattle density areas (active

and abandoned dairy) that averaged 707, 63, 238, and 72 mg kg-1 for the A, E, Bh, and Bw

horizons, respectively. Double acid-extractable concentrations for the low cattle density areas

averaged 16, 4, 34, and 18 mg kg-1 for A, E, Bh, and Bw horizons, respectively. The P forms

(WSP, DAP, TP) in the pastures and forage areas tended to accumulate in the spodic horizon,

which indicated vertical P movement in the low cattle density soil area, also. Soil pH in native

areas was lowest, and it was the highest in the high cattle density areas, which can be related to

high dung and soil Ca and Mg concentrations.









through December. The greatest IVDOM for Pensacola was 623 g kg-1 in December 1998 and

the lowest occurred in August and was 467 g kg-1. All genotypes of bahiagrass tended to drop in

IVDOM during June to August due to a low soluble carbohydrate concentration in above-ground

plant parts.

In comparison to N fertilization, there are relatively few papers that address the effect of P

fertilization on Pensacola bahiagrass nutritive value. However, there has been increasing interest

in P nutrition of bahiagrass because of the role of P in water quality, and this has stimulated

greater research emphasis in recent years.

Ibrikci et al. (1999) found that when triple superphosphate was applied at rates of 0, 17, 34,

51, and 68 kg P ha-l, there was no change in bahiagrass herbage P concentration in the first year,

but in the second year the uptake of inorganic P increased significantly and P concentration

increased with P fertilization. Sumner et al. (1992) found when bahiagrass was fertilized with P

it resulted in increased P concentration and the increase was related to the amount of P applied.

Phosphorus concentration of Pensacola bahiagrass averaged 1.5 g kg-1 when fertilizer rate was 24

kg P, but when P fertilizer rate increased to 49 kg P ha-1 forage P concentration increased to 3.0 g

kg-1 (Burton et al., 1997).

Tiffany et al. (1999) sampled three grazed bahiagrass pastures that were growing on

different soils in Florida. Pastures were fertilized with 40 kg N ha-1. Forage P concentrations in

the first (3.0 g kg-1) and second years (3.1 g kg-1) were similar early in the season. There was a

decrease in P concentration from June to November in both years from approximately 3.0 to 1.3

g kg-1. The forage P concentration in October and November were below the 1.8 g kg-1

requirement for growing beef cattle (NRC, 1996).









a urination event, 11 mm of urine was applied in winter, spring, and summer. Paddocks were

harvested 7, 28, 56, and 84 d each season after the urine application. In the winter, irrigated

white clover-ryegrass total plant N recovered was 1% of applied 15N at 7 d, 2% of applied 15N at

28 d, 5% of applied N15 at 56 d, and 3% of applied N15 at 84 d. This compares with the non-

irrigated subterranean clover plots which had recoveries of 1% at 7 d, 3% at 28 d, 17% at 56 d,

and 20% at 86 d. The lower 5N recovery in the irrigated paddocks can be attributed to the loss of

5N in leachate. In spring there was no difference in plant uptake of 15N among paddocks because

sufficient soil moisture and active plant growth during spring and summer resulted in greater

recovery of 15N from the spring application.

Dung

Dairy manure (dung + urine) was applied annually at four different rates over 2 yr (0, 75,

150, 300 kg N ha-1 in 1990 and 0, 150, 300, and 600 kg N ha-1) to orchardgrass managed for hay

(Kanneganti and Klausner, 1994). Over the 2 yr, average N recovery was 40%. In a single

cutting the crop removed as much as 200 kg N ha-1 when 600 kg N ha-1 was applied; and tissue N

concentration was 45 g kg-.

In a dung pat study, Dickenson and Craig (1990) determined that uncovered and covered

dung still contained 86 to 95% of their original N content, respectively, over an 85-d period. This

has an impact on herbage accumulation under and surrounding dung pats, because most of the N

is still in the pats and not available to the plant.

Impact on Soil and Water

Soil

Urine application on pastures can increase soil pH by 3 units within approximately 1 d due

to hydrolysis of urea to NH4+ (van Groenigen et al., 2005a), although nitrification of the

remaining NH4+ to NO3- leads to a decrease in pH over a 2-wk period (Doak, 1952; Haynes and









0.007) effects for two dung applications and quadratic (P = 0.048) effects for three applications

(Fig. 4-7). This response is similar to that for DM harvested because there was no effect of dung

on herbage N concentration. Thus, as observed for DM harvested, an important factor affecting

the N harvested response appears to be physical interference of dung in Ring 1 that reduced both

DM and N harvested (Figs. 4-3 and 4-7). This effect was most pronounced when dung was

applied two or three times per year.

The effect of application frequency within a level of ring number was assessed to

determine how far from the center of application dung affected N harvested. There were a linear

(P = 0.055) and cubic (P = 0.047) effects of application frequency for Ring 1. Linear effects

reveal the general pattern of decreasing N harvested with increasing numbers of excreta

applications. The cubic effect was significant because two dung applications actually depressed

N harvested more than three applications. The biological significance of the cubic effect is not

clear. There were no effects of application frequency for Rings 2 (P > 0.441) or 3 (P = 0.194)

indicating that dung application had no effect on the N harvested response outside of the

immediate 15-cm radius circle to which it was applied. The minimal positive impact of dung,

despite containing high amounts of N, is attributed to the slow physical breakdown of the dung

pats and low mineralization rates of organic N in dung (Deenen and Middelkoop, 1992).









CHAPTER 1
INTRODUCTION

Rotz et al. (2005) define grassland agriculture as "a farming system that emphasizes the

importance of grasses and legumes in livestock and land management." Planted grasslands and

non-forested rangeland comprise nearly 30% of the USA land area (Barnes and Nelson, 2003)

and occupy more than 4 million ha in Florida (Dubeux et al., 2007). Because of the amount of

area involved, the fate of nutrients within these agroecosystems has important implications for

agricultural production and the environment.

Nutrient management in grasslands has received greater attention in recent years due to

soil nutrient insufficiency and associated pasture degradation in some areas (e.g., Brazil; Boddey

et al., 2004) and excessive applications of nutrients and negative environmental impact in others

(e.g., USA; Woodard et al., 2003). The major nutrient pools in grassland systems are soil,

atmosphere, live and dead plant material, and animals (Mathews et al., 2004; Dubeux et al.,

2007). Addition of livestock to grasslands increases the complexity of the nutrient cycle and the

rate of fluxes among nutrient pools; the latter increases the potential for nutrient loss to the

environment (Boddey et al., 2004). This is due to chemical and biological transformations that

occur during forage digestion, making the forms of nutrients in excreta more readily available for

uptake or loss than those occurring in live plants or plant litter (Jarvis et al., 1995; Rotz et al.,

2005).

In pasture-based livestock production systems, animals gather herbage, utilize a small

proportion of the nutrients and excrete the remaining nutrient compounds in patches (Rotz et al.,

2005). Efficient recovery of these nutrients is hindered by the large quantity of nutrients in a

single dung or urine event and because a disproportionately large number of excreta events occur


































2008 UNA RENEE WHITE









herbage P concentration in urine-treated plots (average of 3.4 g kg-1). These results support the

conclusions of Adeli and Varco (2001) that total P removed in forage grasses is more closely

related to DM production rather than tissue P concentration, which varies much less.

The effect of urine application frequency on P harvested within a ring number was

significant for R1 (linear, P <0.001), R2 (linear, P < 0.001; quadratic, P =0.051), and R3 (linear,

P < 0.001), but there was no effect of application frequency in R4 (Fig. 4-8). Thus like for DM

harvested, the impact of urine on P harvested extended up to 45 cm from the center of the

application or 15 cm beyond the edge of the application (Fig. 4-8).

For dung-treated plots, there was an application frequency X ring number interaction for P

harvested that occurred primarily because of physical interference of dung in R1 (Fig. 4-9). This

led to lower P harvested for two and three dung applications yr- in R1 compared to one or zero

applications (application frequency effect linear, P = 0.047; quadratic, P = 0.041), while in

Rings 2 and 3 there was no effect of application frequency on P harvested (P > 0.401 for Ring 2

and P > 0.137 for Ring 3). This response is reflected in quadratic ring number effects (P = 0.073)

on P harvested for three dung applications, linear (P = 0.004) and quadratic (P = 0.046) effects

for two applications, and no effect for one (P > 0.257) and zero applications (P > 0.289). Similar

to the responses for N and DM harvested, there was no effect of dung application on P harvested

beyond R1, the area actually covered by the dung pat. Lack of an effect of dung on P harvested is

likely associated with the slow breakdown of dung pats in the current study, perhaps due to drier

than normal weather and to little apparent activity of dung beetles which incorporate dung into

the soil and enhance nutrient release (Williams, 1950; Hughes et al., 1975).









Beaty, E.R. and John D. Powell. 1978. Growth and management of Pensacola
bahiagrass. Journal of Soil and Water Conservation 33-34:191-193.

Beaty, E.R., K.H. Tan, R.A. McCreery, and J.B. Jones. 1975. Root-herbage production
and nutrient uptake and retention by bermudagrass and bahiagrass. Journal of
Range Management. 28:385-389.

Blue, W.G. 1970. Fertilizer nitrogen uptake by Pensacola bahiagrass from Leon fine
sand, a Spodosol. Proc XI Int. Grassland Congress, Surfers Paradise, Queensland,
Australia. p. 389-392.

Blue, W.G. 1973. Role of Pensacola bahiagrass stolon-root systems in fertilizer nitrogen
utilization on Leon fine sand. Agronomy Journal. 65: 88-91.

Boddey, R.M., R. Macedo, R.M. Tarre, E. Ferreira, O.C. de Oliveira, C. de P. Rezende,
R.B. Cantarrutti, J.M. Pereira, B.J.R. Alves, and S. Urquiaga. 2004. Nitrogen
cycling in Brachiaira pastures: The key to understanding the process of pasture
decline. Agriculture, Ecosystems and Environment. 103:389-403.

Burson, B.L., and V.H. Watson. 1995. Bahiagrass, dallisgrass, and other Paspalum
species. p. 431-440. In R.F. Barnes et al. (ed.), Forages. Vol. 1. 5th ed. Iowa State
University Press, Ames, IA.

Burton, G.W. 2004. Warm Season (C4) Grasses. In L.E. Moser, B.L. Burson, L.E. Sollenberger
(eds.), Warm-season (C4) Grasses. ASA, CSSA, SSSA. Agronomy Monograph no. 45.
217-255.

Burton, G.W., R.N. Gates, and G.J. Gascho. 1997. Response of Pensacola bahiagrass to rates of
nitrogen, phosphorus and potassium fertilizers. Soil Crop Sci. Soc. Fla. Proc. 56:31-35.

Burton, G.W.. 1967. A search for the origin of Pensacola bahiagrass. Econ. Bot. 21:319-382.

Bussink, D.W., and 0. Oenema. 1998. Ammonia volatilization from dairy farming
systems in temperate areas: A review. Nutrient Cycling in Agroecosystems.
51:19-33.

Carran, R.A., and P.W. Theobald. 1999. Effects of excreta return on properties of a
grazed pasture soil. Nutrient Cycling in Agroecosystems. 56:79-85.

Carran, R.A., P.W. Theobald, and J.P. Evans. 1995. Emission of nitrous oxide from
some grazed pasture soils in New Zealand. Australian Journal of Soil Research.
33:341-352.

Chambliss, C.G., and M.B. Adjei. 2006. Bahiagrass. Publication SS-AGR-36, Agronomy
Department, University of Florida, Gainesville.









ring to show spatial patterns of response. Herbage response for the circle of 45-cm radius was

affected to the greatest extent by application frequency X excreta type interaction. Interaction

occurred because dung application had no effect on most herbage responses, whereas responses

to urine were consistently significant. In urine-treated plots, herbage DM harvested increased

from 2760 at zero applications to 4670 kg ha-1 with three, while over the same application

frequency herbage N concentration increased from 13.3 to 16.2 g kg1, P concentration from 3.45

to 3.87 g P kg-', in vitro digestible organic matter concentration from 569 to 588 g kg-', N

harvested from 37 to 75 kg ha-l, and P harvested from 9 to 18 kg ha-l. Excreta N recovery was

greater from urine than dung and decreased as application frequency increased from one to three

(28 to 18% for urine and 4 to < 1% for dung). Spatial characteristics of response were assessed

within excreta type and were consistently affected by application frequency X ring number

interaction. In urine-treated plots, herbage response generally was greatest near the center of the

urine deposit and decreased as distance from center of excreta application increased. In dung-

treated plots, physical interference by dung resulted in decreased herbage DM, N, and P

harvested in the area under the dung deposit. Urine affected DM harvested and herbage P

concentration up to 15 cm and N concentration and N harvested up to 30 cm beyond the edge of

the urine application. Dung had no effect on any response outside the area physically impacted

by dung. Shallow soil water was affected by treatment at only one sampling date, and N03-N

concentration was > 10 mg L-1 in only three samples during the entire study. In conclusion, urine

has much greater impact on herbage response than dung, both in the area impacted physically by

the excreta and beyond that area. Greater N recovery following single vs. multiple excreta events

per site per year emphasizes the importance to sustainable grassland management of grazing

practices that increase uniformity of excreta deposition.









LIST OF TABLES


Table page

2-1 Number and weight or volume of dung or urine events per day and surface area
covered (adapted from Haynes and Williams, 1993) .....................................................25

3.1 Chemical composition of excreta analyzed immediately after collection (fresh) and
after storage (stored) for up to 8 d at 40C. Each value is the mean of the analysis of
tw o subsam ples. .............................................................................53

3-2 Herbage mass, N, P, and in vitro digestible organic matter (IVDOM) concentrations
of bahiagrass herbage during excreta collection periods in 2006.....................................59

3-3. Composition of fresh dung from Average and High management intensity source
treatments during three collection periods in 2006. Each value reported is the average
across three subsamples from a composite dung sample. Dung was composite
across replicates within a management intensity treatment, so statistical comparisons
of treatm ent effects are not possible. .......................................................................... 61

3-4. Composition of urine from Average and High management intensity source
treatments during three collection periods of 2006. Each value reported is the
average across three subsamples. Urine was composite across replicates within a
management intensity treatment, so statistical comparisons of treatment effects are
n ot p o ssib le. ............................................................ ................ 6 1

3-5 Nutrients applied to dung treatments in bahiagrass swards. Calculations are based on
chemical analyses of fresh dung (Table 3-3) and a 2-kg fresh weight dung
application to a circle of 30-cm diameter. Data are expressed as kg of N, P, and K
ap p lied h a-1 ............................................................................... 62

3-6 Nutrients applied to urine treatments in bahiagrass swards. Calculations are based on
the chemical analyses of urine (Table 3-4) and a 2-L volume of urine applied to a
circle of 60-cm diameter. Data are expressed as kg of N, P, and K applied ha-1...............62

3-7 Effect of N fertilization on bahiagrass herbage responses for treatments to which no
excreta w as applied. .......................... .......................... .. .. .. ................. 65

3-8 Monthly rainfall totals for 2006 for the research location and the 30-yr average for
Island Grove, FL. Island Grove is located 10 km from the research location and is the
nearest site for which 30-yr data exist. ........................................ .......................... 65

A-1 P values for interactions and main effects on response variables discussed in Chapter
3 ........................................................... ................................. . 1 2 1

A-2 P values for main effects, interactions, and polynomial contrasts for ring number for
dung-treated plots as discussed in Chapter 4. ........................................ ...............122










Table 2-1. Continued
Reference Stock Mean Weight of Area Mean Volume Area
type number of single covered number of single covered
defecations defecation by of urination by
per day (kg wet defecation urinations (L) urination
wt) (m2) per day (m2)
Frame (1971) Dairy 11 2.7 -- 11 1.9
cow
MacDiarmid Dairy 13.9 1.82 0.07
and Watkin cow
(1972)
Robertson Dairy -- -- 10 2
(1972) cow
During and Beef -- 0.05
Weeda steer
(1973)
Richards and Dairy -- 0.05 -- 0.49
Wolton cow
(1976)
Weeda Beef 10.5
(1979) steer
Williams et Dairy -- 0.16
al. (1990) cow

Chemical composition

Nutrients in urine are in plant-available forms or are rapidly mineralized within a few days

(Mathews et al., 1996). Because urine is 500 to 800 g urea-N kg-1 of total-N, it is hydrolyzed

very rapidly (Rotz et al., 2005), resulting in a release of available nutrients. The urea in the urine

is broken down by urease which is an enzyme produced by microbes in the soil. Urine contains

many amino acids, such as hippuric acid. The hippuric acid in urine was reported to have a

controlling effect on both hydrolysis of urine N and on NH3 volatilization (Doak, 1952;

Whitehead et al., 1989). When urine-N concentrations are lowered or the amount of urine N is

affected by altering forage source, feed additives, or grazing regimes, there may be a significant

effect on N20 emissions, due to lower amounts of N being applied to the soil (Oenema et al.,









application frequency interaction in the initial analysis, these data were analyzed by excreta type

and polynomial contrasts were used to determine the nature of the response to application

frequency. Water data were analyzed by sampling date using the same statistical approach

described for herbage data from the excreta application experiment.

An additional follow-up analysis was carried out for the data from the three control

treatments. In this case, the treatment was N fertilizer amount, so the effect of N on herbage

responses was determined using polynomial contrasts.

For all response variable from both experiments, differences were deemed significant when

P < 0.10. Data presented are least squares means.

Results and Discussion

Characteristics of Excreta Source Pastures

There were no interactions of sampling date and management intensity for herbage

characteristics of excreta source pastures, but interaction means are presented so that excreta

nutrient concentrations can be considered relative to pasture characteristics at time of excreta

collection (Table 3-2). There were no management intensity effects for herbage mass (3030 kg

ha-1), herbage P concentration (2.0 g kg-1), or herbage IVDOM concentration (520 g kg-1), but

herbage N concentration was greater for High than for Average management intensity (17.1 vs.

14.4 g kg-1). Sampling date affected herbage mass and IVDOM only, although there was a trend

(P = 0.117) toward greater herbage N concentration in July than in the other months. Mass was

similar in June (2820 kg ha-1) and September (2810 kg ha-1), but was greatest in July (3470 kg

ha-1). The second application of N fertilizer occurred on 19 July, 10 d after the second

application of N fertilizer, and this timing along with typical seasonal growth patterns (Stewart et

al., 2007) likely explain the greater herbage mass in July and the trend toward greater N

concentration. Herbage IVDOM was similar in June and July (558 and 540 g kg-1) but was least









Herbage Phosphorus Concentration

Herbage P concentration was affected by application frequency (P = 0.003) and the

application frequency X excreta type interaction (Table A-i). Interaction occurred because there

was no effect of dung application frequency on herbage P concentration (P > 0.583), but there

were linear (P = 0.025) and quadratic (P = 0.002) effects on herbage P concentration in urine-

treated plots (Fig. 3-4). At Application Frequencies 1 and 3, P concentration in herbage from

urine-treated plots had lesser and greater P concentrations (P = 0.004 and 0.035), respectively,

than herbage from dung-treated plots (Fig. 3-4). There was no effect of excreta type at

Application Frequency 2 (P = 0.752). Herbage P concentration in dung-treated plots remained in

a narrow range between 3.33 and 3.45 g kg-1. Absence ofbahiagrass DM harvested and herbage

P concentration response to dung application frequency indicate that almost none of the large

amounts of P applied (up to nearly 600 kg P ha-1 for the three applications yr- treatment) were

taken up by the plant.

The strong quadratic effect of urine application frequency on tissue P concentration

occurred because of a decline from 3.45 with no urine applied to 2.83 g P kg-1 with one urine

application. The initial decline is likely related to greater herbage DM accumulation following

one urine application with essentially no additional P applied (0.2-0.3 kg P ha-1; Table 3-6).

Powell et al. (1998) reported that when urine from rams was applied to pearl millet [Pennisetum

glaucum (L.) Br.], the urine not only increased pH, but also increased available P, especially

during the first week after application. In the current study, as application frequency increased

from one to three yr-1, herbage P concentration increased from 2.83 to 3.87 g P kg-1. This large

increase in P concentration is not due to greater P application because those amounts changed

only from 0.23 (Average) or 0.30 (High) kg P ha-1 for one urine application to 1.16 (Average) or

0.98 (High) kg P ha-1 for three applications (Table 3-6). Thus, the response may be due to greater









Dung

Quantity and distribution in grazed pasture

Cattle defecate approximately 12 times d-1 (Peterson and Gerrish, 1996), with the reported

range from 8 to 16 times d-1 (Barrow, 1967; Wilkinson and Lowrey, 1973; Peterson and Gerrish,

1996) and an average wet weight of 1.5 to 2.7 kg per defecation (Haynes and Williams, 1993;

Table 2.1). The area covered by a single dung pat is generally from 0.05 to 0.09 m2. Like urine,

dung deposition is often not uniform across the pasture and is concentrated in high traffic areas.

Chemical composition

Dung contains undigested herbage residues, products of animal metabolism, ingested soil,

and a large biomass of microorganisms (Rotz et al., 2005). Most of the nutrients in dung are not

immediately plant available; the main factors affecting decomposition and disappearance of dung

are microbial decomposition, weathering, disintegration of pats due to invertebrates, and

consumption and removal of dung by insects and lumbricids (Holter, 1979; Lee and Wall, 2006).

Holter (1979) found hot and dry summers slowed dung disappearance, resulting in 65% of the

dung pat remaining after 65 d compared to other years with pat disappearance (only small

particles or "soil-like material" left) in 50 d. Lee and Wall (2006) estimated that complete

disappearance for dung pats can take 57 to 78 d and 88 to 111 d in spring and summer,

respectively.

Almost all P excreted is in the feces, and more than 70% of P in manure is inorganic P

(Eghball et al., 2005). Nitrogen is also excreted through dung at about 8 g kg-1 of feed consumed.

Nitrogen in dung is only 20 to 25% water soluble, and volatile loss of NH3 is less than 5% (Rotz

et al., 2005). About 10 to 20% is undigested dietary N and the remaining 60% is in bacterial cells

(Whitehead, 1986). On average, "as excreted" cattle feces is 5.2 g N kg-1, 1.9 g P kg-1, and 4.1 g

K kg-1, for an N:P:K ratio of 2.7:1:2.2 (Edwards, 1996).









than Average (57 vs. 47 kg ha-1), and for no excreta plots the response to N fertilizer amount was

linear, increasing from 18 to 42 kg ha-1 as N amount increased from 0 to 120 kg ha-1 (Table 3-7).

Frequency X type interaction occurred because N harvested increased linearly (P < 0.001)

from 37 to 75 kg ha-1 as urine application frequency increased, but there was no effect of dung

application frequency on the response (Fig. 3-6). Excreta type affected N harvested at

Application Frequencies 1 through 3 (P < 0.0001) (Fig. 3-6).

Deenen and Middelkoop (1992) fertilized perennial ryegrass with either 250 or 400 kg N

ha-1 yr- and applied a single application of dung at three different dates. As in the current study,

N harvested in the grass was not significantly different than in the controls which did not receive

dung. Ma et al. (2007) applied 6 kg of sheep dung uniformly to a mixed bunchgrass sward in

Inner Mongolia. The area was harvested 32 and 65 d after dung application. Nitrogen harvested

32 d after application was 48 and 35 kg N ha-1 for dung-treated and control plots, respectively.

Sixty-five days after dung application, herbage N harvested was 32 and 25 kg ha-1 for these two

treatments, respectively. Greater impact of dung in the study by Ma et al. (2007) than in the

current study can be attributed to dung being uniformly spread over the plots and the general

form of sheep dung compared to that of cattle. Thus, there are a range of N harvested responses

to dung in the literature, with amount of N fertilizer, amount of dung, and type of dung affecting

the response.

Cuttle and Bourne (1993) made single urine applications (3.5 L m-2) to different perennial

ryegrass plots at five dates between August and November. Nitrogen harvested ranged from 70

for early applications to 4 kg ha-1 for late applications. For untreated controls, N harvested was

much less, ranging from a low of less than 1 kg ha-1 to a high of 13 kg hal-. Similar results were

also noted by Ball et al. (1979) using perennial ryegrass. They found that N harvested increased









1997). Potassium in urine may be leached or converted to less available forms, which affect

supply (Carran and Theobald, 1999).

Livestock on forage diets have a urine pH of approximately 7.4 (Rotz et al., 2005). The

specific concentrations ofN in cattle urine depend on factors such as diet and water consumption

but normally range from 8 to 15 g N L-1 (Whitehead, 1970; Clough et al., 2004). Reducing cattle

N intake by changing diet composition can lead to lower N concentration in urine with the

volume of urine unchanged, fewer urinations of the same volume but with unchanged N

concentration, or an unchanged number of urinations but with a smaller volume of urine and

unchanged N concentration (Bussink and Oenema, 1998; van Groenigen et al., 2005a). Salts and

other feed additives will dilute the concentration of N in urine through more water consumption

(van Groenigen et al., 2005a).

Effect on herbage production and nutritive value

Nutrients in urine enhance overall pasture productivity disproportionately to the area

physically covered by the excreta site (Peterson and Gerrish, 1996). Recycled nutrients can

account for up to 70% of annual pasture production in low-input systems (Mathews et al., 1996).

Short-term effects of urine have a greater impact on productivity than dung (Carran and

Theobald, 1999). When cattle and sheep were corralled and combined dung and urine applied to

pearl millet (Pennisetum glaucum L.), the millet plots had an average of 53% greater seasonal

yield than where only dung was applied (Powell et al., 1998).

Decau et al. (2003) conducted an experiment using perennial ryegrass planted in three

types of soil. Cow urine spiked with 15N was applied to the plots at a rate of 7 mg N L-1 (urea N

= 4.61 mg L-1, and N = 2.47 mg L-1) and top dressed with 15 g N m2 yr1 of ammonium nitrate.

The controls, which did not receive any form of N in Year 1, had a forage dry matter production

of 5 to 6 Mg ha-1. Application of urine in the spring increased DM yields by 17 to 33%, but the





























To Jacob.









that these applications occurred earlier in the growing season, allowing more time for nutrient

uptake by the bahiagrass.

Another reason for low capture of N could be growth patterns of bahiagrass. Blue (1973)

noted that fertilizer N recovery by bahiagrass was low early in stand life due to large amounts of

N stored in the rhizome-root system. Although our plots were long-time established stands,

visual observations suggest that low soil fertility or some other factor was limiting overall stand

vigor at the start of the experiment. Thus our plots may have responded somewhat like

establishing stands in terms of building up the rhizome-root system and capturing significant

amounts of N in storage organs. Core samples will be taken following completion of the second

year of the study to assess the impact of excreta and control treatments on storage organ mass

and N concentration.









Williams, 1992; van Groenigen et al., 2005a). Powell et al. (1998) found that when urine from

rams was applied to pearl millet, the urine not only increased soil pH from 4.5 to 9.5, but also

increased available P, especially during the first week after application. The pH in the top 15 cm

of the soil remained elevated for approximately 128 d, and Bray I-P appeared to decrease

compared to the untreated control, perhaps indicating movement downward of P. Not only were

there decreases in soil ammonium levels in the urine patches 1 d after urine application, but also

a large increase in soil nitrate levels to a depth of 30 to 45 cm.

When plots of 'Nandi' setaria (Setaria sphacelata var. sericea) received 1.4 L of urine (37-

48 g N m2), the greatest increase in pH in the 0- to 0.5-cm soil layer occurred 2 to 6 h after urine

application from an initial pH of 5 to a pH of 8, and after the second day there was a linear

decline from pH of 8 to a pH of 4.5 (Vallis et al., 1982). Mineralization of urea occurred by 2 h

after application, with only 200 g kg-1 of the original amount remaining, and by 14 h after

application there was less than 20 g kg-1 present. This caused nitrate to accumulate after the first

day, with a maximum NO3s concentration in the 0- to 1.5-cm soil layer on the seventh day.

Powell et al. (2006) reported feces applied to silt loams generally increased net soil

inorganic N (IN), but when feces was applied to a sandy loam soil it caused net IN to decrease

for 112 d followed by a rapid increase. This can be attributed to soil texture and chemical

properties which had the greatest impact on soil N mineralization. Another possible reason is

when microbial populations are deprived of supplied organic C input, biologically active pools of

soil organic matter decline and this leads to a net release of soil nutrients. Also, when fed high

protein diets cattle feces produced higher net soil IN than feces from low protein diets. Cows fed

different diets (corn silage-low CP; corn silage-high CP; alfalfa-low CP; alfalfa-high CP) had









that P cycling through grazing animals was inefficient in the short term because of negligible

return of P through urination, the small area of coverage by manure each grazing season, and the

low mobility and spatial unavailability of P in manure (Peterson and Gerrish, 1996).

There have been experiments showing benefits of excreta application to pastures, but these

impacts have not exclusively been linked to P addition. One such study by Dalrymple et al.

(1994) reported forage yield increases over a no-excreta control of 220 and 880 kg ha-1 in areas

affected by manure piles and urinations, respectively (Peterson and Gerrish, 1996), although this

benefit was not due solely to P addition. During and Weeda (1973; cited by Peterson and

Gerrish, 1996) determined that a cattle dung pat affected forage growth in a zone 5-fold larger

than the area covered by the pat. Forage P yields around the 0.25-m2 area increased 23%.

Summary

The southeastern USA is an important area of livestock and bahiagrass production.

Because of increasing cost of fertilizer nutrients and growing concern regarding the impact of

agricultural systems on the environment, there is a need for greater understanding of nutrient

relationships in grazed pasture. There is little information in the literature evaluating the impact

of type and number of excreta applications to pasture on leaching of N, changes in soil nutrient

concentration, and herbage growth and nutritive value. The objective of the research reported in

the chapters that follow is to characterize the effects of cattle excreta application on 1) bahiagrass

yield, chemical composition, and nutrient recovery and 2) nitrate leaching to shallow ground

water.









also increase overall pasture herbage accumulation and, based on results from Chapter 3,

increase the efficiency of N recovery.









can result in slower nutrient release (Holter, 1979; Dickenson and Craig, 1990; Yokoyama et al.,

1991).

Table 3-7. Effect of N fertilization on bahiagrass herbage responses for treatments to which no
excreta was applied.
Response Nitrogen fertilizer applied (kg ha-1) Polynomial Standard
variable; 0 60 120 contrast Error
DM harvested (kg ha-1) 1310 2450 3060 L*, Q 125
N concentration (g kg-1) 13.6 12.7 13.9 NS 0.46
P concentration (g kg-1) 3.87 3.35 3.54 NS 0.22
IVDOM (g kg-1) 559 567 572 L* 7.3
N harvested (kg ha-) 17.6 31.1 42.4 L** 1.4
P harvested (kg ha-) 5.0 9.6 11.1 L** 0.6
L = linear, Q = quadratic; **, P < 0.01; *, NS, P > 0.10; Letter followed by no symbol, P < 0.10
Dry matter (DM) harvested; in vitro digestible organic matter (IVDOM)

Table 3-8. Monthly rainfall totals for 2006 for the research location and the 30-yr average for
Island Grove, FL. Island Grove is located 10 km from the research location and is the
nearest site for which 30-yr data exist.
Month 2006 30-yr Average'


January
February
March
April
May
June
July
August
September
October
November
December
Annual


---------------------------------------------mm--- ------------------
22 107
147 83
4 85
52 75
12 119
130 180
233 167
159 172
93 115


1031


1330









recovery of N, 49%, for the 100 kg N ha-1 yr1 fertilizer treatment. The higher level of N fertilizer

resulted in a decrease of urine N uptake in above-ground parts of ryegrass.

Ball et al. (1979) applied 300 (N300) and 600 kg N ha-1 (N600) as a mixture of urea and

urine to perennial ryegrass-white clover plots in New Zealand. The urine was obtained from beef

cattle fed pasture clippings in a barn. The ryegrass-white clover sward had an apparent N

recovery of 37 and 23% for N300 and N600, respectively. Total herbage N yields increased with

increasing N application, 143 to 202 kg N ha-1 as amount of N applied increased from 300 to 600

kg.

Thompson and Fillery (1998) conducted three experiments applying sheep urine spiked

with 15N at different times of the year to a rotation system including wheat (Triticum aestivum

L.) pasture. Applications were either to pasture residues (Experiments 1 and 2), which were

sown to wheat, or to growing pasture in winter-spring (Experiment 3). In Experiments 1 and 2,

urine was applied in November 1990, April 1991, October 1991, January 1992, and March 1992

(9.8 g N m2, 46.1 gN m-2, 4.6 g N m2, 15.6 g N m-2, and 13.6 g N m2, respectively). Wheat

recoveries in November 1992 were 4, 7, and 12% of 15N applied. For Experiment 3, the urine
-2
was applied in August and September 1992 (12.3 and 25.9 g N m-2, respectively). Nine days after

urine application in August, 14% of the applied 15N was taken up by the pasture plants, and after

6 wk 53% had been recovered. In September, 47% of the 15N applied was recovered by the

growing plants.

Pakrou and Dillon (1995) collected urine from dairy cows and spiked it with 15N-labelled

urea. The urine was applied to either irrigated white clover-perennial ryegrass paddocks, which

received 25 to 30 mm of irrigation in a single application per week, or non-irrigated subterranean

clover (Trifolium subterraneum L.)-based paddocks with annual grasses and weeds. To simulate













18

17

R-16

o 15

14


0 13 ---Urine-3
Z ---Urine-2
12 Urine-1
SUrine-0
11

10
1 2 3 4
Ring


Figure 4-4. Ring number by urine application frequency interaction (P < 0.001) on total-season
herbage N concentration during 2006. Ring number effects were linear for three (P <
0.001) and for two (P < 0.001) applications, but not significant for one (P > 0.171)
and zero (P > 0.128) urine applications. Standard error of a treatment mean was 0.25
g kg-1. Ring Numbers 1 through 4 refer to sampling areas 0 to 15, 15 to 30, 30 to 45,
and 45 to 60 cm from the center of the urine application, respectively.









Research Unit, Citra, FL (29.41N latitude, 82.02 W longitude)was used for applications of

excreta. Site characteristics and the rationale for choosing these locations were described in

Chapter 3.

Treatments and Design

Treatments applied to ungrazed bahiagrass plots were two management intensities (the

same N amounts as were applied to excreta source pastures, i.e., 60 [Average] and 120 [High] kg

ha-1 yr1), two types of excreta (dung and urine), and three frequencies of excreta application

(one, two, or three times per year). The 2 x 2 x 3 factorial accounted for 12 treatments. In

addition, there were three control treatments that received no excreta and were fertilized with N

at 0, 60 (no excreta control for Average management intensity plots), or 120 kg ha-1 yr-1 (no

excreta control for High management intensity plots). The 15 treatments were replicated three

times in a randomized complete block design. Plots were 3 x 3 m in area with a 1-m bahiagrass

alley surrounding each plot.

Average and High management intensity source pastures were defined based on N

fertilizer amount and stocking rate. Average management pastures received 60 kg N ha-1 yr1 and

were stocked with two yearling heifers hal-, and High pastures received 120 kg N ha-1 yr-1 and

were stocked with four yearling heifers ha-l. These two management intensities were selected to

represent fertilization and stocking regimes that are common in the Florida livestock industry

(Average) or represent approximately the most intensive management applied to grazed

bahiagrass (High). In addition, based on previous work by Stewart et al. (2007) it is expected that

these treatments will result in forage that varies in nutritive value, especially N concentration,

and that these differences could affect the composition of dung and urine.









of greater P fertilization for at least 15 to 20 yr, even though P fertilization was discontinued in

1987, 12 yr before the start of this study.

In many soils, high P-sorbing oxide components keep leachate P levels well below

eutrophication thresholds (0.01 to 0.05 mg L-1); vertical flux of P is potentially significant in

areas with shallow ground water and coarse-textured soils with little P-sorbing capacity (Elliot et

al., 2002). Phosphorus loss is a concern on sandy, high water table soils with limited P-holding

capacity (Mathews et al., 1996), like those in Florida. Phosphorus loss increases with input rate

or soil test P level (Rotz et al., 2005). Phosphorus soil test values over 330 kg P ha-1 (Mehlich 1)

in the upper 20 cm of soil greatly increases the chance for loss from the system, which will affect

water quality (Sigua et al., 2004).

Mineralization

Mineralization of organic P compounds represents an important P source for plants,

especially in soils with low levels of bioavailable P (Fuentes et al., 2006). Mineralization of

organic P in the soil is catalyzed by various enzymes (phosphatases), including phytase (Eghball

et al., 2005). These enzymes play a fundamental role in the P cycle allowing orthophosphate to

be released from organic and inorganic compounds and increasing the bioavailable P (Fuentes et

al, 2006). Mineralization is greatest when soil moisture is near field capacity and declines with

soil drying. For P to become available for plants, phosphatase enzymes must breakdown the

organic P compound (i.e., phosphate diester-nucleic acids, phosphoprotein, phospholipids, or

phosphate monoester-glucose-6-phosphate, nucleotides) into phosphate (Fuentes et al., 2006).

Recovery of Excreta Phosphorus by Grassland Plants

It has been reported that dry beef cattle manure contains between 2.94 (Iyamuremye et al.,

1996) and 4.02 g kg-1 ofP (Griffin et al., 2003). Manure contains significant amounts ofP that

can be utilized for crop production (Eghball et al., 2005). Wilkinson and Lowrey (1973) reported










N ha- (Haynes and Williams, 1993; Clough et al., 2004). Pakrou and Dillon (1995) collected

fresh urine from dairy cows grazing clover (Trifolium spp.)-perennial ryegrass (Lolium perenne

L.) pastures. The N loading under the urine patches ranged between 650 (autumn) and 1370 kg N

ha-1 (summer) in the soil. Applying this much N increases the likelihood of nutrient losses.

Table 2-1. Number and weight or volume of dung or urine events per day and surface area
covered (adapted from Haynes and Williams, 1993).


Reference


Johnstone-
Wallace and
Kennedy
(1994)
Castle et al.
(1950)
Hancock
(1950)
Goodall
(1951)
Waite et al.
(1951)
Doak (1952)

Hardison et
al. (1956)
Petersen et
al. (1956)
MacLusky
(1960)
Davies et al.
(1962)
Wardrop
(1963)
Hogg (1968)

Weeda
(1967)


Stock Mean
type number of


Beef
cow


Dairy
cow
Dairy
cow
Dairy
cow
Dairy
cow
Dairy
cow
Dairy
cow
Dairy
cow
Dairy
cow
Dairy
cow
Dairy
cow
Dairy
cow
Beef
steer


Weight of


Area Mean Volume


single covered


defecations defecation
per day (kg wet
wt)
11.8 1.77


11.6

12.2

12


by
defecation
(m2)
0.06


Area


number of single covered


of
urinations
per day
8.5


urination
(L)


by
urination
(m2)
m-


9.8

10.1

11


1.48

2.27


15.4

12

11.6

12

16.1




10.5


0.09




0.05

0.07


9.4

8




10

12.1


0.28




0.19




0.18









REFERENCES


Adeli, Ardeshir, and Jac J. Varco. 2001. Swine lagoon effluent as a source of nitrogen
and phosphorus for summer forage grasses. Agron. J. 93:1174-1181.

Arthington, John, Patrick Behlen, and Fritz Roka. 2003. Effect of stocking rate on
measures of cow-calf productivity and nutrient loads in surface water runoff.
EDIS. May 2003. http://edis.ifas.ufl.edu/AN141 .

Allen, A.G., S.C. Jarvis, and D.M. Headon. 1996. Nitrous oxide emissions from soils due to
inputs of nitrogen from excreta return by livestock on grazed grassland in the UK. Soil
Biology & Biochemistry 28:597-607.

Anger, M., C. Hoffman, and W. Kuhbauch. 2003. Nitrous oxide emissions from artificial urine
patches applied to different N-fertilized swards and estimated annual N20 emissions for
differently fertilized pastures in an upland location in Germany. Soil Use and
Management 52:191-201.

Ball, P.R., and D.R. Keeney. 1981. Nitrogen losses from urine affected areas of a New
Zealand pasture. Proceedings of the XIV International Grassland Congress.
Edited by J. Allan Smith and Virgil W. Hays. June 15-24. p. 42-344.

Ball, Roger, D.R. Keeney, P.W. Theobald, and P. Nes. 1979. Nitrogen balance in urine
affected areas of a New Zealand pasture. Agron. J. 71:309-314.

Bardgett, R.D., and D.A. Wardle. 2003. Herbivore-mediated linkages between
aboveground and belowground communities. Ecology 84:2258-2268.

Barnes, R.F., and C.J. Nelson. 2003. Forages and grasslands in a changing world. p. 3-23. In R.F.
Barnes et al. (ed.), Forages: An introduction to grassland agriculture. Iowa State Press,
Ames, IA.

Barrow, N.J. 1967. Some aspects of the effects of grazing on the nutrition of pastures. J.
Aust. Inst. Agric. Sci. 33:254-262.

Barton, L., A. Schipper, G.F. Barkle, M. McLeod, T.W. Speir, M.D. Taylor, A.C.
McGill, A.P. van Schaik, N.B. Fitzgerald, and S.P. Pandey. 2005. Waste
Management: Land application of domestic effluent onto four soil types: plant
uptake and nutrient leaching. J. Environ. Qual. 34: 635-643.

Bathurst, N.O. 1952. The amino acids of sheep and cow urine. J. Agric. Sci. 42:476-478.

Beaty, E.R., R.H. Brown and J.B. Morris. 1968. Response of Pensacola bahiagrass to
intense clipping. p. 538-542. In Proc. 11th Int. Grassl. Congress. Surfers
Paradise, Australia.









compacted areas as "hot spots" of N20 emission from pastures (Anger et al., 2003; Carran et al.,

1995; and van Groenigen et al., 2005b). Allen et al. (1996) reported a shift from N20

consumption to N20 production during warm conditions. Clough et al. (2004) found when

synthetic urine was applied to limed soil cores with different water-filled pore spaces (WFPS)

there were significant differences in N20 emission and soil pH. Soils at field capacity (54%

WFPS) that had a pH greater than or equal to 5.9 produced the least amount of N20, while the

N20 flux was greatest in saturated (80% WFPS) treatments with a pH of 4.7.

Under suitable conditions, i.e. moisture, temperature, pH, etc., N20 can be reduced in the

soil and released as benign N2 (Firestone, 1982; Clough et al., 2004). van Groenigen et al.

(2005b) evaluated treatments including urine+dry soil, urine+dung, urine+compaction, and

dung+soil with different water-filled pore space during an incubation period of 103 d at 160C.

Nitrous oxide fluxes were measured 27 times during the 103 d. They reported a greater emission

of N20 for the urine+dung and urine+compaction treatments, while urine+dry soil had the lowest

emissions likely due to greater immediate penetration into the soil.

Ammonia volatilization is another important pathway of N emission from the grassland

and will be greatest at high temperatures and high soil pH (Boddey et al., 2004). Yokoyama et al.

(1991) found when dung beetles colonized dung pats NH3 volatilization ceased during the first

week; although, denitrification increased 23% and was greater than for the uncolonized dung

pats (12.2%). Denitrification increased due to supplies of available energy and N03-N and the

consumption of 02 by microorganisms.

It is reported that hippuric acid in urine has a controlling effect on both hydrolysis of urine

N and on NH3 volatilization, which may also have an affect on N20 emission factors (Doak,

1952; Whitehead et al., 1989; van Groenigen et al., 2005a). van Groenigen et al. (2005a) theorize









Cherney, Debbie J. R., Jerome H. Cherney, and Elena A. Mikhailova. 2002.
Orchardgrass and tall fescue utilization of nitrogen from dairy manure and
commercial fertilizer. Agron. J. 94:405-412.

Clough, Tim J., Francis M. Kelliher, Robert R. Sherlock and Colleen D. Ford. 2004. Lime and
soil moisture effects on nitrous oxide emissions from a urine patch. Soil Science Society
of America Journal. 68:1600-1609.

Coleman, S.W., J.E. Moore, and J.R. Wilson. 2004. Quality and utilization. p. 267-308.
In L.E. Moser, B.L. Burson, L.E. Sollenberger (eds.), Warm-season (C4) Grasses.
ASA-CSSA-SSSA, Madison, WI.

Correll, David L. 1996. Environmental impact of pasture systems on surface water quality. p.
13-28. In R.E. Joost and C.A. Roberts (ed.) Nutrient cycling in forage systems. Potash
and Phosphate Institute and The Foundation for Agronomic Research, Manhattan,
Kansas.

Crowder, L.V., and H.R. Chheda. 1982. p. 177. In Tropical grassland husbandry. Longman
Publishers, New York.

Cuttle, S.P., and P.C. Bourne. 1993. Uptake and leaching of nitrogen from artificial urine
applied to grassland on different dates during the growing season. Plant and Soil.
150:77-86.

Dai, Xiaobing. 2000. Impact of cattle dung deposition on the distribution pattern of
plant species in an alvar limestone grassland. Journal of Vegetation Science.
11:715-724.

Dalrymple, R.L., R. Stevens, T. Carroll, and B. Flatt. 1994. Forage production benefits
from nutrient recycling via beef cattle and how to manage for nutrient recycling in
a grazing cell. p. 269-273. In American Forage and Grassland Council Proc.
Lancaster, PA. 6-10 March 1994.

Day, T.A., and J.K. Detling. 1990. Grassland patch dynamics and herbivore grazing
preference following urine deposition. Ecology. 71:180-188.

de Klein, C.A.M. 2001. An analysis of environmental and economic implications of nil
and restricted grazing systems designed to reduce nitrate leaching from New
Zealand dairy farms. II. Pasture production and cost/benefit analysis. New
Zealand Journal of Agricultural Research. 44:217-235.

Decau, M.L., J.C. Simon, and A. Jacquet. 2003. Fate of urine nitrogen in three soils
throughout a grazing season. J. Environ. Qual. 32:1405-1413.









Nair et al. (2007) found that combined root systems of pine (Pinus elliotti) and bahiagrass

may absorb soil nutrients more completely than grasses alone in a treeless system. This

conclusion was based upon the reduction of WSP in tree-based systems. This superior absorption

resulted in greater P uptake in the silvopasture system and less loss of nutrients to surface water.

The silvopasture received 6 kg P ha-1 annually and was grazed; the treeless system was planted to

bahiagrass and received a single inorganic P application of 6 kg P ha-1 in 2003 and was grazed.

Soil samples were taken at a range of depths (0-5, 5-15, 15-30, 30-50, 50-75, and 75-100 cm).

The treeless pasture had a greater Melich-1 P as depth increased up to 50 cm. This could be due

to the silvopasture having a higher soil Al concentration which ranged from 350 to 1040 mg kg-1

compared to the treeless pasture (280-560 mg Al kg-1). This high concentration of Al will adsorb

P and form non-available compounds.

Phosphorus can move laterally from agricultural fields either in dissolved or particulate

(attached to soil particles) forms (Elliot et al., 2002). When P is lost through runoff it causes

negative environmental effects in surface water, such as eutrophication. The eutrophication of

freshwater by increased inputs of P from agricultural runoff is a concern in many areas of the

USA, particularly Florida (Sarkar and O'Connor, 2004). Arthington et al. (2003) found that

using three different stocking rates (1.49, 2.63, and 3.48 ha cow -1) of pregnant Brahman cows

did not affect concentrations or loads of total P or N in runoff from planted summer or mixed

winter pastures. They also found that control pastures, containing no cattle, provided similar

amounts of total P and N in runoff water compared to pastures containing cattle, although it

should be noted that these stocking rates are very low. The summer pastures consistently

delivered greater P loads (1.28 kg ha-1) in runoff than the winter pastures (0.03 kg ha-1) because









associated with greater IVDOM. This was true for Frequency 2 where both N concentration and

IVDOM decreased with increasing ring number and this was also the tendency for Frequency 3.

For frequencies 0 and 1, neither N concentration nor IVDOM decreased as ring number

increased. The departure from the expected trend occurred with Frequency 3, and in particular

Ring 3 within that frequency treatment (Fig. 4-5). No other papers were found in the literature

that reported digestibility responses to urine application.


600

595

590

585

580

575

0 570

565
5 Urine-3
4 Urine-2

555 --Urine-1
Urine-O
550
1 2 3 4
Ring


Figure 4-5. Ring number X urine application frequency interaction (P = 0.054) for in vitro
digestible organic matter (IVDOM) concentration during 2006. Ring number effects
were linear (P<0.0001) for two applications of urine and not significant for three (P >
0.144), one, and zero applications (P > 0.272). Standard error of a treatment mean
was 2.4 g kg-1. Ring Numbers 1 through 4 refer to sampling areas 0 to 15, 15 to 30,
30 to 45, and 45 to 60 cm from the center of the urine application, respectively.










van Groenigen, Jan Willem, Gerard L. Velthof, Frank J.E. van der Bolt, An Vos and
Peter J. Kuikman. 2005b. Season variation in N20 emissions from urine patches:
Effects of urine concentration, soil compaction and dung. Plant and Soil. 273:15-
27.

Wachendorf, Christine, Friedhelm Taube, and Michael Wachendorf 2005. Nitrogen
leaching from 15N labeled cow urine and dung applied to grassland on a sandy
soil. Nutrient Cycling in Agroecosystems. 73:89-100.

West, C.P., A.P. Mallarino, W.F. Wedin, and D.B. Marx. 1989. Spatial variability of soil
chemical properties in grazed pastures. Soil Sci. Soc. Am. J. 53:784-789.

White, R.E. 1988. Leaching. p. 193-211. In J.R. Wilson (ed.) Advances in nitrogen cycling in
agricultural ecosystems. CAB International, Wallingford, CT, USA.

Whitehead, D.C. 1970. The role of nitrogen in grassland productivity. Commonwealth
Agricultural Bureau, Aberystwyth, Wales, UK.

Whitehead, D.C. 1986. Sources and transformations of organic nitrogen in intensively
managed grassland soils. p. 47-58. In H.G. van der Meer, J.C. Ryden, and G.C. Ennik.
(ed.) Nitrogen fluxes in intensive grassland systems. Martinus Nijhoff, Dordrecht.
Wageningen, The Netherlands

Whitehead, D.C., D.R. Lockyer, N. Raistrick. 1989. Volatilization of ammonia from urea
applied to soil: influence of hippuric acid and other constituents of livestock urine. Soil
Biology & Biochemistry 21:803-808.

Wilkinson, S. R., and R.W. Lowrey. 1973. Cycling of mineral nutrients in pasture
ecosystems. p. 247-315. In G.W. Butler and R.W. Bailey (ed.) Chemistry and
biochemistry of herbage. Vol. 2. Academic Press, New York, NY.

Williams, C.H. 1950. Studies on soil phosphorus. II. The nature of native and residual
phosphorus in some South Australian soils. J. Agric. Sci., Camb. 40:243-256.

Williams, M.J., and A.C. Hammond. 1999. Rotational vs. continuous intensive stocking
management of bahiagrass pasture for cows and calves. Agron. J. 91:11-16.

Williams, P.H., and R.J. Haynes. 1995. Effect of sheep, deer, and cattle dung on herbage
production and soil nutrient content. Grass and Forage Science. 50:263-271.

Williams, P.H., and R.J. Haynes. 2000. Transformations and plant uptake of urine N and
S in long and short-term pastures. Nutrient Cycling in Agroecosystems. 56:109-
116.









in small areas where cattle congregate, e.g., near to shade, water, and supplemental feed sources

(Mathews et al., 2004; Sollenberger et al., 2002).

It has been suggested that rotational stocking with short grazing periods, i.e., many

paddocks per pasture, decreases the opportunity and tendency of animals to congregate in

lounging areas by intensifying competition for feed and shortening residency periods (Haynes

and Williams, 1993). In Florida, it was found that rotationally stocked pastures where grazing

periods were short (1 to 7 d) had greater spatial uniformity in time spent by cattle, excreta

deposition, and soil nutrient concentration than continuously stocked pastures (Dubeux, 2005).

Thus, it is likely that greater uniformity of excreta deposition can be achieved by imposing

rotational stocking with short grazing periods. Whether this intensification of grazing

management results in more efficient nutrient cycling may depend on the degree to which more

uniform excreta deposition enhances nutrient recovery by grassland plants and avoids excessive

nutrient accumulation in soils or nutrient loss to surface or ground water.

There is little information in the literature evaluating the impact of type and number of

excreta applications to pasture on leaching of N, changes in soil nutrient concentration, and

herbage growth and nutritive value. The objective of the reported research in this thesis is to

characterize the effects of cattle excreta application on 1) bahiagrass (Paspalum notatum Flugge)

herbage accumulation, chemical composition, and nutrient recovery and 2) nitrate leaching to

shallow ground water under a bahiagrass sod. Companion studies not reported in this document

will assess the effect of these factors on changes in soil nutrient concentration over time.

Bahiagrass was chosen for this research because it is the most widely used of the planted pasture

species in Florida.









medium N (MN; 40 and 10 g N m-2 in Years 1 and 2), or high N rate (HN: 80 and 20 g N m-2 in

Years 1 and 2) and were clipped every 10 (short interval; SI) or 21 d (long interval; LI) (Hirata,

1996). Root mass averaged across the 2-yr period was greatest with greatest N rate and longest

cutting interval (LN/LI: 595 g m-2; MN/LI: 597 g m-2; HN/LI: 613 g m-2). There also were effects

of N rate and cutting interval on leaf mass. Specifically, as N rate and cutting interval increased

so did leaf mass (LN/SI: 81 g m-2; LN/LI: 89 g m-2; HN/SI: 148 g m-2; HN/LI: 288 g m-2), but

leaf mass was relatively small compared to root mass.

Adaptation to Environments

Bahiagrass is grown throughout Florida due to its tolerance of a wide range of soil

conditions, including low fertility, drought, and short-term flooding (Gates et al., 2004), pH up to

6, and its ability to withstand close grazing (Burson and Watson, 1995; Williams and Hammond,

1999). Following its introduction to Florida, it spread to the Gulf Coast and Coastal Plains of the

Southeast USA and has become naturalized in these regions (Gates et al., 2004). Its area of

growth extends north to North Carolina and west to Texas and southeastern Oklahoma (Gates et

al., 2004).

Use in Production Systems

Approximately 75% of the 1.4 million ha of planted pasture in Florida are dominated by

bahiagrass which supports 1 million head of beef cows (Gates et al., 2004; Mislevy et al.,

2005). Bahiagrass is also used for hay, although in the Lower South, most hay is produced from

higher yielding warm-season perennial grasses such as bermudagrass [Cynodon dactylon (L.)

Pers.], dallisgrass (Paspalum dilatatum Poir.), and stargrass (Cynodon nlemfuensis Vanderyst)

(Robinson, 1996; Taliaferro et al., 2004). Bahiagrass is widely used for low maintenance turf,

especially in highway rights-of-way throughout the Southeast (Gates et al., 2004).









Summary and Conclusions

The objective of this research was to characterize the effects of cattle excreta type and

application frequency on 1) herbage yield, chemical composition, and nutrient recovery and 2)

nitrate leaching to shallow ground water under bahiagrass swards managed at two different

intensities. The study was carried out during the period from June 2006 through June 2007.

The High and Average management intensity treatments imposed on the pastures grazed

by livestock (excreta source pastures) had limited impact on herbage characteristics of grazed

bahiagrass. Herbage N concentration was greater for High than Average, but herbage mass, P

concentration, and IVDOM were not affected. There were sampling date effects on pasture

herbage mass with greatest mass occurring in July compared to June and September, and herbage

CP tended to be greater in July than other dates. Dung nutrient concentration was quite consistent

between treatments and across the three collection periods. Urine nutrient concentration was

similar between treatments in June (not analyzed statistically), but it was quite high for the

Average treatment relative to the High treatment in July and for High relative to Average in

September. Herbage characteristics did not vary seasonally to the same degree as urine chemical

composition, so the seasonal changes in urine composition, especially the high nutrient

concentrations in July, are not understood.

In the plots to which excreta was applied, there were interactions of excreta application

frequency X excreta type for all herbage responses except for excreta N recovery. Interaction

occurred because dung application had no effect on herbage responses, whereas responses to

urine were consistently significant. Greater herbage response to urine than dung was expected

because of the high proportion of dung N that is in an organic form and the greater availability to

plants of nutrients in urine. The general absence of response to dung was not expected and could

be attributed to a number of factors including physical interference of the dung pat, the high









CHAPTER 5
SUMMARY AND CONCLUSIONS

Pasture-based, forage-livestock systems in Florida are planted primarily to bahiagrass

(Paspalum notatum Flugge) because of its tolerance to close grazing and adaptation to a wide

range of soil conditions including low fertility, drought, and short-term flooding. Bahiagrass

pastures support approximately 1 million head of beef cattle in Florida. Even though these

pastures typically receive low amounts of fertilizer and are stocked at low to moderate rates, the

quantity of land area occupied and the number of animals supported statewide on bahiagrass

pasture make nutrient management a key issue. The importance of nutrient management also is

related to the sensitivity of Florida ecosystems, especially in terms of nutrient impacts on water

quality (Nair et al., 2007). On agricultural lands, the lateral flow of P to surface water,

particularly in Central and South Florida (Nair et al., 2007), and leaching of nitrates to ground

water, more commonly in North Florida, are critical concerns (Woodard et al., 2002).

One of the many challenges faced by livestock producers utilizing pasture as a feed

resource is avoidance of over accumulation of livestock wastes in certain areas of the pasture.

Uneven distribution of soil nutrients, due to non-uniform excreta deposition, is thought to lead to

leaching and volatilization of N and runoff of P and other nutrients from so-called nutrient hot

spots in pastures, yet the effect of in situ excreta deposition on plant, soil, and water responses

has received limited attention. The research reported in this thesis has as its objectives to

characterize the impact of frequency and type of excreta application on bahiagrass herbage dry

matter harvested, chemical composition, and excreta nutrient recovery, and to measure P and

NO3-N concentration in shallow soil water under excreta applications.

Treatments were two management intensities (Average and High), two types of excreta

(dung and urine), and three frequencies of excreta application (one, two, or three times per year).









BIOGRAPHICAL SKETCH

Una Renee White was born in Low Moor, Virginia, on 21 August 1982 and grew up in

Millboro, Virginia. She was raised on a beef farm, where she tagged along with her dad feeding

cattle. She moved away in 2003 to Raleigh, North Carolina, to attend North Carolina State

University, where she received her B.S. in Agronomy. Upon completing her M.S. she would

like to work for the government and then start her own consulting company. One day she hopes

to move back to Virginia and have a small farm.











3300


3100

2900

S2700

S2500

j 2300

S2100

S1900

1700


1500


Ring


Figure 4-3. Ring number by dung application frequency interaction (P < 0.027) on herbage DM
harvested during 2006. Ring number effects were quadratic for three (P = 0.025)
applications, linear (P < 0.001) and quadratic (P = 0.008) for two applications, and
not significant for one (P > 0.370) or for zero applications (P > 0.100). Standard error
of a treatment mean was 106 kg ha-1. Ring Numbers 1 through 3 refer to sampling
areas 0 to 15, 15 to 30, and 30 to 45 cm from the center of the dung application,
respectively.


-4-Dung-0
-- &Dung-1
-*-Dung-3
--I&Dung-2









N fertilizer level), so the response to excreta was consistent across levels of management

intensity. One reason for the lack of interaction in the current study may be that fertilizer N

levels applied were much less than those used by Deenen and Middelkoop (1992).

For the three control treatments to which no excreta was applied in the current study, the

DM harvested response to N application had linear and quadratic terms, increasing from to 1310

to 2450 to 3060 kg ha-1 as N fertilizer applied increased from 0 to 60 to 120 kg ha-1 (Table 3-7).

Responses to management intensity in both excreta-treated and no-excreta plots were primarily

due to the differences in N fertilization and are consistent with bahiagrass responses to N

fertilizer observed in the literature (Burton et al., 1997; Twidwell et al., 1998). Total-season DM

harvested in the current study was less than reported in many previous studies including those by

Burton et al. (1997), Twidwell et al. (1998), and Interrante et al. (2007). The two primary reasons

for low yields in the current research are rainfall and previous management at the plot site.

Rainfall totaled 722 mm during March through October 2006 compared to a 30-yr average of

962 mm (Table 3-8). In addition, the research site had not been fertilized for at least 12 yr, so

overall stand vigor was relatively low at the beginning of the trial and this affected first-year DM

harvested.

There was excreta type X application frequency interaction for bahiagrass DM harvested.

Interaction occurred because there was no effect of dung application frequency on DM harvested

(Fig. 3-2), but there were linear and quadratic effects of urine application frequency. The

response to urine increased from 2760 kg ha-1 at zero applications to 4670 kg ha-1 with three

applications. The quadratic effect was significant because the greatest increase in DM harvested

was with the first urine application, while additional urine applications had less effect on DM









Sumner, Sid, Wayne Wade, Jim Selph, Jerry Southwell, Vicky Hoge, Pat Hogue, Ed Jennings,
and Travis Seawright. 1992. Fertilization of established bahiagrass pasture in Florida.
Florida Cooperative Extension Service Circular 916.

Taliaferro, C.M., F.M. Rouquette, Jr., and P. Mislevy. 2004. Bermudagrass and stargrass. p.
417-438. In L.E. Moser, B.L. Burson and L.E. Sollenberger (ed.) Warm-season (C4)
grasses. ASA Monogr. 45. ASA, CSSA, SSSA, Madison, WI.

Tayler, J.C. and J.E. Rudman. 1966. The distribution of herbage at different heights in 'grazed'
and 'dung patch' areas of a sward under two methods of grazing management. J. Agric.
Sci. 66:29-39.

Thomas, R.J. 1992. The role of the legume in the nitrogen cycle of productive and sustainable
pastures. Grass For. Sci. 47:133-142.

Thomas, R.J., and N.M. Asakawa. 1993. Decomposition of leaf litter from tropical forage
grasses and legumes. Soil Biology and Biochemistry 25:1351-1361.

Thompson, R.B., and I.R.P Fillery. 1998. Fate of urea nitrogen in sheep urine applied to soil at
different times of the year in the pasture-wheat rotation in south Western Australia. Aust.
J. Agric. Res. 49:495-510.

Tiffany, M.E., L.R. McDowell, G.A. O'Connor, F.G. Martin, and N.S. Wilkinson. 1999.
Variation of forage and extractable soil minerals over two grazing seasons in
North Florida. Commun. Soil Sci. Plant Anal. 30:2743-2754.

Twidwell, E., W.D. Pitman, and G.J. Cuomo. 1998. Bahiagrass production and management.
Louisiana State Univ. Publication # 2697, Baton Rouge, LA.

Valk, H., and M.E.J. Hobbelink. 1992. Supplementation of grazing dairy cows to reduce
environmental pollution. In 14th General Meeting of the European Grassland
Federation. Lahti, Finland. European Grassland Federation. p. 400-405.

Vallis, I., L.A. Harper, V.R. Catchpoole, and K.L. Weier. 1982. Volatilization of
ammonia from urine patches in a subtropical pasture. Aust. J. Agric. Res. 33:97-
107.

van der Meer, H.G., and M.G. Van Uum-Van Lohuyzen. 1986. The relationship between
inputs and outputs of nitrogen in intensive grassland systems. p. 1-15. In H.G. van der
Meer, J.C. Ryden, and G.C. Ennik. (ed.), Nitrogen fluxes in intensive grassland systems.
Martinus Nijhoff, Dordrecht. Wageningen, The Netherlands.

van Groenigen, Jan Willem, Peter J. Kuikman, Willy J.M. de Groot, and Gerard L.
Velthof. 2005a. Nitrous oxide emission from urine-treated soil as influenced by
urine composition and soil physical conditions. Soil Biology & Biochemistry
37:463-473.









lounging areas by intensifying competition for feed and shortening residency periods (Haynes

and Williams, 1993). In Florida, it was found that rotationally stocked pastures where grazing

periods were short (1 to 7 d) had greater spatial uniformity in time spent by cattle, excreta

deposition, and soil nutrient concentration than continuously stocked pastures (Dubeux, 2005).

Thus, it is likely that greater uniformity of excreta deposition can be achieved by imposing

rotational stocking with short grazing periods. Whether this intensification of grazing

management results in more efficient nutrient cycling may depend on the degree to which more

uniform excreta deposition enhances nutrient recovery by grassland plants and avoids excessive

nutrient accumulation in soils or nutrient loss to surface or ground water.

There is little information in the literature evaluating the impact of type and number of

excreta applications to pasture on leaching of N, recovery of nutrients in harvested herbage, and

changes in herbage growth and nutritive value. The objective of this research was to characterize

the effects of cattle excreta application on 1) bahiagrass yield, chemical composition, and

nutrient recovery and 2) nitrate leaching to shallow ground water.

Materials and Methods

Experimental Sites

There were two sites used for this study. One location served as the source for excreta and

the other site was used for excreta application to bahiagrass. Excreta source pastures were

located at the University of Florida Beef Research Unit, northeast of Gainesville, FL (29.72 N

latitude, 82.350 W longitude). This site was chosen because well-established 'Pensacola'

bahiagrass pastures, fencing, and animals were readily available. Soils at this site were classified

as Spodosols of the Pomona (sandy, siliceous, hyperthermic Ultic Alaquods) and Smyrna series

(sandy, siliceous, hyperthermic Aeric Alaquods). Average soil pH was 5.5, and Mehlich-I

extractable P, K, Mg, and Ca were 17, 29, 50, and 392 mg kg1, respectively.









A-3 P values for main effects, interactions, and polynomial contrasts for ring number for
urine-treated plots as discussed in Chapter 4 ............................................................. 123









During the June excreta collection, samples were taken to ensure that chemical

composition did not change markedly during the 8 d of excreta storage prior to application, i.e.,

to ascertain that what was applied to the plots was similar in composition to what the animal

would deposit fresh. This was accomplished by taking two subsamples from each of two separate

urine and dung events. One of the two subsamples from each event was analyzed immediately;

the other subsamples were stored in the same manner as the urine and dung that were eventually

applied to the plots. At the date of excreta application to plots, these stored subsamples were

analyzed to assess changes in chemical composition during the storage period (Table 3-1).

Urine samples for analysis were acidified with concentrated sulfuric acid to pH 2 to avoid

N volatilization. Dung was analyzed for total N, organic N, NH4-N, P, and K, and urine was

analyzed for N, NH4-N, urea-N, P, and K. Dung total N was analyzed using the Kjeldahl

method- AOAC 984.13; NH4-N analyzed using Distillation AOAC 941.04; organic N by

difference (Total N Ammonia-N); and P and K were analyzed using Thermo IRIS Advantage

HX Inductively Coupled Plasma Radial Spectrometer. Urine total N was analyzed using AOAC

2001.11 Block digestion and Foss 2300 or 2400 Analyzer; urea and ammonia AOAC 941.04,

Analysis by Thermo IRIS Advantage HX Inductively Coupled Plasma (ICP) Radial

Spectrometer; and for P and K by Thermo IRIS Advantage HX Inductively Coupled Plasma

(ICP) Radial Spectrometer.









Table 3-3. Composition of fresh dung from Average and High management intensity source
treatments during three collection periods in 2006. Each value reported is the average
across three subsamples from a composite dung sample. Dung was composite across
replicates within a management intensity treatment, so statistical comparisons of
treatment effects are not possible.
Dung Application Date
Nutrient 20 June 1 August 15 September
Average High Average High Average High
g kg-1 fresh dung
Total N 1.97 2.10 2.03 2.13 2.25 1.85
NH3-N 0.07 0.08 0.08 0.10 0.09 0.08
Organic N 1.89 2.02 1.95 2.03 2.16 1.77
Total P 0.64 0.74 0.71 0.67 0.72 0.57
Total K 1.37 1.49 1.32 1.32 1.46 1.43
tAverage refers to pastures fertilized with 60 kg N ha- yrf and stocked at 2 yearling heifers
ha-1, while High refers to pastures fertilized with 120 kg N ha-1 yr1 and stocked at two yearling
heifers ha-1.

Table 3-4. Composition of urine from Average and High management intensity source
treatments during three collection periods of 2006. Each value reported is the average
across three subsamples. Urine was composite across replicates within a
management intensity treatment, so statistical comparisons of treatment effects are
not possible.
Urine Application Date
Nutrient 20 June 1 August 15 September
Average High Average High Average High
g kg-1 urine
Total N 1.60 1.58 4.04 2.97 0.98 1.88
NH3-N 0.00 0.00 0.00 0.00 0.00 0.00
Urea-N 0.84 0.88 1.82 1.71 0.37 0.86
Total P 0.0033 0.0042 0.0049 0.0047 0.0027 0.0049
Total K 2.20 2.39 3.77 2.76 1.38 2.27
t Average refers to pastures fertilized with 60 kg N ha-1 yr1 and stocked at 2 yearling heifers
ha-l, while High refers to pastures fertilized with 120 kg N ha-l yr1 and stocked at two yearling
heifers ha-1.

Nutrients Applied in Excreta

Quantity of N, P, and K ha-1 applied at each of three application frequencies for Average

and High management intensities are shown in Tables 3-5 (dung) and 3-6 (urine). Amount of

nutrient applied per unit land area is greater for dung than urine because of the smaller land area

to which dung was applied.









harvested. For each level of excreta application frequency from one through three, urine-treated

plots outyielded dung-treated plots (P < 0.003; Fig. 3-2).

Day and Detling (1990) applied simulated urine to a mixture of little bluestem

[Schizachyrium scoparium (Michx.) Nash] and 'Kentucky' bluegrass (Poapratensis L.). Urine-

treated areas outyielded the no-urine control 3870 to 2720 kg ha-1. Norman and Green (1958)

applied urine to a mixture of cool-season grasses, legumes, and forbs in spring. One month after

application, herbage DM harvested was 1870 kg ha-1 for treated plots compared to 1160 kg ha-1

for the untreated control and 3 mo after application the difference was 2050 vs. 1160 kg ha-1,

respectively. Thus, urine application is associated with yield increases that can carry over for at

least several months.

One reason for lack of response to dung application was the interference effect on herbage

under the pat. Reduction in DM harvested under the dung pat (circle of 15-cm radius) compared

with the ring extending 15 cm beyond the edge of the dung pat was greater than 500 kg ha-l

averaged across application frequencies (Chapter 4). Hirata et al. (1991) described interference

of dung pats on herbage beneath them or even death of herbage under the pat, resulting in

reduced yields (Hirata et al., 1991). MacDiarmid and Watkin (1971) reported that 75% of grass

tillers and rooted nodes of clover (Trifolium spp.) stolons under the dung pat were dead within 15

d of application. This resulted in a significant reduction in yield from the area of the pat. Another

reason for the lack of a positive yield response in the dung-treated plots is the high proportion of

organic N in dung relative to the high proportion of total urine N that is urea N (Table 3-3 and 3-

4). Norman and Green (1958) suggested that chemical differences between dung and urine

accounted for lesser response to dung. Additionally in the current study, there was relatively little

dung beetle (Scarabaeidae) activity apparent as well as drier than normal weather, both of which









(L.) Moench] were planted. Plant dry matter yield and N uptake were more affected by soil type

than feces type (diets fed to animals producing feces for the study were corn silage-low CP; corn

silage-high CP; alfalfa-low CP; alfalfa-high CP; Powell et al., 2006). Soil type accounted for 67,

65, and 82% of the variability in N uptake by oat and sorghum; feces from the low CP diets

decreased oat dry matter production likely due to N immobilization (Powell et al., 2006).

Pearl millet was grown in pots and the soil was amended with fecal material (with

fertilizer N or without fertilizer N) or with leaves. Dry matter yield was measured from 60 to 240

d after amendment application. Pearl millet amended with feces accumulated about 37% more P

than when amended with leaves (Acacia trachycarpa, Combretum glutinosum, Guiera

senegalensis, Pterocarpus erinaceus, Pennisetum glaucum, and Vigna unguiculata), perhaps due

to immobilization of P in the leaves. Feces without fertilizer N did not affect pearl millet dry

matter yield or N and P uptake, while feces with fertilizer N resulted in a yield response likely

due to enhanced N mineralization (Powell et al., 1999).

Cherney et al. (2002) applied three different nutrient treatments (NT) to orchardgrass

(Dactylis glomerata L.) and tall fescue (Festuca arundinacea Schreb.). Treatments were

inorganic N fertilization at 196 kg ha-1 (NT 1) or dairy cow manure which was applied at two

rates (X and 2X and termed NT2 and NT3) in 1995, 1996, and 1997. Rates of N applied in

manure were 123 and 246 kg N ha1 in 1995, 134 and 268 kg N ha-1 in 1996, and 143 and 286 kg

N ha-1 in 1997. The relative N recovery in 1995 through 1997 for orchardgrass was 72, 58, and

45%, respectively, for NT 1, NT2, and NT3. Tall fescue had a relative N recovery of 76% for NT

1, 58% for NT 2, and 48% for NT 3. They concluded that perennial grasses can utilize large

quantities of applied N, even N from dung.









Rhoads et al. (1997) found similar results but with different rates of P (0, 84, and 168 kg ha-').

In this study, they found that increasing P fertilizer from 0 to 84 kg ha-1 increased yield from 9.8

to 10.7 Mg DM ha-1 in the first year and 7.3 to 8.8 Mg DM ha-1 in the second year.

Herbage Nutritive Value

There is extensive literature on the effects of N fertilization on the chemical composition

and digestibility of Pensacola bahiagrass herbage (Beaty et al., 1975; Burton et al., 1997;

Twidwell et al., 1998; Newman et al., 2006; Stewart et al., 2007). Similar to many other C4

grasses, Pensacola bahiagrass crude protein (CP) increases with increasing N fertilization, while

the response of herbage in vitro digestible organic matter (IVDOM) is less clear.

Beaty et al. (1975) showed that Pensacola bahiagrass N concentration increased

significantly when N fertilization increased from 0 to 224 and again from 224 to 672 kg N ha-1.

Newman et al. (2006) fertilized bahiagrass with two levels of N fertilizer, 80 and 320 kg N ha-1,

and harvested them by clipping to a 5-cm stubble height every 7 wk. Bahiagrass fertilized with

320 kg N ha-1 had greater IVDOM than bahiagrass fertilized with 80 kg N ha-1 (471 and 432 g

kg- respectively). Herbage CP was 79 and 58 g kg-1 for the 320 and the 80 kg N ha-1 treatments,

respectively. Burton et al. (1997) found bahiagrass N concentrations increased from 10.6 to 17 g

N kg-1 as fertilizer N applications increased from 56 to 448 kg N ha-1. Stewart et al. (2007) found

that bahiagrass herbage CP increased as amount of N fertilizer increased from 40 to 120 kg N

ha-1 (99 to 113 g CP kg-1), and from 120 to 360 kg N ha-1 (113 to 140 g CP kg-1).

In Florida, bahiagrass was fertilized with 112 kg N, 30 kg P, and 52 kg K ha-1 yr-1, and this

resulted in average CP and IVDOM concentrations from April through December of 142 and 563

g kg-1, respectively, in Year 1, 140 and 510 g kg-1 in Year 2, and 163 and 485 g kg-1 in Year 3

(Mislevy et al., 2005). These authors characterized seasonal changes in bahiagrass nutritive

value. Pensacola's greatest CP (186 g kg-1) occurred in April 1998 and then decreased from June









Herbage Nitrogen Concentration

Herbage N concentration was affected by management intensity (P = 0.039), application

frequency (P < 0.001), excreta type (P < 0.001), and application frequency X excreta type

interaction (P < 0.001) (Table A-i). High management intensity plots had greater N

concentration than Average intensity (14.7 and 14.2 g kg-1, respectively). For control plots, N

concentration ranged only from 12.7 to 13.9 g kg-1 and was not affected by amount of N fertilizer

applied (Table 3-7). Pensacola bahiagrass herbage N concentration increased from 9.9 to 13.8 g

kg-1 when N fertilizer amount increased from 0 to 224 kg ha-1 (Beaty et al., 1975). In Louisiana,

Pensacola bahiagrass N concentration increased from 16.8 to 23.0 g kg-1 as N fertilizer amount

increased from 0 to 450 kg ha-1 (Twidwell et al., 1998), while in Georgia it increased from 10.6

to 16.8 g kg-1 as N fertilizer increased from 56 to 450 kg ha-1 (Burton et al., 1997). Thus,

bahiagrass is responsive to N fertilizer, but the range in N applied was relatively small in the

current study.

There was excreta type X application frequency interaction for herbage N concentration.

Interaction occurred because there was no effect of frequency of dung application on N

concentration (P > 0.246), but there were linear (P < 0.001) and cubic effects (P < 0.001) for

response to urine application frequency (Fig. 3-3). The linear effect was associated with an

increase from 13.3 to 16.2 g kg-1 as application frequency increased from zero to three. The cubic

effect occurred because there was little change in N concentration between zero and one

application, a large increase between one and two applications, and little change between two

and three applications. Urine-treated plots had greater herbage N concentration than dung-treated

plots for Application Frequencies 2 and 3 (P < 0.001), but there was no excreta type effect for a

single application (P = 0.495; Fig. 3-3).










Table A-2. P values for main effects, interactions, and polynomial contrasts for ring number for dung-treated plots as discussed in
Chapter 4.

Excreta Response Source of Variation
Type variable Management Application Ring # (R) MI x AF MI x R AF x R MI x AF x R
Intensity (MI) Frequency (AF)
Dung DM Harvested <0.0001 0.1247 L-0.0918 0.3522 0.9914 0.0271 0.8681
Q-0.0009
Herbage N <0.0001 0.2144 L-0.5948 0.1771 0.7763 0.6869 0.7432
Q-0.1608
Herbage P 0.1498 0.9932 L-0.6190 0.4429 0.3765 0.8761 0.6519
Q-0.4989
Herbage IVDOM 0.0005 0.1888 L-0.0108 0.7276 0.7383 0.2491 0.9789
Q-0.4093
N Harvested <0.0001 0.2143 L-0.0328 0.5474 0.9210 0.0194 0.8024
Q-0.0009
P Harvested <0.0001 0.1845 L-0.1447 0.5862 0.5769 0.0584 0.4858
Q-0.0030
L = linear and Q = quadratic effect of ring number.









Total Phosphorus Harvested

Total-season P harvested was affected by main effects of management intensity,

application frequency, and excreta type (P < 0.001) and the interactions of intensity X excreta

type (P = 0.033) and application frequency X excreta type (P < 0.001) (Table A-i). The

management intensity X excreta type interaction occurred because for dung-treated plots the

magnitude of the advantage of High over Average management intensity (10.9 vs. 8.0 kg ha-1; P

< 0.001) was greater than the advantage in urine-treated plots (13.7 vs. 12.4 kg ha-l; P = 0.054).

Phosphorus harvested increased linearly from 5.0 to 11.1 as N fertilizer increased from 0 to 120

kg ha-1 for the no-excreta controls (P = 0.002) (Table 3-7).

Interaction of application frequency X excreta type occurred because there was no effect of

dung application frequency on P harvested (P > 0.570), but there were linear (P < 0.001) and

quadratic effects (P = 0.052) of urine application frequency on the response (Fig 3-7).

Phosphorus harvested increased from 9 to 18 kg ha-1 as urine application frequency increased

from zero to three. Herbage P harvested was greater for urine- than dung-treated plots for

Application Frequencies 2 (P = 0.007) and 3 (P < 0.001) and tended to be greater for Frequency

1 (P = 0.115).

Newman et al. (2005) harvested bahiagrass hay that was treated with different levels of N

fertilizer (0, 50, 67, and 100 kg N ha-1 harvest-'). Phosphorus harvested was 32, 39, 53, and 55 kg

ha-1 for these four treatments, respectively. The response was primarily yield driven as DM

harvested increased from 9.7 to 20.5 Mg ha-1, but herbage P concentration varied only from 3.7

(at the zero N rate) to 3.0 g kg-1 (at the 100 N rate). The values reported in the current study are

much lower than those reported by Newman et al. (2005) due to lower N rates in some cases and

likely greater N losses due to volatilization for others, especially multiple applications of urine.

There are few studies in the literature that measure tissue P response or P removal under excreta









A long-term, ungrazed stand of Pensacola bahiagrass at the Plant Science Research Unit,

Citra, FL (29.41N latitude, 82.02 W longitude) was used for application of excreta. The

ungrazed area was used for this purpose so that amount and frequency of excreta applications

could be controlled. Also, this location was chosen because the soils are well-drained sands,

avoiding potential subsurface lateral flow of water and nutrients that could occur under the

Spodosols at the Beef Research Unit. This could result in mixing of nutrients leaching from

adjoining plots and preclude the drawing of meaningful conclusions regarding water quality from

the treatments imposed. Also, this site had not been fertilized for the past 12 yr, so there was

minimal carryover effect of previous treatments. The soils at the application site are classified as

Tavares or Candler fine sands (sandy hyperthermic, uncoated Typic Quartzipsamments).

Average soil pH was 5.5, and Mehlich-I extractable P, K, Mg, and Ca were 34, 70, 18, and 164

mg kg-1, respectively.

Treatments and Design

Treatments were two excreta sources (from here forward referred to as Average or High

pasture management intensity), two types of excreta (dung and urine), and three frequencies of

excreta application (one, two, or three times per year). Average and High management intensity

were defined based on N fertilizer amount and stocking rate. Average management pastures

received 60 kg N ha-1 yr1 and were stocked with two yearling heifers ha-1, and High pastures

received 120 kg N ha-yr-1 and were stocked with four yearling heifers ha-1. These two

management intensities were selected to represent fertilization and stocking regimes that are

common in the Florida livestock industry (Average) or represent approximately the most

intensive management applied to grazed bahiagrass (High). In addition, based on previous work

by Stewart et al. (2007) it is expected that these treatments will result in forage that varies in









In spite of the magnitude of the response to urine, significant effects did not extend large

distances beyond the edge of the urine application. For DM harvested and herbage P harvested,

these effects extended only 15 cm beyond the edge of the area to which urine was applied, while

this distance was 30 cm for herbage N concentration and N harvested. In contrast, herbage DM

harvested and N and P concentrations in areas outside of Ring 1 were similar for plots treated

with dung and the no dung control. Thus, unlike urine there was no measurable effect of dung

beyond the immediate area to which it was applied. Lack of positive response to dung was due to

physical interference, as already mentioned, but it was accentuated by limited apparent dung

beetle activity, the drier than normal weather that may have reduced breakdown of dung, and the

high proportion of nutrients in dung that are in an organic form and only slowly released for

plant growth.

Data from both chapters provide justification for management practices that increase

uniformity of excreta deposition on pastures. Responses to single applications of urine are often

significant, especially in terms of DM harvested, and additional applications have lesser impact.

Single applications of dung and urine result in a greater percentage N recovery than multiple

applications. In addition, by avoiding repeated dung applications in the same area, there is

opportunity to avoid much of the negative impact of physical interference. Thus management

practices that increase the uniformity of excreta distribution should also increase herbage

accumulation on pastures and increase the efficiency of recapture by plants of nutrients in

excreta. Dubeux (2005) and Dubeux et al. (2006) provide evidence that rotational stocking,

especially rotational stocking with short grazing periods, increases uniformity of excreta

distribution. Though rotational stocking is not widely used at present by Florida livestock

producers, rapidly increasing fertilizer costs and concerns about the impact of cattle excreta on









Excreta was collected from animals grazing bahiagrass pasture managed at Average and High

intensities and applied to ungrazed bahiagrass plots fertilized with the same amount of N as the

excreta source pasture from which the excreta was obtained. Average and High management

intensities imposed on the excreta source pastures were defined based on N fertilizer amount and

stocking rate. Average management intensity source pastures received 60 kg N ha-1 yr1 and were

stocked with two yearling heifers ha-l, and High pastures received 120 kg N ha-1 yr1 and were

stocked with four yearling heifers ha-1. These two management intensities were selected to

represent fertilization and stocking regimes that are common in the Florida livestock industry

(Average) or represent approximately the most intensive management applied to grazed

bahiagrass (High) in Florida. Excreta from cattle grazing Average management intensity pastures

was applied to bahiagrass plots that received 60 kg N ha-1 yr and excreta from cattle grazing

High management intensity pastures was applied to plots that received 120 kg N ha-1 yrf.

Data were reported in two chapters. Chapter 3 describes shallow soil water responses and

those of herbage within a 45-cm radius of the excreta application. In Chapter 4, the spatial

patterns of herbage responses, beginning at the center of excreta application and moving away

from the application, were described.

For herbage responses in the circle of 45-cm radius surrounding the excreta application

(Chapter 3), the primary treatments affecting the responses were management intensity and the

excreta application frequency X excreta type interaction. For plots receiving excreta, as

management intensity increased from Average to High, forage DM harvested increased from

3230 to 3850 kg ha-l. High management intensity plots had greater N concentration than

Average intensity (14.7 and 14.2 g kg-1, respectively), greater IVDOM than Average plots (577

vs. 567 g kg-1), greater N harvested than Average (57 vs. 47 kg ha-1), and greater P harvested









Table 3.1. Chemical composition of excreta analyzed immediately after collection (fresh) and
after storage (stored) for up to 8 d at 40C. Each value is the mean of the analysis of two
subsamples.
Urine Dung
Constituent Fresh Stored Fresh Stored
------------------ ---------- g kg--------------
Total N 0.41 0.49 2.13 2.07
NH3-N 0 0 0.075 0.070
Urea 0.29 0.26 t t
Organic N t t 2.05 2.00
Total P 0.0024 0.0029 0.75 0.72
Total K 0.90 0.90 1.42 1.37
Not analyzed for this constituent within this excreta type.

Plot Management

All ungrazed plots to which excreta would subsequently be applied were fertilized on 7

June with 18 kg P and 66 kg K ha-1. Nitrogen application depended upon excreta-source

treatment. Plots that received excreta from the High management intensity source pastures were

fertilized with 120 kg N ha1 yr-1, split equally in two applications of 60 kg N ha-1 (7 June and 16

Aug. 2006). Plots that received excreta from the Average management intensity plots were

fertilized with 60 kg N ha-1 yr- in two applications of 30 kg N ha-1 (on the same dates as High).

Following fertilization on 7 June, the entire plot area was irrigated with 25 mm of water because

of spring drought and lack of early summer rainfall (rainfall in March through May at this

location was 68 mm compared to a 30-yr average of 279 mm). Magnesium sulfate was applied to

all plots on 6 July 2006 to address low soil Mg levels. It was applied at 135 kg ha-1 to provide 27

kg Mg ha- and 36 kg S ha-.

Excreta Application

All plots were staged to a 5-cm stubble on 19 June 2006. All experimental units except the

no-excreta controls received dung or urine on 20 June. Subsequent applications were on 1

August and 13 September. Only plots receiving two or three excreta applications yr-i received









samples and 20 disk heights were taken at each sampling date. Herbage mass was predicted from

disk heights, and a single calibration equation was used across all pastures and the three

sampling dates. The equation was herbage mass (kg ha1) = 340 (disk height [in cm]) 137 (r2 =

0.821). At the same sampling times, hand-plucked samples of the top 10 cm of the forage canopy

were taken to represent cattle diets. Samples were collected from 30 representative locations per

pasture, composite, dried at 600C, and analyzed for N and P concentration using a micro-

Kjeldahl technique (Gallaher et al., 1975) and for in vitro digestibility using a two-stage

procedure (Moore and Mott, 1974).

Excreta Collection and Analyses

The yearling heifers used for grazing the excreta source pastures were selected for docile

temperament following careful screening of a large group of animals of similar age and weight.

Initially 12 heifers were chosen. During the winter and spring before the start of the trial, they

were fed hay and small amounts of concentrate several times each week to develop a routine of

close proximity to people. Based on their response to this interaction, the group was reduced to

eight heifers, the number needed for the grazing study.

Stanchions were erected in a corral near the pastures, and animals were moved to the corral

several times per week during spring before the start of the trial. Concentrate feed was placed in

troughs in front of the stanchions so that heifers were required to put their heads into the

stanchions to gain access to the feed. After several weeks of this process, the animals were

locked in the stanchions to become accustomed to short periods of restraint. This activity was

suspended at least a week before dung and urine collection began.

During the trial, animals were brought to the stanchions for collection of urine and dung.

Urine and dung were collected during three periods (12-16 June; 26-28 July; and 7-8 Sept. 2006)

leading up to the three dates (20 June, 1 Aug., and 13 Sept. 2006) when excreta was applied to









concentration of organic N in dung as a proportion of total N, limited apparent activity of dung

beetles, and a drier than normal year leading to rapid drying and crusting of dung.

High management intensity applied to the ungrazed bahiagrass plots increased herbage

response over Average for all response variables except P concentration and N recovery. With

the exception of excreta type X management intensity interaction for P harvested, there were no

interactions of other treatments with management intensity. This indicates that both excreta types

and the three excreta application frequencies evaluated in this study had similar effects on

response variables across the range of management intensities tested.

Excreta N recovery was greater for urine than dung averaging 22 and 2%, respectively.

Recovery was also affected by excreta application frequency, decreasing from 28 to 18% as

urine application frequency increased from one to three yr-, and from 4 to less than 1% as dung

application frequency increased from one to three. These recoveries are in the lower part of the

range reported in the literature, but it does not appear that leaching losses explain this response.

Significant responses of shallow soil water N03-N concentrations occurred at only one sampling

date (15 Sept. 2006), and these values were less than 0.15 mg L-1. Greater concentrations

occurred in individual wells at two other dates in August, but these were not consistent across

replicates within a treatment. Warm, dry weather may have increased volatilization losses of

urine N (Russelle, 1996), and bahiagrass has been reported to store large quantities of N in

rhizomes and roots (Blue, 1973). Low values for dung recovery and absence of herbage response

to dung suggest limited mineralization of nutrients in dung during the course of the growing

season. Greater percent recovery of nutrients in excreta occurred with single excreta applications

suggesting that grazing management practices which increase uniformity of excreta deposition

will likely increase efficiency of nutrient cycling in grazed grasslands.



































Figure 3-1. This drawing (not to scale) shows one plot or experimental unit and the areas to
which urine and dung were applied. In plots where urine was applied, 1 L was applied
to the area labeled 30 (inside a 30-cm diameter circle) and 1 L to the area labeled 60
(the area outside the 30-cm diameter circle but inside the 60-cm diameter circle). In
plots where dung was applied, 2 kg fresh weight of dung was applied to the area
labeled 30.

Forage Harvest and Herbage Analyses

Forage was harvested every 28 d following initial excreta application (20 June), and

harvest dates were 18 July, 15 Aug., 12 Sept., and 10 Oct. 2006. To characterize the effect of the

excreta application on herbage yield and nutritive value, herbage was clipped at each harvest

date to a 5-cm stubble in a 90-cm diameter circle (0.636-m2 area), the center of which was the

center of the circles of dung or urine application. Harvested herbage was dried at 600C to a

constant weight and weighed. It was subsequently ground to pass 1-mm screen and analyzed for









Jaramillo and Detling (1992) reported that herbage N concentration in urine-affected areas

of western wheatgrass [Pascopyrum smithii (Rydb. A. Love)] was greater than in control

patches. Herbage N concentration averaged 30 and 16 g kg-1 for affected and unaffected areas.

Ledgard et al. (1982) found similar effects of urine on N concentration in a perennial ryegrass

(Lolium perenne L.)-white clover (Trifolium repens L.) mixture. Ryegrass herbage N

concentration in urine patches was 47 g kg-1 compared to 30 g kg-1 for unaffected areas.

The range in N concentration in dung-treated plots was only 13.2 to 13.6 g kg-1 (P >

0.246). Lack of N concentration response to dung application frequency can be attributed to

reasons similar to those for absence of a DM harvested response to dung. Others have found

some impact of dung on herbage N. Perennial ryegrass herbage from dung-affected areas had N

concentration of 12.3 compared to 10.8 g kg-1 for unaffected herbage (Jorgensen and Jensen,

1997). Similarly, Greenhalgh and Reid (1968) stated that N concentration of perennial ryegrass

herbage near a dung deposit was greater than similar herbage growing in harvested fields (32 vs.

27 g kg- respectively).









Gates, Roger N., Camilo L. Quarin, and Carlos G.S. Pedreira. 2004. Bahiagrass. p. 651-675. In
L.E. Moser et al. (ed.) Warm-season (C4) Grasses. Agronomy Monograph no. 45. ASA,
CSSA, SSSA. Madison, WI.

Graetz, D.A., V.D. Nair, K.M. Portier, and R.L Voss. 1999. Phosphorus accumulation in
manure-impacted Spodosols of Florida. Agriculture, Ecosystem and
Environment. 75:31-40.

Greenhalgh, J.F.D., and G.W. Reid. 1968. The effects of grazing intensity on herbage
consumption and animal production. III. Dairy cows grazed at two intensities on
clean or contaminated pasture. Journal of Agric. Sci., Camb. 71:111-222.

Griffin, T.S., C.T. Honeycutt, and Z. He. 2003. Changes in soil phosphorus from
manure application. Soil Sci. Soc. Am. J. 67:645-653.

Griffith, W.K., and L.S. Murphy. 1996. Macronutrients in forage production. p. 13-28. In R.E.
Joost and C.A. Roberts (ed.) Nutrient cycling in forage systems. Potash and Phosphate
Institute and The Foundation for Agronomic Research, Manhattan, Kansas.

Hack-ten Broeke, M.J.D., W.J.M. De Groot and J.P. Dijkstra. 1996. Impact of excreted
nitrogen by grazing cattle on nitrate leaching. Soil Use and Management 12:190-198.

Hambleton, L.G. 1977. Semiautomated method for simultaneous determination of
phosphorus, calcium and crude protein in animal feeds. J.A.O.A.C. 60:845-852.

Haynes, R.J., and P.H. Williams. 1992. Changes in soil solution composition and pH in
urine-affected areas of pasture. J. Soil Sci. 43:323-334.

Haynes, R.J., and P.H. Williams. 1993. Nutrient cycling and soil fertility in the grazed
pasture ecosystem. Adv. Agron. 49:119-199.

Henzell, E.F., and Ross, P.J. 1973. The nitrogen cycle of pasture ecosystems. p. 227-246. In
O.W. Butler and R.W. Bailey (ed.), Chemistry and biochemistry of herbage. Academic
Press, London, UK.

Hirata, Masahiko. 1994. Response of bahiagrass (Paspalum notatum Flugge) sward to
nitrogen fertilization rate and cutting interval. J. Japan Grassl. Sci. 40:313-324.

Hirata, M., Y. Ogawa, N. Koyama, K. Shindo, Y. Sugimoto, M. Higashiyama, S. Ogura,
and K. 2006. Fukuyama. Productivity of bahiagrass pastures in South-western
Japan: Synthesis of data from grazing trials. J. Agronomy & Crop Science.
192:79-91.

Hirata, Masahiko, Yasuhiro Sugimoto, and Masahiko Ueno. 1990. Quantification of
cattle dung-related processes in bahiagrass pasture. JARQ. 23:189-195.
















20




A
C-
o15













0
I


5 ---Urine-3
-- Urine-2
-~ Urine-1
U Urine-0

1 2 3 4
Ring


Figure 4-8. Ring number X urine application frequency interaction (P < 0.001) for bahiagrass
herbage P harvested during 2006. Ring number effects were linear (P < 0.001) and
cubic (P = 0.0419) for three urine applications, linear for two (P < 0.001) and one
application (P = 0.019), and not significant (P > 0.194) for zero applications.
Standard error of a treatment mean was 0.4 kg ha-1. Ring Numbers 1 through 4 refer
to sampling areas 0 to 15, 15 to 30, 30 to 45, and 45 to 60 cm from the center of the
urine application, respectively.









Excreta Application

Plots were staged to a 5-cm stubble height on 19 June 2006. All experimental units except

the no-excreta controls received dung or urine on 20 June. Subsequent applications were on 1

August and 13 September. Only plots receiving two or three excreta applications yr-1 received

excreta on 1 August, while on 13 September only those plots receiving three applications yr

were treated.

Quantity of dung and urine and the area to which they were applied were determined based

on values reported for cattle in the literature (Haynes and Williams, 1993) and the author's

personal observation of the cattle used in this study. Haynes and Williams (1993) indicated that

surface area ranges from 0.16 to 0.49 m2 for urine and 0.05 to 0.09 m2 for dung. Urine volume is

said to range between 1.6 and 2.2 L and fresh dung mass from 1.5 to 2.7 kg (Haynes and

Williams, 1993).

In the current study, 2 L of urine constituted one application, and it was applied to an area

of 0.283 m2, a 30-cm radius circle, with the center being the center of the plot (Fig. 4-1). Of the 2

L of urine, 1 L was applied to a 15-cm radius circle with the center being the center of the plot

(area of 0.071 m2), and 1 L was applied to the area outside the 15-cm radius circle but inside a

circle of 30-cm radius. This was done to reflect the likelihood of greater concentration of urine

closer to the center of the affected area in a natural urine deposit. The urine was applied using a

sprinkler head on a watering can, and rate of application was controlled so that runoff from the

application area was minimal. Dung was applied at 2 kg fresh weight to an area of 0.071 m2,

based on the range in surface area for dung applications (0.05 to 0.09 m2) in the review by

Haynes and Williams (1993). This was a circle of 15-cm radius with the midpoint being the

center of the plot (Fig. 4-1). The dung was evenly distributed across this area. Subsequent









Characteristics of Cattle Excreta

A large proportion of the nutrients consumed by grazing livestock are returned to the

pasture in animal excreta (Sollenberger et al., 2002). In grazed pastures, soil nutrient

redistribution occurs as animals consume forage from throughout the pasture but concentrate

excreta return in areas around water and shade where they spend more time (Mathews et al.,

1994). Nutrients in excreta are much more readily available for plant uptake or loss to the

environment than nutrients in plant matter. Nitrogen losses from dung and urine are particularly

sensitive to climatic and edaphic conditions (Boddey et al., 2004), thus excreta deposition can

hasten N depletion in extensively managed grasslands. Leaching of nitrate N is a major pathway

of N loss, while gaseous N emissions from dung and urine occur mainly in the form of NH3 and

only a small portion of N20 and NO is emitted (Pineiro et al., 2006).

Urine

Quantity and distribution in grazed pasture

Cattle urinate approximately 8 to 10 times d-1 (Carran and Theobald, 1999; Peterson and

Gerrish, 1996) with a volume of 10 to 25 L d-1 (Mathews et al, 1996). A given urine spot covers

0.28 to 0.37 m2 (Haynes and Williams, 1993). The literature on frequency, amount, and area

impacted by excreta was summarized (Table 2-1) by Haynes and Williams (1993). The quantity

and distribution of urine will vary depending on the season; during warmer weather cattle will

consume more water which will increase frequency of urination and dilute the concentration of

nutrients. Also during warmer weather, cattle will seek shade and water causing more urine to be

deposited there (Sugimoto et al., 1987). Soil P and K concentrations of 100 and 1000 mg kg-',

respectively, in the upper 7.5 cm of the soil profile were reported in a zone 10 to 20 m from the

water source for cattle grazing in a three-paddock rotational system (West et al., 1989; Peterson

and Gerrish, 1996). Nitrogen loading in a urine patch can reach up to the equivalent of 1000 kg









Russelle, Michael P. 1996. Nitrogen cycling in pasture systems. In Nutrient cycling in forage
systems. p. 125-166. In R.E. Joost and C.A. Roberts (ed.) Nutrient cycling in forage
systems. Potash and Phosphate Institute and The Foundation for Agronomic Research,
Manhattan, Kansas.

Russelle, M.P., B.A. Browne, N.B. Turyk, and B. Pearson. 2005. Denitrification under
pastures on permeable soils helps protect ground water quality. p. 98. In S.C. Jarvis et al.
(ed.), Optimisation of nutrient cycling and soil quality for sustainable grasslands.
Wageningen Academic Publishers, Wageningen, The Netherlands.

Saarijarvi, K., P. Virkajarvi, H. Heinonen-Tanski, and I. Taipalinen. 2004. N and P
leaching and microbial contamination from intensively managed pasture and cut
sward on sandy soil in Finland. Agriculture, Ecosystems, and Environment. 104:
621-630.

Sakar, D., and G.A. O'Connor. 2004. Plant and soil responses to biosolids-phosphorus in two
Florida soils with high phosphorus content. Communications in Soil Science and Plant
Analysis. 35:1569-1589.

SAS Institute. 2007. SAS/STAT user's guide. Release version 9.1. SAS Inst., Cary,
NC.

Saunders, W.M.H. 1984. Mineral composition of soil and pasture from areas of grazed
paddocks, affected and unaffected by dung and urine. New Zealand Journal of
Agricultural Research. 27:405-412.

Schachtman, Daniel P., Robert J. Reid and S.M. Ayling. 1998. Phosphorus uptake by
plants: From soil to cell. Plant Physiol. 116:447-453.

Scott, J.M. 1920. Bahiagrass. University of Florida Agricultural Experiment Stations,
Gainesville, FL.

Shigaki, Francirose, Andrew Sharpley and Luis Ignacio Prochnow. 2006. Animal based
agriculture, phosphorus management and water quality in Brazil: Options for the
future. Sci. Agric. (Piracicaba, Braz.) 63:194-209.

Sigua, G.C., M.J. Williams, and S.W. Coleman. 2004. Levels and changes of soil
phosphorus in subtropical beef cattle pastures. Comm. Soil Sci. Plant Anal.
35:975-990.

Silva, R.G., K.C. Cameron, H.J. Di, and E.E. Jorgensen. 2005. A lysimeters study to
investigate the effect of dairy effluent and urea on cattle urine N losses, plant
uptake and soil retention. Water, Air and Soil Pollution. 164:55-78.

Skerman, P.J., and F. Riveros. 1989. Tropical grasses. FAO, Rome.









LIST OF FIGURES


Figure page

3-1 This drawing (not to scale) shows one plot or experimental unit and the areas to
which urine and dung were applied. In plots where urine was applied, 1 L was
applied to the area labeled 30 (inside a 30-cm diameter circle) and 1 L to the area
labeled 60 (the area outside the 30-cm diameter circle but inside the 60-cm diameter
circle). In plots where dung was applied, 2 kg fresh weight of dung was applied to
the area labeled 30. .........................................................................55

3-2 Excreta type X excreta application frequency interaction (P = 0.063) for bahiagrass
DM harvested during 2006. There was no effect of dung application frequency on
DM harvested (P > 0.375), but there were linear (P < 0.001) and quadratic (P =
0.097) effects of urine application frequency on the response. Standard error of a
treatment mean was 103 kg ha-1. Dry matter harvested was greater for urine- than
dung-treated plots for Application Frequencies 1 (P = 0.002), 2 (P = 0.003), and 3 (P
< 0 .0 0 1)............... .................. ......... ................................................ 6 6

3-3 Excreta type X excreta application frequency interaction (P < 0.001) for bahiagrass
herbage N concentration during the 2006 growing season. There was no effect of
dung application frequency on N concentration (P > 0.246), but there were linear (P
< 0.001) and cubic (P < 0.001) effects of urine application frequency on the
response. Standard error of a treatment mean was 1.5 g kg-. Herbage N
concentration was greater for urine- than dung-treated plots for Application
Frequencies 2 and 3 (P < 0.001), but there was no excreta type effect for a single
application (P = 0.495). ..................... .................... ................... ...........69

3-4 Excreta type X excreta application frequency interaction (P = 0.002) for bahiagrass
herbage P concentration during the 2006 growing season. There was no effect of
dung application frequency on P concentration (P > 0.583), but there were linear (P
= 0.025) and quadratic (P = 0.002) effects of urine application frequency on the
response. Standard error of a treatment mean was 0.07 g kg-. Herbage P
concentration was affected by excreta type for Application Frequencies 1 and 3 (P =
0.004 and 0.035, respectively), but there was no effect of excreta type at Frequency 2
(P = 0 .752). ................................................................................ 7 1

3-5 Excreta type X excreta application frequency interaction (P = 0.099) for bahiagrass
herbage in vitro digestible organic matter (IVDOM) concentration during the 2006
growing season. There was no effect of dung application frequency on IVDOM (P >
0.300), but there was a linear (P = 0.004) effect of urine application frequency on the
response. Standard error of a treatment mean was 3.1 g kg1. Herbage IVDOM
concentration is different for Frequency 3 (P = 0.006) but not different for
Frequencies 2 (P = 0.132) and 1 (P = 0.445).......... ..... ................. ............... 73

3-6 Excreta type X excreta application frequency interaction (P < 0.001) for N harvested
in bahiagrass herbage during the 2006 growing season. There was no effect of dung









observations in the field suggest that rate of flow of natural urine deposition is more rapid and

may be associated with greater lateral movement of urine across the soil surface before it soaks

in. Despite this difference, data from the current study are comparable with several other

experiments.

With dairy cows, Lantinga et al. (1987) reported that urine affected growth in a 0.68-m2

area. Haynes and Williams (1993) indicated that the area covered by a urine event ranges from

0.16 to 0.49 m2, and de Klein (2001) reported that the area affected is 2.3 times the area to which

urine is applied. Using the mid-point of the range proposed by Haynes and Williams (1993) and

the factor suggested by de Klein (2001) results in a calculated affected area of 0.75 m2,

comparable to that measured in the current study. When artificial urine was applied to perennial

ryegrass that was fertilized with 250 kg N ha-1 yr1, effects were measured up to only 15 cm from

the edge of the urine patch (Deenen and Middelkoop, 1992). This corresponds to the results of

the current study because urine was applied to a circle of 30-cm radius and herbage accumulation

was affected only through a 45-cm radius.

As in urine-treated plots, there was application frequency X ring number interaction in

dung-treated plots, but the response was very different than for urine. The primary factor driving

the interaction was the negative impact of dung application, particularly multiple applications, on

DM harvested in Ring 1 (Fig. 4-3). This resulted in quadratic effects of ring number on DM

harvested for three (P = 0.025) dung applications, linear (P < 0.001) and quadratic (P = 0.008)

effects for two applications, and no effect for one (P > 0.370) or for zero applications (P > 0.100)

(Fig. 4-3). Reduction in DM harvested in Ring 1, where the dung was applied, ranged from 500

to more than 1000 kg ha- for plots to which multiple applications of dung were made.









Replicate was a random effect. Polynomial contrasts were used to assess the response to ring

number and excreta application frequency. Differences were considered significant when P <

0.10. Data presented are least squares means.

In this chapter, only ring main effects and interactions with ring will be reported. This is

done for two reasons. First, our objective in this chapter is to assess spatial patterns of response,

and differences among rings are how this is characterized. Also, because of the concentric circle

sampling approach used, the amount of land area sampled was different for each ring (i.e., R1

through R4). Thus, when SAS calculates a main effect mean for the other treatment factors (i.e.,

excreta application frequency and management intensity) and the frequency x intensity

interaction means, the value is a mean across levels of ring number. When calculated in this

manner, these means are not weighted to account for the different areas of the rings. In Chapter

3, all data presented were from the entire 0.636-m2 circular area around the excreta deposit

encompassed by Rings 1 through 3, and they were calculated by weighting the responses for the

specific areas of each of those three rings. Thus, the best assessment of the effects of

management intensity and excreta application frequency main effects and the interaction of these

two factors are the data presented in Chapter 3.

Results and Discussion

Dry Matter Harvested

There was a main effect (P < 0.001) of ring number on DM harvested for dung-treated

plots, but there also was application frequency X ring number interaction (P = 0.027) (Table A-

2). Likewise for urine-treated plots, there was a main effect (P < 0.001) of ring number and an

excreta application frequency X ring number interaction (P < 0.001) (Table A-3).

Application frequency X ring number interaction occurred for urine-treated plots because

DM harvested decreased linearly (P < 0.001) as ring number increased for Application









nutritive value, especially N concentration, and that these differences likely will affect the

composition of dung and urine.

The 2 x 2 x 3 factorial resulted in 12 treatments. In addition there were three control

treatments that received no excreta and were fertilized with N at 0, 60, or 120 kg ha-1 yr-. The 15

treatments were replicated three times in a randomized complete block design. Plots were 3 x 3

m in area with a 1-m bahiagrass alley surrounding each plot.

Excreta Source Pastures

There were two replicates of each management intensity (Average and High) arranged in

completely randomized design. Pastures were stocked continuously and pasture size was 1 ha for

Average and 0.5 ha for High. Each pasture was grazed with two crossbred yearling heifers

(Angus x Brahman) with average initial weight of 408 kg, and animal care was monitored

according to Institutional Animal Care and Use Committee Protocol Number D655. Heifers were

provided with access to water, trace mineral mix (minimum concentrations of Ca, 12; P, 6; NaC1,

19; K, 0.8; Mg, 1.0; S, 0.4; and Fe, 0.4 g [100 g]-1), and artificial shade (3.1 x 3.1 m) on all

treatments.

All pastures were fertilized with 17 kg P and 66 kg K ha-1 on 15 May 2006. The Average

treatment received 60 kg N ha-1 yr- in two equal applications of 30 kg ha-1 that were made on 15

May and 15 July 2006. On the same days, the High treatment received 60 kg N ha-1 for a total of

120 kg N ha-1 yr-. Grazing was initiated on 26 May 2006. Fertilization occurred in advance of

initiation of grazing so that bahiagrass herbage would reflect treatment effects by the time cattle

first entered the pastures.

The pastures were sampled to characterize herbage mass and nutritive value during each

excreta collection period. Herbage mass was quantified using double sampling with the indirect

sampling tool being a disk meter (Stewart et al., 2005). On each of the four pastures, five double










as N treatment increased from 0 N and no urine applied, to 300 or 600 kg fertilizer N ha-1 plus a

single urine application. Herbage N harvested was 143, 188, and 202 kg N ha- for the three

treatments, but absence a no-urine control for each fertilizer level makes it impossible to draw

conclusions about the impact of urine.


80

70

60

50

S40

c 30
I
z


- Urine
-Dung


Application Frequency (Number yr1)


Figure 3-6. Excreta type X excreta application frequency interaction (P < 0.001) for N harvested
in bahiagrass herbage during the 2006 growing season. There was no effect of dung
application frequency on N harvested (P > 0.404), but there was a linear (P < 0.001)
effect of urine application frequency on the response. Standard error of a treatment
mean was 1.5 kg ha-l. Herbage N harvested was greater for urine- than dung- treated
plots for Frequencies 1 through 3 (P < 0.001).









S o il ....................................................... 3 8
W after ....................... ..... ............................................ .......................................... 4 0
Fate of Phosphorus from Excreta Applied to Grazed Grass Swards.............. ...................41
Phosphorus Cycle in Grazed Grasslands ............................ .. ........................................ 41
Pathway of Phosphorus Loss from the Agroecosystem ............................................41
Leaching losses and runoff......... .................... ................. .... ............... 41
M ineralization ........................................44
Recovery of Excreta Phosphorus by Grassland Plants.......................... ............... 44
S u m m a ry ....................................................... ....................... ................ 4 5

3 BAHIAGRASS HERBAGE DRY MATTER HARVESTED, NUTRIENT
CONCENTRATION, AND NITROGEN RECOVERY FOLLOWING EXCRETA
D E P O S IT IO N ................................................................................................................... 4 6

In tro d u ctio n ................... ...................4...................6..........
M materials and M methods ................................... ... .. .......... ....... ...... 47
Experim mental Sites ................................................... ............ ........... 47
Treatm ents and D design .................. ...................................... .. ............ 48
E x creta Source P astures ......................................................................... ................... 49
E xcreta C collection and A nalyses......................................................................... ...... 50
Plot M anagem ent ........................................................ .......... .. ............. 53
Excreta A application ................................................................... .. ..... ... 53
Forage Harvest and Herbage Analyses......................... ..................55
Lysimeter Placement, Water Sampling, and Water Analyses.......................................56
Data Presentation and Statistical Analyses................................ ........................ 57
R results and D iscu ssion ............. ...... ........................................ .................. .. .... .... ....... 58
Characteristics of Excreta Source Pastures ........................................... ...............58
Excreta Com position .................................. .. .. .. ...... .. ............59
N nutrients A applied in Excreta .................................................... .................. ............... 61
Herbage Dry Matter Harvested from Excreta-Treated Plots................ ..................62
Herbage Nitrogen Concentration..................... ....... ............................. 67
H erbage Phosphorus Concentration ........................................ .......................... 70
Herbage In Vitro Digestible Organic M atter........................................ ............... 72
T total N nitrogen H arrested ........................................................................ .................. 73
Total Phosphorus H arrested .......................................... ............... ............... .76
E xcreta N nitrogen R recovery ..................................................................... ..................78
Sum m ary and C onclu sions ......................................................................... ....................84

4 SPATIAL PATTERNS OF BAHIAGRASS HERBAGE ACCUMULATION AND
NUTRIENT CONCENTRATION RESPONSES TO TYPE AND FREQUENCY OF
E X C R E TA D E P O SIT IO N .......................................................................... .....................86

In tro du ctio n ................... ...................8...................6..........
M materials and M methods ................................... ... .. .......... ....... ...... 87
Treatm ents and D design .................. ...................................... .. ............ 88
Excreta Source Pastures .......................................... ................... ........ 89
E xcreta C collection and A nalyses......................................................................... ...... 89


6









the first cutting following applications of urine in spring, summer, and autumn. In the current

study, the lack of significance for the single application treatment likely is due to the fact that it

occurred at the beginning of the experimental period and because the herbage N concentrations

reported are weighted averages across four harvests that occurred over a 112-d period following

that application. Application Frequencies 2 and 3 received urine at Day 42 and Days 42 and 84 in

this period, respectively, and their effect was measurable across the time period of the study.

The effect of application frequency within each ring number was explored to determine

how far from the center of urine application an effect on herbage N concentration could be

detected. This response had strong linear (P < 0.001) and cubic (P < 0.001) effects of application

frequency for Rings 1 through 3. With increasing urine application frequency, herbage N

concentration increased, but the effect was minimal between zero and one application and

likewise between two and three applications (Fig. 4-4). For Ring 4, the data began to converge

markedly, but the linear effect (P = 0.029) of application frequency remained. This indicates that

the effect of urine application on herbage N concentration extended 30 cm beyond the edge of

application and 15 cm further than the effect on DM harvested (Fig. 4-2).




Full Text

PAGE 1

1 NUTRIENT DYNAMICS IN BAHIAGRASS SWAR DS IMPACTED BY CATTLE EXCRETA By UNA RENEE WHITE A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2008

PAGE 2

2 2008 UNA RENEE WHITE

PAGE 3

3 To Jacob.

PAGE 4

4 ACKNOWLEDGMENTS Sincere appreciation and thanks are extende d to Dr. Lynn E. Sollenberger, advisor and mentor, for his time and effort to make my time here a great experience. Also, I would like to thank the advisory committee, Dr. Donald A. Gr aetz, Dr. Joao M.B. Vendramini, and Dr. Yoana C. Newman for serving on the committee, provi ding assistance, and for reviewing the thesis. Special thanks are given to Dr. Kenneth R. Woodard for his help in planning and carrying out the research, including his insightful input and assistance constructing lysimeters. My thanks go to those who assisted in both field and lab work. Si ndy Interrante, Kesi Liu, Miguel Castillo, and Tiberio Saraiva collected excreta and harvested. Dwight Thomas and Sid Jones were essential in helping with the project. Thanks are due to Richard Fethiere and the group of the Forage Evaluation Support Laborat ory for conducting sample analysis. I am grateful to Dr. Newman for her assi stance with statistical analyses. I would like to thank my family for their suppo rt, love, and prayers. I want to thank the Leech family for their support. Last but not leas t I want to thank W. Jacob Leech for his support and love. I am grateful for hi s time collecting dung beetles.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........8 LIST OF FIGURES................................................................................................................ .......10 ABSTRACT....................................................................................................................... ............13 CHAPTER 1 INTRODUCTION.................................................................................................................15 2 LITERATURE REVIEW......................................................................................................17 Introduction................................................................................................................... ..........17 Bahiagrass and Its Use in Florida...........................................................................................18 Center of Origin and Introduction to Florida..................................................................18 Morphological Characteristics.........................................................................................18 Adaptation to Environments............................................................................................19 Use in Production Systems..............................................................................................19 Pensacola Bahiagrass.......................................................................................................20 Bahiagrass Response to Nitrogen and Phosphorus.................................................................20 Herbage Accumulation....................................................................................................20 Herbage Nutritive Value..................................................................................................22 Characteristics of Cattle Excreta............................................................................................24 Urine.......................................................................................................................... .....24 Quantity and distribution in grazed pasture..............................................................24 Chemical composition..............................................................................................26 Effect on herbage production and nutritive value.....................................................27 Dung........................................................................................................................... .....29 Quantity and distribution in grazed pasture..............................................................29 Chemical composition..............................................................................................29 Effect on herbage production and nutritive value.....................................................30 Fate of Nitrogen from Excreta Applied to Grazed Grass Swards..........................................32 Nitrogen Cycle in Grazed Grasslands.............................................................................32 Pathways of Nitrogen Loss from the Agroecosystem.....................................................33 Gaseous losses..........................................................................................................33 Leaching losses.........................................................................................................35 Denitrification losses................................................................................................35 Nitrification...............................................................................................................36 Recovery of Excreta Nitrogen by Grassland Plants........................................................36 Urine.........................................................................................................................36 Dung..........................................................................................................................38 Impact on Soil and Water...............................................................................................38

PAGE 6

6 Soil........................................................................................................................... .38 Water.........................................................................................................................40 Fate of Phosphorus from Excreta Applied to Grazed Grass Swards......................................41 Phosphorus Cycle in Grazed Grasslands.........................................................................41 Pathway of Phosphorus Loss from the Agroecosystem..................................................41 Leaching losses and runoff.......................................................................................41 Mineralization..........................................................................................................44 Recovery of Excreta Phosphorus by Grassland Plants...................................................44 Summary........................................................................................................................ .........45 3 BAHIAGRASS HERBAGE DRY MA TTER HARVESTED, NUTRIENT CONCENTRATION, AND NITROGEN RECOVERY FOLLOWING EXCRETA DEPOSITION..................................................................................................................... ....46 Introduction................................................................................................................... ..........46 Materials and Methods.......................................................................................................... .47 Experimental Sites...........................................................................................................47 Treatments and Design....................................................................................................48 Excreta Source Pastures..................................................................................................49 Excreta Collection and Analyses.....................................................................................50 Plot Management.............................................................................................................53 Excreta Application.........................................................................................................53 Forage Harvest and Herbage Analyses............................................................................55 Lysimeter Placement, Water Sa mpling, and Water Analyses.........................................56 Data Presentation and Statistical Analyses......................................................................57 Results and Discussion......................................................................................................... ..58 Characteristics of Excreta Source Pastures.....................................................................58 Excreta Composition.......................................................................................................59 Nutrients Applied in Excreta...........................................................................................61 Herbage Dry Matter Harvested from Excreta-Treated Plots...........................................62 Herbage Nitrogen Concentration.....................................................................................67 Herbage Phosphorus Concentration................................................................................70 Herbage In Vitro Digestible Organic Matter...................................................................72 Total Nitrogen Harvested................................................................................................73 Total Phosphorus Harvested............................................................................................76 Excreta Nitrogen Recovery.............................................................................................78 Summary and Conclusions...................................................................................................84 4 SPATIAL PATTERNS OF BAHIAGRA SS HERBAGE ACCUMULATION AND NUTRIENT CONCENTRATION RESPONSES TO TYPE AND FREQUENCY OF EXCRETA DEPOSITION.....................................................................................................86 Introduction................................................................................................................... ..........86 Materials and Methods.......................................................................................................... .87 Treatments and Design....................................................................................................88 Excreta Source Pastures..................................................................................................89 Excreta Collection and Analyses.....................................................................................89

PAGE 7

7 Plot Management.............................................................................................................90 Excreta Application.........................................................................................................91 Forage Harvest and Laboratory Analyses.......................................................................92 Statistical Analysis..........................................................................................................93 Results and Discussion......................................................................................................... ..94 Dry Matter Harvested......................................................................................................94 Nitrogen Concentration.................................................................................................100 Herbage Phosphorus Concentration..............................................................................103 Herbage IVDOM...........................................................................................................104 Herbage Nitrogen Harvested.........................................................................................106 Herbage Phosphorus Harvested.....................................................................................109 Summary and Conclusions...................................................................................................112 5 SUMMARY AND CONCLUSIONS..................................................................................115 APPENDIX A STATISTICAL TABLES.....................................................................................................121 REFERENCES..................................................................................................................... .......124 BIOGRAPHICAL SKETCH.......................................................................................................139

PAGE 8

8 LIST OF TABLES Table page 2-1 Number and weight or volume of dung or urine events per day and surface area covered (adapted from Haynes and Williams, 1993)........................................................25 3.1 Chemical composition of excreta analyzed immediately after co llection (fresh) and after storage (stored) for up to 8 d at 4oC. Each value is the mean of the analysis of two subsamples................................................................................................................. .53 3-2 Herbage mass, N, P, and in vitro dige stible organic matter (IVDOM) concentrations of bahiagrass herbage during excr eta collection periods in 2006......................................59 3-3. Composition of fresh dung from Averag e and High management intensity source treatments during three collection periods in 2006. Each value reported is the average across three subsamples from a composite dung sample. Dung was composited across replicates within a management inte nsity treatment, so statistical comparisons of treatment effects are not possible..................................................................................61 3-4. Composition of urine from Average and High management intensity source treatments during three collection peri ods of 2006. Each value reported is the average across three subsamples. Urine wa s composited across replicates within a management intensity treatment, so statis tical comparisons of treatment effects are not possible................................................................................................................... .....61 3-5 Nutrients applied to dung treatments in bahiagrass swards. Calc ulations are based on chemical analyses of fresh dung (Table 3-3) and a 2-kg fresh weight dung application to a circle of 30-cm diameter. Data are expressed as kg of N, P, and K applied ha-1.........................................................................................................................62 3-6 Nutrients applied to urine treatments in bahiagrass swards. Calc ulations are based on the chemical analyses of urine (Table 34) and a 2-L volume of urine applied to a circle of 60-cm diameter. Data are expressed as kg of N, P, and K applied ha-1...............62 3-7 Effect of N fertilization on bahiagrass herbage responses for treatments to which no excreta was applied............................................................................................................65 3-8 Monthly rainfall totals for 2006 for the re search location and the 30-yr average for Island Grove, FL. Island Grove is located 10 km from the research location and is the nearest site for which 30-yr data exist...............................................................................65 A-1 P values for interactions and main effect s on response variables discussed in Chapter 3.............................................................................................................................. ..........121 A-2 P values for main effects, interactions, and polynomial contrasts for ring number for dung-treated plots as discussed in Chapter 4...................................................................122

PAGE 9

9 A-3 P values for main effects, interactions, and polynomial contrasts for ring number for urine-treated plots as discussed in Chapter 4...................................................................123

PAGE 10

10 LIST OF FIGURES Figure page 3-1 This drawing (not to scale) shows one plot or experimental unit and the areas to which urine and dung were applied. In pl ots where urine was applied, 1 L was applied to the area labeled 30 (inside a 30-cm diameter circle) and 1 L to the area labeled 60 (the area outside the 30-cm diamet er circle but inside the 60-cm diameter circle). In plots where dung was applied, 2 kg fresh weight of dung was applied to the area labeled 30............................................................................................................ .55 3-2 Excreta type X excreta application frequency interact ion (P = 0.063) for bahiagrass DM harvested during 2006. There was no e ffect of dung application frequency on DM harvested (P > 0.375), but there were linear (P < 0.001) a nd quadratic (P = 0.097) effects of urine application frequenc y on the response. Standard error of a treatment mean was 103 kg ha-1. Dry matter harvested was greater for urinethan dung-treated plots for Application Frequenc ies 1 (P = 0.002), 2 (P = 0.003), and 3 (P < 0.001)....................................................................................................................... .......66 3-3 Excreta type X excreta app lication frequency interaction ( P < 0.001) for bahiagrass herbage N concentration during the 2006 growing season. There was no effect of dung application frequency on N concentration ( P > 0.246), but there were linear ( P < 0.001) and cubic ( P < 0.001) effects of urine ap plication frequency on the response. Standard error of a treatment mean was 1.5 g kg-1. Herbage N concentration was greater for urinethan dung-treated plots for Application Frequencies 2 and 3 ( P < 0.001), but there was no excret a type effect for a single application ( P = 0.495)......................................................................................................69 3-4 Excreta type X excreta app lication frequency interaction ( P = 0.002) for bahiagrass herbage P concentration during the 2006 growing season. There was no effect of dung application frequency on P concentration ( P > 0.583), but there were linear ( P = 0.025) and quadratic ( P = 0.002) effects of urine application frequency on the response. Standard error of a treatment mean was 0.07 g kg-1. Herbage P concentration was affected by excreta type for Application Frequencies 1 and 3 ( P = 0.004 and 0.035, respectively), but there was no e ffect of excreta type at Frequency 2 ( P = 0.752)...................................................................................................................... ...71 3-5 Excreta type X excreta app lication frequency interaction ( P = 0.099) for bahiagrass herbage in vitro digestible organic ma tter (IVDOM) concentr ation during the 2006 growing season. There was no effect of dung application frequency on IVDOM ( P > 0.300), but there was a linear ( P = 0.004) effect of urine application frequency on the response. Standard error of a treatment mean was 3.1 g kg-1. Herbage IVDOM concentration is different for Frequency 3 ( P = 0.006) but not different for Frequencies 2 ( P = 0.132) and 1 ( P = 0.445).....................................................................73 3-6 Excreta type X excreta app lication frequency interaction ( P < 0.001) for N harvested in bahiagrass herbage du ring the 2006 growing season. There was no effect of dung

PAGE 11

11 application frequency on N harvested ( P > 0.404), but there was a linear ( P < 0.001) effect of urine application frequency on th e response. Standard error of a treatment mean was 1.5 kg ha-1. Herbage N harvested was greater for urinethan dungtreated plots for Frequencies 1 through 3 ( P < 0.001)...................................................................75 3-7 Excreta type X excreta app lication frequency interaction ( P < 0.001) for P harvested in bahiagrass herbage du ring the 2006 growing season. There was no effect of dung application frequency on P harvested ( P > 0.570), but there were linear ( P < 0.001) and quadratic effects ( P = 0.052) of urine application frequency on the response. Standard error of a treatment mean was 0.35 kg ha-1. Herbage P harvested was greater for urinethan dung-treated plots for Application Frequencies 2 ( P = 0.007) and 3 ( P < 0.001) and tended to be greater for Frequency 1 ( P = 0.115)..........................77 3-8 Excreta type main effect ( P < 0.001) for excreta N recove ry in harvested bahiagrass herbage during the 2006 growing season. Th ere was no excreta type X excreta application frequency interaction ( P = 0.543), but there was a linear effect ( P = 0.064) of excreta application frequency on N re covery. Standard error of a treatment mean was 1.8%..................................................................................................................80 3-9 Excreta type X applicati on frequency interaction for NO3-N concentration in shallow soil water on 15 Sept. 2006. There was a qua dratic effect of urine application frequency on NO3-N concentration ( P = 0.057), and there was a linear effect ( P = 0.0394) of dung application frequency on th e response. There was no effect of excreta type at any level of application frequency ( P > 0.181). Standard error of a treatment mean was 0.03 mg L-1........................................................................................83 4-1 Diagram of harvested rings to quantify sp atial pattern of res ponse to dung and urine application (R1, circle of 15-cm radius; R 2, 15to 30-cm radius; R3, 30to 45-cm radius; R4, 45to 60-cm radius)........................................................................................93 4-2 Ring number by urine applic ation frequency interaction ( P < 0.001) on herbage DM harvested during 2006. Ring number effects were linear for three ( P < 0.001) and two applications ( P < 0.001), linear ( P < 0.001) and quadratic ( P = 0.058) for one application, and not significant ( P > 0.352) for zero applications. Standard error of a treatment mean was 76 kg ha-1. Ring Numbers 1 through 4 re fer to sampling areas 0 to 15, 15 to 30, 30 to 45, and 45 to 60 cm fr om the center of th e urine application, respectively................................................................................................................... .....98 4-3 Ring number by dung applica tion frequency interaction ( P < 0.027) on herbage DM harvested during 2006. Ring number e ffects were quadratic for three ( P = 0.025) applications, linear ( P < 0.001) and quadratic ( P = 0.008) for two applications, and not significant for one ( P > 0.370) or for zero applications ( P > 0.100). Standard error of a treatment mean was 106 kg ha-1. Ring Numbers 1 through 3 refer to sampling areas 0 to 15, 15 to 30, and 30 to 45 cm from the center of the dung application, respectively....................................................................................................99

PAGE 12

12 4-4 Ring number by urine application freque ncy interaction (P < 0.001) on total-season herbage N concentration during 2006. Ring numbe r effects were linear for three (P < 0.001) and for two (P < 0.001) applications, but not significant for one (P > 0.171) and zero (P > 0.128) urine applications. Sta ndard error of a treatment mean was 0.25 g kg-1. Ring Numbers 1 through 4 refer to sampling areas 0 to 15, 15 to 30, 30 to 45, and 45 to 60 cm from the center of the urine application, respectively...........................102 4-5 Ring number X urine applic ation frequency interaction ( P = 0.054) for in vitro digestible organic matter (IVDOM) concen tration during 2006. Ring number effects were linear ( P <0.0001) for two applications of ur ine and not significant for three ( P > 0.144), one, and zero applications ( P > 0.272). Standard error of a treatment mean was 2.4 g kg-1. Ring Numbers 1 through 4 refer to sampling areas 0 to 15, 15 to 30, 30 to 45, and 45 to 60 cm from the center of the urine applic ation, respectively............105 4-6 Ring number X urine applic ation frequency interaction ( P < 0.001) for bahiagrass herbage N harvested during 2006. The ring number effect was linear for three ( P < 0.001), two ( P < 0.001), and one ur ine application ( P < 0.001) and there was no effect of ring number ( P >0.167) for the no urine cont rol. Standard error of a treatment mean was 1.2 kg ha-1. Ring Numbers 1 through 4 refer to sampling areas 0 to 15, 15 to 30, 30 to 45, and 45 to 60 cm fr om the center of th e urine application, respectively................................................................................................................... ...108 4-7 Ring number X dung applica tion frequency interaction ( P = 0.019) for bahiagrass herbage N harvested during 2006. The ri ng number effect was quadratic ( P = 0.048) for three applications and linear ( P < 0.001) and quadratic ( P = 0.007) for two dung applications, but was not significant ( P > 0.288) for one and no dung applications ( P > 0.167). Standard error of a treatment mean was 1.5 kg ha-1. Ring Numbers 1 through 3 refer to sampling areas 0 to 15, 15 to 30, and 30 to 45 cm from the center of the dung application, respectively...............................................................................109 4-8 Ring number X urine applic ation frequency interaction ( P < 0.001) for bahiagrass herbage P harvested during 2006. Ring number effects were linear ( P < 0.001) and cubic ( P = 0.0419) for three urine app lications, linear for two ( P < 0.001) and one application ( P = 0.019), and not significant ( P > 0.194) for zero applications. Standard error of a treatment mean was 0.4 kg ha-1. Ring Numbers 1 through 4 refer to sampling areas 0 to 15, 15 to 30, 30 to 45, and 45 to 60 cm from the center of the urine application, respectively.........................................................................................111 4-9 Ring number X dung application frequenc y interaction (P = 0.058) for bahiagrass herbage P harvested during 2006. Ring number effects were quadratic (P = 0.073) for three dung applications, lin ear (P = 0.004) and quad ratic (P = 0.046) for two applications, and not significant for one (P > 0.257) or for zero applications (P > 0.289). Standard error of a treatment mean was 0.4 kg ha-1. Ring Numbers 1 through 3 refer to sampling areas 0 to 15, 15 to 30, and 30 to 45 cm from the center of the dung application, respectively.........................................................................................112

PAGE 13

13 Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science NUTRIENT DYNAMICS IN BAHIAGRASS SWAR DS IMPACTED BY CATTLE EXCRETA By Una Renee White May 2008 Chair: Lynn E. Sollenberger Major: Agronomy Most nutrients consumed by grazing livestock are returned to pasture in excreta, but excreta effects on forage respons es, plant nutrient recovery, a nd shallow soil water quality are not well defined. The objective was to determine the effect of management intensity, excreta type, and number of excreta applications on bahiagrass ( Paspalum notatum Flgge) herbage dry matter (DM) harvested and nutritiv e value, excreta nutrient recove ry in harvested herbage, and nutrient concentration in shallow soil water. Trea tments were the factorial combinations of two management intensities (Average and High, 60 and 120 kg N ha-1, respectively), two excreta types (dung and urine), and three applic ation frequencies (1, 2, or 3 season-1). Three control treatments received no excreta and 0, 60, or 120 kg N ha-1. Dung and urine were collected from animals grazing bahiagrass pasture and applied to ungrazed bahiagrass plots. A urine application was 2 L distributed to a 60-cm diameter circle, and a dung app lication was 2 kg fresh weight applied to a 30-cm diameter circle. Concentric rings (radii of 0-15, 15-30, 30-45, and for urine plots, 45-60 cm from the center of excreta application) were cl ipped monthly from July through October 2006 to measure forage responses, and lysimeters were placed 1.3 m below soil level directly under the excreta appli cation to measure nutrient concentrations in soil water. Herbage data are reported for a circle of radius 45 cm from the center of the excreta deposit and also by

PAGE 14

14 ring to show spatial patterns of response. Herbag e response for the circle of 45-cm radius was affected to the greatest extent by application frequency X excret a type interac tion. Interaction occurred because dung application had no effect on most herbage responses, whereas responses to urine were consistently significant. In urin e-treated plots, herbage DM harvested increased from 2760 at zero applications to 4670 kg ha-1 with three, while over the same application frequency herbage N concentration increased from 13.3 to 16.2 g kg-1, P concentration from 3.45 to 3.87 g P kg-1, in vitro digestible organic matte r concentration from 569 to 588 g kg-1, N harvested from 37 to 75 kg ha-1, and P harvested from 9 to 18 kg ha-1. Excreta N recovery was greater from urine than dung and decreased as ap plication frequency increas ed from one to three (28 to 18% for urine and 4 to < 1% for dung). Sp atial characteristics of response were assessed within excreta type and were consistently a ffected by application fr equency X ring number interaction. In urine-treated plots, herbage respon se generally was greatest near the center of the urine deposit and decreased as distance from center of excreta application increased. In dungtreated plots, physical interference by dung re sulted in decreased herbage DM, N, and P harvested in the area under the dung deposit. Ur ine affected DM harvested and herbage P concentration up to 15 cm and N concentration a nd N harvested up to 30 cm beyond the edge of the urine application. Dung had no effect on any response outside the area physically impacted by dung. Shallow soil water was affected by tr eatment at only one sampling date, and NO3-N concentration was > 10 mg L-1 in only three samples during the entire study. In conclusion, urine has much greater impact on herbage response th an dung, both in the area impacted physically by the excreta and beyond that area. Greater N recove ry following single vs. multiple excreta events per site per year emphasizes th e importance to sustainable grassland management of grazing practices that increase unif ormity of excreta deposition.

PAGE 15

15 CHAPTER 1 INTRODUCTION Rotz et al. (2005) define grassland agricultu re as a farming system that emphasizes the importance of grasses and legumes in livestock and land management. Planted grasslands and non-forested rangeland comprise nearly 30% of the USA land area (Barnes and Nelson, 2003) and occupy more than 4 million ha in Florida (Dubeux et al., 2007). Beca use of the amount of area involved, the fate of nutrien ts within these agroecosystems has important implications for agricultural production a nd the environment. Nutrient management in grasslands has receiv ed greater attention in recent years due to soil nutrient insufficiency and associated pastur e degradation in some areas (e.g., Brazil; Boddey et al., 2004) and excessive applicat ions of nutrients and negative environmental impact in others (e.g., USA; Woodard et al., 2003). The major nut rient pools in grassland systems are soil, atmosphere, live and dead plant material, and animals (Mathews et al., 2004; Dubeux et al., 2007). Addition of livestock to gras slands increases the complexity of the nutrient cycle and the rate of fluxes among nutrient pools; the latter in creases the potential for nutrient loss to the environment (Boddey et al., 2004). This is due to chemical and biological transformations that occur during forage digestion, making the forms of nutrients in excreta more readily available for uptake or loss than those occurring in live plants or plant litter (Jarvis et al., 1995; Rotz et al., 2005). In pasture-based livestock pr oduction systems, animals gather herbage, utilize a small proportion of the nutrients and excrete the remain ing nutrient compounds in patches (Rotz et al., 2005). Efficient recovery of these nutrients is hindered by the larg e quantity of nutrients in a single dung or urine event and because a dispropor tionately large number of excreta events occur

PAGE 16

16 in small areas where cattle congregate, e.g., near to shade, water, and supplemental feed sources (Mathews et al., 2004; Solle nberger et al., 2002). It has been suggested that rotational stoc king with short grazing periods, i.e., many paddocks per pasture, decreases the opportunity and tendency of animals to congregate in lounging areas by intensifying competition for f eed and shortening resi dency periods (Haynes and Williams, 1993). In Florida, it was found that rotationally stocked pastures where grazing periods were short (1 to 7 d) had greater spat ial uniformity in time spent by cattle, excreta deposition, and soil nutrient con centration than continuously st ocked pastures (Dubeux, 2005). Thus, it is likely that greate r uniformity of excreta deposi tion can be achieved by imposing rotational stocking with short grazing periods. Whether this intensification of grazing management results in more efficient nutrient cycling may depend on the degree to which more uniform excreta deposition enhances nutrient rec overy by grassland plants and avoids excessive nutrient accumulation in soils or nutri ent loss to surface or ground water. There is little information in the literature evaluating the imp act of type and number of excreta applications to pastur e on leaching of N, changes in soil nutrient concentration, and herbage growth and nutritive value. The objective of the reported research in this thesis is to characterize the effects of cattle excreta application on 1) bahiagrass ( Paspalum notatum Flgge) herbage accumulation, chemical composition, and nut rient recovery and 2) nitrate leaching to shallow ground water under a bahiagrass sod. Comp anion studies not reported in this document will assess the effect of these factors on cha nges in soil nutrient concentration over time. Bahiagrass was chosen for this rese arch because it is the most wide ly used of the planted pasture species in Florida.

PAGE 17

17 CHAPTER 2 LITERATURE REVIEW Introduction Bahiagrass ( Paspalum notatum Flgge) is a warm-season perennial pasture grass that is important in Florida and throughout the Gulf Coast Region of the southern USA (Chambliss and Adjei, 2006). It is growing on approximately 1.1 m illion ha in Florida where it is used as the primary feed for the nearly 1 million head of beef cows (NASS: Florida, 2006). Due to the quantity of land area occupied and the number of animals supported on bahiagrass pasture, nutrient management is a key issue. One of the many challenges faced by livestock producers utilizing pasture as a feed resource is avoidance of over accumulation of livestock wastes in certain areas of the pasture. Uneven distribution of soil nutrients, due to nonuniform spatial deposition of ex creta, is thought to lead to le aching and volatilization of N and runoff of P and other nutrients from so-called nutrient hot spots in pastures, yet the effect of in situ excreta deposition on plant, soil, and water responses has not been studied in detail. The reported research objectives were to characteri ze the impact of frequency and type of excreta application on bahiagrass herbage production, ch emical composition, and nutrient recovery, and to measure the nitrate-N concentration in sh allow ground water under excreta applications. This literature review will focus on a de scription of bahiagrass and its growth characteristics. Subsequently, excreta quantit y, composition, deposition pa tterns, and effects on herbage productivity and nutritive value will be described. Finally, nutrien t release from excreta and its impact on plant, water, and soil will be explored.

PAGE 18

18 Bahiagrass and Its Use in Florida Center of Origin and Introduction to Florida Bahiagrass is native to South America and wa s described in 1810 us ing a plant collected from St. Thomas Island by Schrader and Venten at (Gates et al., 2004). Common bahiagrass is particularly abundant in Brazil, eastern Bolivia, Paraguay, and nor theastern Argentina, but the original distribution of the races of var. saurae was confined to Corrientes, Entre Rios, and the eastern edge of Santa Fe Provinces in Argen tina (Gates et al., 2004). Pensacola bahiagrass belongs to the var. saurae Scott (1920) reported that bahiag rass was first introduced into the USA by the Bureau of Plant Industry and grown by the Florida Agricultural Experiment Station in 1913 (Gates et al., 2004). Currently, bahiagra ss is widespread throughout the southern USA and Central and South America (Skerman and Rive ros, 1989; Hirata et al., 2006; Chambliss and Adjei, 2006). Morphological Characteristics Bahiagrass is a sod forming, warm-season pere nnial (Skerman and Riveros, 1989; Hirata et al., 2006; Gates et al., 2004). It has strong, shallow, horizontal rhizomes formed by short internodes (Gates et al., 2004). Pensacola bahiagrass has a decumb ent growth habit with most leaves originating from rhizomes near the so il surface (Beaty et al., 1968). Approximately 40% of Pensacola biomass is within 3 cm of the so il surface (Beaty et al., 1968). Bahiagrass leaves are attached to growing rhizomes which are produ ced continuously as long as leaves are being produced (Beaty and Powell, 1978). Upon death of leaves, soluble constituents such as N are translocated from the aging leaf to the rhiz ome, root, or young phytomers (Beaty and Powell, 1978). Bahiagrass allocates significan t mass to rhizomes and roots (Pedreira and Brown, 1996a). Bahiagrass plots were fertilized with either a low N (LN: 0 and 5 g N m-2 in Years 1 and 2),

PAGE 19

19 medium N (MN; 40 and 10 g N m-2 in Years 1 and 2), or high N rate (HN: 80 and 20 g N m-2 in Years 1 and 2) and were clipped every 10 (short interval; SI) or 21 d (long interval; LI) (Hirata, 1996). Root mass averaged across the 2-yr period was greatest with greatest N rate and longest cutting interval (LN/LI: 595 g m-2; MN/LI: 597 g m-2; HN/LI: 613 g m-2). There also were effects of N rate and cutting interval on leaf mass. Speci fically, as N rate and cu tting interval increased so did leaf mass (LN/SI: 81 g m-2; LN/LI: 89 g m-2; HN/SI: 148 g m-2; HN/LI: 288 g m-2), but leaf mass was relatively small compared to root mass. Adaptation to Environments Bahiagrass is grown throughout Florida due to its tolerance of a wide range of soil conditions, including low fertility, drought, and short-term flooding (G ates et al., 2004), pH up to 6, and its ability to withstand close grazing (Burson and Watson, 1995; Williams and Hammond, 1999). Following its introduction to Florida, it spread to the Gulf Coast and Coastal Plains of the Southeast USA and has become naturalized in th ese regions (Gates et al., 2004). Its area of growth extends north to North Carolina and west to Texas and southeastern Oklahoma (Gates et al., 2004). Use in Production Systems Approximately 75% of the 1.4 million ha of pl anted pasture in Florida are dominated by bahiagrass which supports ~ 1 million head of beef cows (Gates et al., 2004; Mislevy et al., 2005). Bahiagrass is also used for hay, although in the Lower South, most hay is produced from higher yielding warm-season perennial grasses such as bermudagrass [ Cynodon dactylon (L.) Pers.], dallisgrass ( Paspalum dilatatum Poir.), and stargrass ( Cynodon nlemfuensis Vanderyst) (Robinson, 1996; Taliaferro et al., 2004). Bahiagra ss is widely used for low maintenance turf, especially in highway rights-of-way thr oughout the Southeast (Gates et al., 2004).

PAGE 20

20 Pensacola Bahiagrass Pensacola bahiagrass belongs to P. notatum var. saurae and was first discovered growing near docks in Pensacola, FL; it is assumed that th e seed arrived on a ship from Argentina prior to 1926 (Finlayson, 1941; Burton, 1967; Gates et al., 2004). Hoveland (2000) stated that the introduction and release of Pens acola bahiagrass was a major achievement in the development of grasslands in the southern USA. Pensacola, a dipl oid, is more cold tolerant than the tetraploid bahiagrasses, including Argentine (Gates et al., 2004). Pensacola is the most widely grown of the bahiagrass cultivars and is very drought tolerant, although growth during dry periods is minimal. Its winter hardiness and survival under heavy grazing are due to its prostrate and rhizomatous nature (Pedreira and Brown, 1996a). These traits also contribute to its tolerance of continuous stocking (Pedreira and Brown, 1996b). Bahiagrass Response to Nitrogen and Phosphorus Herbage Accumulation Plant production responses are a ffected by species, stage of gr owth, amount and time of N applied, and environmental conditions following fertilizer application (Crowder and Chheda, 1982). The addition of N influences yield as lo ng as no other element is limiting (Crowder and Chedda, 1982). Without adequate P or other nutrie nts, plant growth is restricted, yields are lower, use efficiency of N is affected, and lower profits result (Gri ffith and Murphy, 1996). Dry matter accumulation of planted warm-s eason perennial grasses, like bahiagrass, depends on the amount of N applied and has pr onounced seasonal charact eristics. Dry matter yields of Pensacola bahiagrass were 3000 to 4000 kg ha-1 without applied N (Blue, 1970) and 12 000 kg ha-1 or more when 224 kg N ha-1 was applied (Sigua et al ., 2004). Beaty et al. (1975) showed that Pensacola herbage accu mulation increased from 3700 to 4360 kg ha-1 when N

PAGE 21

21 fertilization increased from 0 to 224 kg ha-1. Herbage accumulation was not different when N fertilization increased from 224 to 672 kg ha-1. Mislevy et al. (2005) found th at Pensacola produced 50 to 60 % of its total seasonal yield during long days of June, July, and August, a nd average annual yield was between 10 000 and 12 000 kg ha-1 when 112 kg N ha-1 yr-1 was applied. Total seasonal yield was 13 400 kg ha-1 in the first year of the 3-yr study, while yields in the second and third y ears were 11 000 and 6300 kg ha-1. The reason for this drastic decrease in Year 3 was dr ought. Total rainfall for the 1st, 2nd, and 3rd years was 1740, 1270, and 813 mm, respectively. Burton et al. (1997) fertilized Pensacola bahiag rass at differing rates of N, P, and K, the lowest being 56 kg N, 24 kg P, and 46 kg K ha-1; the highest was 448 kg N, 49 kg P, and 278 kg K ha-1. Average dry matter harvested over the 3-yr period was 10 620 kg ha-1, with the highest yield of 15 070 kg ha-1 from the greatest fertilizer amount. In Louisiana, Twidwell et al. (1998) found that a single appl ication of 224 kg N ha-1 to bahigrass increased annual yield from 4040 kg ha-1 for the zero control to 11 900 kg ha-1. Bahiagrass response to P fertilization has been le ss consistent than to N, due in part to bahiagrass ability to access soil P from below the depth sampled for soil analysis (Mylavarapu et al., 2007), especially when growing in Spodosols. In a st udy conducted in South Central Florida, bahiagrass herbage yield increased 12% a bove the zero control wh en fertilized with 15 kg P ha-1 (Ibrikci et al., 1999). Burton et al. (1997) found bahiagrass P rates had no effect on DM yield across a range of N and P rates. In that st udy, bahiagrass was fertil ized with varying rates of N, P, and K (from 56 kg N, 24 kg P, and 46 kg K ha-1 as the lowest fertilizer rate, to the highest of 448 kg N, 49 kg P, 279 kg K ha-1). McCaleb et al. (1966) found that Pensacola bahiagrass yield was affected by P fertilizati on only for an initial increment of 6 kg P ha-1 yr-1.

PAGE 22

22 Rhoads et al. (1997) found similar results but with different rates of P (0, 84, and 168 kg ha-1). In this study, they found that increa sing P fertilizer from 0 to 84 kg ha-1 increased yield from 9.8 to 10.7 Mg DM ha-1 in the first year and 7.3 to 8.8 Mg DM ha-1 in the second year. Herbage Nutritive Value There is extensive literature on the effects of N fertilization on the chemical composition and digestibility of Pensacola bahiagrass herbage (Beaty et al., 1975; Burton et al., 1997; Twidwell et al., 1998; Newman et al., 2006; St ewart et al., 2007). Similar to many other C4 grasses, Pensacola bahiagrass crude protein (CP) increases with increasing N fertilization, while the response of herbage in vitro digestible organic matter (IVDOM) is less clear. Beaty et al. (1975) showed that Pensaco la bahiagrass N concentration increased significantly when N fertilization increased fr om 0 to 224 and again from 224 to 672 kg N ha-1. Newman et al. (2006) fertilized bahiagrass wi th two levels of N fertilizer, 80 and 320 kg N ha-1, and harvested them by clipping to a 5-cm stubble height every 7 wk. Bahiagrass fertilized with 320 kg N ha-1 had greater IVDOM than bahiagra ss fertilized with 80 kg N ha-1 (471 and 432 g kg-1, respectively). Herbage CP was 79 and 58 g kg-1 for the 320 and the 80 kg N ha-1 treatments, respectively. Burton et al. (1997) found bahiagrass N concentrations increased from 10.6 to 17 g N kg-1 as fertilizer N applications increased from 56 to 448 kg N ha-1. Stewart et al. (2007) found that bahiagrass herbage CP increased as amount of N fertilizer increased from 40 to 120 kg N ha-1 (99 to 113 g CP kg-1), and from 120 to 360 kg N ha-1 (113 to 140 g CP kg-1). In Florida, bahiagrass was fertilized with 112 kg N, 30 kg P, and 52 kg K ha-1 yr-1, and this resulted in average CP and IVDOM concentrat ions from April through December of 142 and 563 g kg-1, respectively, in Year 1, 140 and 510 g kg-1 in Year 2, and 163 and 485 g kg-1 in Year 3 (Mislevy et al., 2005). These authors characterize d seasonal changes in bahiagrass nutritive value. Pensacolas greatest CP (186 g kg-1) occurred in April 1998 and then decreased from June

PAGE 23

23 through December. The greatest I VDOM for Pensacola was 623 g kg-1 in December 1998 and the lowest occurred in August and was 467 g kg-1. All genotypes of bahiagrass tended to drop in IVDOM during June to August due to a low solu ble carbohydrate concen tration in above-ground plant parts. In comparison to N fertilization, there are relatively few papers that address the effect of P fertilization on Pensacola bahiagra ss nutritive value. However, there has been increasing interest in P nutrition of bahiagrass because of the role of P in water quality, and this has stimulated greater research emphasis in recent years. Ibrikci et al. (1999) found that when triple superphos phate was applied at rates of 0, 17, 34, 51, and 68 kg P ha-1, there was no change in bahiagrass herb age P concentration in the first year, but in the second year the uptak e of inorganic P increased sign ificantly and P concentration increased with P fertilization. Sumner et al. ( 1992) found when bahiagrass was fertilized with P it resulted in increased P concentration and the increase was related to the amount of P applied. Phosphorus concentration of Pensaco la bahiagrass averaged 1.5 g kg-1 when fertilizer rate was 24 kg P, but when P fertilizer rate increased to 49 kg P ha-1 forage P concentration increased to 3.0 g kg-1 (Burton et al., 1997). Tiffany et al. (1999) sampled three grazed bahiagrass pastures that were growing on different soils in Florida. Pastur es were fertilized with 40 kg N ha-1. Forage P concentrations in the first (3.0 g kg-1) and second years (3.1 g kg-1) were similar early in the season. There was a decrease in P concentration from June to Nove mber in both years from approximately 3.0 to 1.3 g kg-1. The forage P concentration in Oct ober and November were below the 1.8 g kg-1 requirement for growing beef cattle (NRC, 1996).

PAGE 24

24 Characteristics of Cattle Excreta A large proportion of the nutrients consumed by grazing livestock are returned to the pasture in animal excreta (Sollenberger et al., 2002). In grazed pastures, soil nutrient redistribution occurs as animals consume fora ge from throughout the pasture but concentrate excreta return in areas around water and shade where they spend more time (Mathews et al., 1994). Nutrients in excreta are much more read ily available for plant uptake or loss to the environment than nutrients in plant matter. Nitr ogen losses from dung and urine are particularly sensitive to climatic and edaphic conditions (B oddey et al., 2004), thus excreta deposition can hasten N depletion in extensively managed grasslands. Leaching of nitrate N is a major pathway of N loss, while gaseous N emissions from dung and urine occur mainly in the form of NH3 and only a small portion of N2O and NO is emitted (Pineiro et al., 2006). Urine Quantity and distribution in grazed pasture Cattle urinate approximately 8 to 10 times d-1 (Carran and Theobald, 1999; Peterson and Gerrish, 1996) with a volume of 10 to 25 L d-1 (Mathews et al, 1996). A given urine spot covers 0.28 to 0.37 m2 (Haynes and Williams, 1993). The litera ture on frequency, amount, and area impacted by excreta was summarized (Table 21) by Haynes and Williams (1993). The quantity and distribution of urine will vary depending on the season; during warmer weather cattle will consume more water which will increase frequency of urination and dilute the concentration of nutrients. Also during warmer weather, cattle will seek shade and water causing more urine to be deposited there (Sugimoto et al., 1987). Soil P and K concentrations of 100 and 1000 mg kg-1, respectively, in the upper 7.5 cm of the soil prof ile were reported in a zone 10 to 20 m from the water source for cattle grazing in a three-paddock rotational syst em (West et al., 1989; Peterson and Gerrish, 1996). Nitrogen loading in a urine pa tch can reach up to th e equivalent of 1000 kg

PAGE 25

25 N ha-1 (Haynes and Williams, 1993; Clough et al., 2004). Pakrou and Dillon (1995) collected fresh urine from dairy cows grazing clover ( Trifolium spp.)-perennial ryegrass ( Lolium perenne L.) pastures. The N loading under the urine pa tches ranged between 650 (autumn) and 1370 kg N ha-1 (summer) in the soil. Applyi ng this much N increases the lik elihood of nutrient losses. Table 2-1. Number and weight or volume of dung or urine ev ents per day and surface area covered (adapted from Haynes and Williams, 1993). Reference Stock type Mean number of defecations per day Weight of single defecation (kg wet wt) Area covered by defecation (m2) Mean number of urinations per day Volume of single urination (L) Area covered by urination (m2) JohnstoneWallace and Kennedy (1994) Beef cow 11.8 1.77 0.06 8.5 --Castle et al. (1950) Dairy cow 11.6 --9.8 --Hancock (1950) Dairy cow 12.2 --10.1 --Goodall (1951) Dairy cow 12 1.48 -11 --Waite et al. (1951) Dairy cow -2.27 ----Doak (1952) Dairy cow ----1.6 -Hardison et al. (1956) Dairy cow 15.4 -0.09 9.4 --Petersen et al. (1956) Dairy cow 12 --8 -0.28 MacLusky (1960) Dairy cow 11.6 -0.05 ---Davies et al. (1962) Dairy cow 12 -0.07 10 2.2 0.19 Wardrop (1963) Dairy cow 16.1 --12.1 --Hogg (1968) Dairy cow -----0.18 Weeda (1967) Beef steer 10.5 -----

PAGE 26

26 Table 2-1. Continued Reference Stock type Mean number of defecations per day Weight of single defecation (kg wet wt) Area covered by defecation (m2) Mean number of urinations per day Volume of single urination (L) Area covered by urination (m2) Frame (1971) Dairy cow 11 2.7 -11 1.9 -MacDiarmid and Watkin (1972) Dairy cow 13.9 1.82 0.07 ---Robertson (1972) Dairy cow ---10 2 -During and Weeda (1973) Beef steer --0.05 ---Richards and Wolton (1976) Dairy cow --0.05 --0.49 Weeda (1979) Beef steer 10.5 -----Williams et al. (1990) Dairy cow -----0.16 Chemical composition Nutrients in urine are in plant-available forms or are rapidly mineralized within a few days (Mathews et al., 1996). Because urine is 500 to 800 g urea-N kg-1 of total-N, it is hydrolyzed very rapidly (Rotz et al., 2005), resu lting in a release of available nutrients. The urea in the urine is broken down by urease which is an enzyme produced by microbes in the soil. Urine contains many amino acids, such as hippuric acid. The hip puric acid in urine was reported to have a controlling effect on both hydrol ysis of urine N and on NH3 volatilization (Doak, 1952; Whitehead et al., 1989). When urine-N concentra tions are lowered or the amount of urine N is affected by altering forage source, feed additives, or grazing regimes, there may be a significant effect on N2O emissions, due to lower amounts of N being applied to the soil (Oenema et al.,

PAGE 27

27 1997). Potassium in urine may be leached or conv erted to less available forms, which affect supply (Carran and Theobald, 1999). Livestock on forage diets have a urine pH of approximately 7.4 (Rotz et al., 2005). The specific concentrations of N in cattle urine depe nd on factors such as diet and water consumption but normally range from 8 to 15 g N L-1 (Whitehead, 1970; Clough et al., 2004). Reducing cattle N intake by changing diet composition can lead to lower N concentration in urine with the volume of urine unchanged, fewer urinations of the same volume but with unchanged N concentration, or an unchanged number of urinat ions but with a smaller volume of urine and unchanged N concentration (Bussink and Oenema, 1998; van Groenigen et al., 2005a). Salts and other feed additives will dilute the concentratio n of N in urine through more water consumption (van Groenigen et al., 2005a). Effect on herbage production and nutritive value Nutrients in urine enhance overall pasture productivity disproportio nately to the area physically covered by the excreta site (Peter son and Gerrish, 1996). Recycled nutrients can account for up to 70% of annual pasture production in low-input systems (Mathews et al., 1996). Short-term effects of urine have a greater impact on productivity than dung (Carran and Theobald, 1999). When cattle and sheep were co rralled and combined dung and urine applied to pearl millet ( Pennisetum glaucum L.), the millet plots had an average of 53% greater seasonal yield than where only dung was applied (Powell et al., 1998). Decau et al. (2003) conducted an experiment using perennial ryegrass planted in three types of soil. Cow urine spiked with 15N was applied to the plots at a rate of 7 mg N L-1 (urea N = 4.61 mg L-1, and 15N = 2.47 mg L-1) and top dressed with 15 g N m-2 yr-1 of ammonium nitrate. The controls, which did not receive any form of N in Year 1, had a forage dry matter production of 5 to 6 Mg ha-1. Application of urine in the spring increased DM yi elds by 17 to 33%, but the

PAGE 28

28 summer application increased yield only 10 to 15%, probably due to greater volatilization losses in summer. Total N uptake for the spring and su mmer applications was greater than the fall. Average urinary N uptake in the harvested herb age over the 2-yr period was approximately 40%; for the spring harvest urinary N uptake ranged fro m 52 to 60% across soil types. The fall uptake of urinary N was poor compared to the spring and ranged from 16 to 41%. Silva et al. (2005) applied dair y cow urine to perennial ryegra ss-white clover plots. Plots received urine+urea (1000 + 4000 kg N ha-1 yr-1), urine alone (1000 kg N ha-1 yr-1), or a control with no N applied. Dry matter producti on over a 2-yr period was 10 700 kg ha-1 yr-1 (control), 19 415 kg ha-1 yr-1 (urine), and 20 000 kg ha-1 yr-1 (urine+urea). Uptake of N by the control treatment was approximately half that (364 kg ha-1 yr-1) of the urine (610 kg ha-1 yr-1) and urine+urea (705 kg ha-1 yr-1) treatments. In another study, perennial ry egrass cv. Concord and white clover cv. Grasslands Huia were used in short-term pasture rotations (Williams and Haynes, 2000). The pasture had been planted previously to cereal and grain crops for ~ 20 yr and then planted to grass-clover 2 yr before start of the experiment. Nitrogen con centration in herbage was measured 14, 33, 77, and 133 d after application of sheep urine. Resulti ng concentrations were 46, 31, 29, and 22 g N kg-1, respectively. In the long-term past ure, a pasture that was sown ~25 yr prior to the experiment and rotationally stocked by sheep, N c oncentration in herbage decreased slightly compared to the short-term pasture and was 33, 28, 24, and 24 g kg-1, respectively, for the four time periods after urine application. Clearly the impact of urine is greatest soon after application and diminishes with time.

PAGE 29

29 Dung Quantity and distribution in grazed pasture Cattle defecate approximately 12 times d-1 (Peterson and Gerrish, 1996), with the reported range from 8 to 16 times d-1 (Barrow, 1967; Wilkinson and Lo wrey, 1973; Peterson and Gerrish, 1996) and an average wet weight of 1.5 to 2.7 kg per defecation (Haynes and Williams, 1993; Table 2.1). The area covered by a single dung pat is generally from 0.05 to 0.09 m2. Like urine, dung deposition is often not uniform across the pasture and is concen trated in high traffic areas. Chemical composition Dung contains undigested herbage residues, prod ucts of animal metabolism, ingested soil, and a large biomass of microorganisms (Rotz et al., 2005). Most of the nut rients in dung are not immediately plant available; th e main factors affecting deco mposition and disappearance of dung are microbial decomposition, weathering, disinteg ration of pats due to invertebrates, and consumption and removal of dung by insects and lu mbricids (Holter, 1979; Lee and Wall, 2006). Holter (1979) found hot and dry summers slowed dung disappearance, resulting in 65% of the dung pat remaining after 65 d compared to othe r years with pat disappearance (only small particles or soil-like materi al left) in 50 d. Lee and Wall (2 006) estimated that complete disappearance for dung pats can take 57 to 78 d and 88 to 111 d in spring and summer, respectively. Almost all P excreted is in the feces, and mo re than 70% of P in manure is inorganic P (Eghball et al., 2005). Nitrogen is also excreted through dung at about 8 g kg-1 of feed consumed. Nitrogen in dung is only 20 to 25% wate r soluble, and volatile loss of NH3 is less than 5% (Rotz et al., 2005). About 10 to 20% is undigested dietar y N and the remaining 60% is in bacterial cells (Whitehead, 1986). On average, as excreted cattle feces is 5.2 g N kg-1, 1.9 g P kg-1, and 4.1 g K kg-1, for an N:P:K ratio of 2.7:1:2.2 (Edwards, 1996).

PAGE 30

30 Powell et al. (2006) describe fecal N pools as being either endogenous N or undigested feed N. The endogenous N pool consists of orga nic N forms which are readily available and contributes to crop N requirements the year of application. The undigest ed feed N mineralizes slowly in soil and is plant unavailab le until broken down by microorganisms. Effect on herbage production and nutritive value Studies have evaluated the impact of cattle dung on herbage production (Williams and Haynes, 1995). Dung deposits initially reduced he rbage yield owing to smothering. However, after 40 d, herbage around the edges of the dung patch responded positively to the dung and more dry matter was produced in this patch than in the contro l patch during th e first 12 mo. Dung applied to pearl millet resu lted in lower yields than excreta including both dung and urine that was collected from a corral where cattle and sheep were penned. This is due to the more readily available nutrients in the urine co mpared to the dung (Powell et al., 1998). Hirata et al. (1990) studied the effect of dung on bahiagrass herbage growth when dung pats from dairy cows were applied in June and August; there was also a control with no dung applied. The herbage dry matter growth rate was 0.2 g (0.04 m-2) d-1 from the initial application of the dung pat on 1 June until 29 June. The greatest herbage growth rate was observed from 1 to 12 July and was 0.8 g (0.04 m-2) d-1. The greatest herbage growth rate for the August dung application did not occur until th e following year when from 29 June to 12 July growth was 1.2 g (0.04 m-2) d-1. They found that some plants covered by the dung pat were killed, thus explaining the reduced herbage growth. In a greenhouse trial, feces was added to three different soils (Fine-loamy, mixed, superactive, frigid Oxyaquic Glossudalfs; fi ne-silty, mixed, superactive, mesic, Typic Argiudolls; coarse-loamy, mixed, superactive, frigi d, Haplic Glossudalfs) such that a total of 350 kg N ha-1 was mixed with 800 g of soil. Oat ( Avena sativa L.) and sorghum [ Sorghum bicolor

PAGE 31

31 (L.) Moench] were planted. Plant dry matter yiel d and N uptake were more affected by soil type than feces type (diets fed to animals producing fe ces for the study were corn silage-low CP; corn silage-high CP; alfalfa-low CP; alfalfa-high CP; Powell et al., 2006). Soil type accounted for 67, 65, and 82% of the variability in N uptake by oa t and sorghum; feces from the low CP diets decreased oat dry matter production likely due to N immobilization (Powell et al., 2006). Pearl millet was grown in pots and the so il was amended with fecal material (with fertilizer N or without fertilizer N) or with leaves. Dry matter yield was measured from 60 to 240 d after amendment application. Pearl millet amen ded with feces accumulated about 37% more P than when amended with leaves ( Acacia trachycarpa Combretum glutinosum Guiera senegalensis Pterocarpus erinaceus Pennisetum glaucum and Vigna unguiculata ), perhaps due to immobilization of P in the leaves. Feces with out fertilizer N did not affect pearl millet dry matter yield or N and P uptake, while feces with fertilizer N resulted in a yield response likely due to enhanced N mineraliza tion (Powell et al., 1999). Cherney et al. (2002) applied three differe nt nutrient treatments (NT) to orchardgrass ( Dactylis glomerata L.) and tall fescue ( Festuca arundinacea Schreb.). Treatments were inorganic N fertilization at 196 kg ha-1 (NT 1) or dairy cow manur e which was applied at two rates (X and 2X and termed NT2 and NT3) in 1995, 1996, and 1997. Rates of N applied in manure were 123 and 246 kg N ha-1 in 1995, 134 and 268 kg N ha-1 in 1996, and 143 and 286 kg N ha-1 in 1997. The relative N recovery in 1995 through 1997 for orchardgrass was 72, 58, and 45%, respectively, for NT1, NT2, and NT3. Tall fescue had a relative N recovery of 76% for NT 1, 58% for NT 2, and 48% for NT 3. They concl uded that perennial grasses can utilize large quantities of applied N, even N from dung.

PAGE 32

32 Fate of Nitrogen from Excreta A pplied to Grazed Grass Swards Nitrogen Cycle in Grazed Grasslands The major N pools in grazed grasslands ar e the soil, vegetatio n, grazing animals, and atmosphere. Fluxes among pools are a function of c limate, soil microbiota, forage species, and herbivores. Considering all terrestrial ecosystem s, the atmospheric N pool is 16 000 times greater than the sum of the soil and biotic N pools (Rus selle, 1996); however, it is available to plants only through biological N fixation (Marschner, 1995). In grasslands, the soil is the second largest N reservoir and is affected by soil OM, soil microbial biomass, fixed NH4 +, and to a lesser extent, plant-available inorganic N (Ste venson and Cole, 1999). The below-ground soil mesofauna are also important components of th e soil pool, and the rhizos phere may contain from 4500 to 24 000 kg N ha-1 (Henzell and Ross, 1973). These am ounts are far greater than the 20 and 400 kg N ha-1 reported in live herbage of tropical forages (Dubeux et al., 2007). Litter is another very important N pool, because along with the soil microbiota it constitutes the link between N in metabolically active plant tissues and N available for pl ant uptake (Thomas and Asakawa, 1993). Nutrients cycle among pools in the grassla nd and the rate of flux between pools is associated with fertilization pr actice, stocking rate, soil microf auna, climate, soil chemical characteristics, and plant species. Biological pathwa ys of N in pasture soils include nitrification and denitrification (Wrage et al., 2004; Clough et al., 2004). B oddey et al. (2004) found when stocking rate increased from 2 to 4 animals ha-1, N deposited as urine and dung in the paddocks increased from 50 to 90 and 37 to 59 kg ha-1, respectively. Some pathways of loss, such as denitrification and volatilization of NH3, can be reduced by activity of dung beetles and earthworms which incorporate feces into soil and eliminate anaerobic zones around the dung pats

PAGE 33

33 (Dubeux et al., 2007). The different pa thways of nitrogen loss will be discussed in greater detail in the following sections. Pathways of Nitrogen Loss from the Agroecosystem Loss of nutrients from urine to the environm ent occurs primarily via volatilization and leaching (Mathews et al., 1996). Urine spots on sandy soils have a lower ammonia loss due to higher infiltration (Rotz et al., 2005). Urine is particularly su sceptible to gaseous N losses because urine N is not bound to organic compounds (Sollenberger et al., 2002). Nitrogen loss from urine is typically greater than from dung spots (Rotz et al., 2005). Excreta (especially urine) contains a high concentra tion of soluble N that is very susceptible to either gaseous (ammonia volatilization, denitrific ation) or leaching losses. Much excreta N in grazed pasture is deposited in rest areas and around drinking troughs where the vegetation can be so trampled that very little N is recovered in forage production (Boddey et al., 2004). Grazi ng alters N cycling in ecosystems; it may alter fluxes of N to the atmo sphere and change the amount of the different chemical forms released (Pineiro et al., 2006). Gaseous losses Concentrations of atmospheric N2O have increased since preindustrial times (Rockmann et al., 2003), with one of the ma in contributions being from agri cultural soils (Perez et al., 2003; Clough et al., 2004). Excreta on grasslands repres ent high, very local additions of N, which creates optimal conditions for N2O emissions (van Groenigen et al., 2005b). Emissions are typically greater in grazed than ungrazed grassl ands (Clough et al., 2004) due to soil compaction by animals. Compaction can reduce water penetr ation, causing ponding and saturated soils and resulting in reduction of aeration and the increas e of anaerobic environments (Smith et al., 1998; van Groenigen et al., 2005b). Severity of compacti on is affected by soil te xture, and it is limited in sandy-soil environments like Florida. Ther e are several studies which have implicated

PAGE 34

34 compacted areas as hot spots of N2O emission from pastures (Ange r et al., 2003; Carran et al., 1995; and van Groenigen et al., 2005b). Alle n et al. (1996) reported a shift from N2O consumption to N2O production during warm conditions Clough et al. (2004) found when synthetic urine was applied to limed soil cores with different water-fi lled pore spaces (WFPS) there were significant differences in N2O emission and soil pH. Soils at field capacity (54% WFPS) that had a pH greater than or equa l to 5.9 produced the least amount of N2O, while the N2O flux was greatest in saturated (80% WFPS) treatments with a pH of 4.7. Under suitable conditions, i.e. mois ture, temperature, pH, etc., N2O can be reduced in the soil and released as benign N2 (Firestone, 1982; Clough et al ., 2004). van Groenigen et al. (2005b) evaluated treatments including urin e+dry soil, urine+dung, urine+compaction, and dung+soil with different water-filled pore spac e during an incubation period of 103 d at 16oC. Nitrous oxide fluxes were measured 27 times duri ng the 103 d. They reported a greater emission of N2O for the urine+dung and urine+compaction treatm ents, while urine+dry soil had the lowest emissions likely due to greater imme diate penetration into the soil. Ammonia volatilization is anot her important pathway of N emission from the grassland and will be greatest at high temperatures and hi gh soil pH (Boddey et al., 2004). Yokoyama et al. (1991) found when dung beetles colonized dung pats NH3 volatilization ceased during the first week; although, denitrification increased 23% and was greater than for the uncolonized dung pats (12.2%). Denitrification increased due to supplies of available energy and NO3-N and the consumption of O2 by microorganisms. It is reported that hippuric acid in urine has a controlling effect on both hydrolysis of urine N and on NH3 volatilization, which may also have an affect on N2O emission factors (Doak, 1952; Whitehead et al., 1989; van Groenigen et al ., 2005a). van Groenigen et al. (2005a) theorize

PAGE 35

35 it is possible to control hippuric acid concentratio n in urine by changing th e diet. Salts and other additives will dilute the N concentra tion in urine, leading to lower NH3 volatilization (Bussink and Oenema, 1998; van Groeni gen et al., 2005a). Leaching losses Nitrogen leaching is a major environmental c oncern in agroecosystems. Nitrate pollution of groundwater resources affect s drinking water quality and the conservation of natural ecosystems (Steenvoorden et al., 1986). Stout et al. (2000) noted the uneven recycling of N through urine could increase N leac hing and threaten water quality. Leaching is rainfall dependent and likely to be greater in free-draining, sandier soils (Boddey et al., 2004). Nitrogen leaching take s place when precipitation exceeds evapotransporation (Steenvoorden et al., 1986). Perennial grassla nds efficiently use soil water due to deep root systems and long growing season s and also protect the so il surface from erosion and runoff minimizing soil and nutrient loss (K emp and Michalk, 2005). Ru sselle et al. (2005) noted significant nitrate leaching occurs even in low N input systems because available N from excreta patches often exceeds plant uptake capacity. Hydrolysis is fairly rapid, approximately 1 d, resulting in high c oncentrations of NH4 + followed by the development of NO3 approximately 14 d after deposition (van Groenigen et al., 2005a). Denitrification losses Dung application may lead to anaerobicity due to high biological ac tivity and subsequent lowering of the redox potentia l (Monaghan and Barraclough, 1993; van Groenigen et al., 2005b). Manure N in organic form is not lost by de nitrification until af ter oxidation to the NO3 form (Edwards, 1996). Denitrification losses will only occur if there are anaerobic sites which will generally occur after heavy rainfall, and loss will be greater in soils with impeded drainage (Boddey et al., 2004). Denitrification is also pr omoted by high soil temperature, a low rate of

PAGE 36

36 oxygen diffusion, as well as the pr esence of soluble organic matter and nitrate (Luo et al., 1999). Luo et al. (1999) showed that th ere was no significant difference in denitrification rates between saturated and control cores co llected during the cool-wet se ason, but during the warm season denitrification rates were strongly enhanced by NO3 additions. Although the responses to added N in the warm, dry season were not as large as th ose in other seasons, it may be the lack of soil moisture controlled denitrification. Nitrification When urea is applied to soil, it reacts with wa ter in the presence of the urease enzyme and is rapidly converted to NH4 +, a process called hydrolysis (Griffith and Murphy, 1996). The remaining NH4 + is oxidized to NO2 (nitrite) through nitrification. Nitrite is either denitrified or lost from the system or oxidized to NO3 -, the latter being the dominant form of plant-available N in most soils. Nitrate can either be leach ed from the root zone or lost as N2O or N2. The conversion of NH4 + to NO3 leads to a decrease in pH over a period of approximately 2 wk (Doak, 1952; Haynes and Williams, 1992; van Groenige n et al., 2005a). Nitrite will accumulate under high pH following hydrolysis of urea due to Nitrobacter being inhibited (van Groenigen et al., 2005a). Recovery of Excreta Nitrogen by Grassland Plants Urine Leterme et al. (2003) conducted an experiment measuring the fate of urine N over three time periods (spring, summer, and autumn) when a pplied to perennial ryegrass cv. Belfort plots receiving two N fertilizer rates (100 and 300 kg N ha-1 yr-1). Three liters of spiked 15N urine was applied to ryegrass plots, and plant N uptake was determined. They found th at ryegrass recovery of N from urine ranged from 30 to 65%. The autumn application resulted in a relatively higher

PAGE 37

37 recovery of N, 49%, for the 100 kg N ha-1 yr-1 fertilizer treatment. The higher level of N fertilizer resulted in a decrease of urine N upt ake in above-ground parts of ryegrass. Ball et al. (1979) applied 300 (N300) and 600 kg N ha-1 (N600) as a mixture of urea and urine to perennial ryegrass-white clover plots in New Zealand. Th e urine was obtained from beef cattle fed pasture clippings in a barn. The ryeg rass-white clover sward had an apparent N recovery of 37 and 23% for N300 and N600, respec tively. Total herbage N yields increased with increasing N application, 143 to 202 kg N ha-1 as amount of N applied increased from 300 to 600 kg. Thompson and Fillery (1998) conducted three experiments applying sheep urine spiked with 15N at different times of the year to a rotation system including wheat ( Triticum aest ivum L.) pasture. Applications were either to pasture residues (Exp eriments 1 and 2), which were sown to wheat, or to growing pasture in winte r-spring (Experiment 3). In Experiments 1 and 2, urine was applied in November 1990, Apr il 1991, October 1991, January 1992, and March 1992 (9.8 g N m-2, 46.1 g N m-2, 4.6 g N m-2, 15.6 g N m-2, and 13.6 g N m-2, respectively). Wheat recoveries in November 1992 were 4, 7, and 12% of 15N applied. For Experiment 3, the urine was applied in August and September 1992 (12.3 and 25.9 g N m-2, respectively). Nine days after urine application in Augus t, 14% of the applied 15N was taken up by the pasture plants, and after 6 wk 53% had been recovered. In September, 47% of the 15N applied was recovered by the growing plants. Pakrou and Dillon (1995) collected urine from dairy cows and spiked it with 15N-labelled urea. The urine was applied to either irrigate d white clover-perennial ryegrass paddocks, which received 25 to 30 mm of irrigation in a single app lication per week, or nonirrigated subterranean clover ( Trifolium subterraneum L.)-based paddocks with annual grasses and weeds. To simulate

PAGE 38

38 a urination event, 11 mm of urine was applied in winter, spring, and summer. Paddocks were harvested 7, 28, 56, and 84 d each season after the urine application. In the winter, irrigated white clover-ryegrass total plant N recovered was 1% of applied 15N at 7 d, 2% of applied 15N at 28 d, 5% of applied N15 at 56 d, and 3% of applied N15 at 84 d. This compares with the nonirrigated subterranean clover plot s which had recoveries of 1% at 7 d, 3% at 28 d, 17% at 56 d, and 20% at 86 d. The lower 15N recovery in the irrigated paddocks can be attributed to the loss of 15N in leachate. In spring there was no difference in plant uptake of 15N among paddocks because sufficient soil moisture and active plant growth during spring and summer resulted in greater recovery of 15N from the spring application. Dung Dairy manure (dung + urine) was applied annually at four different ra tes over 2 yr (0, 75, 150, 300 kg N ha-1 in 1990 and 0, 150, 300, and 600 kg N ha-1) to orchardgrass managed for hay (Kanneganti and Klausner, 1994). Over the 2 yr average N recovery was 40%. In a single cutting the crop removed as much as 200 kg N ha-1 when 600 kg N ha-1 was applied; and tissue N concentration was 45 g kg-1. In a dung pat study, Dickenson and Craig (1990) determined that uncovered and covered dung still contained 86 to 95% of their original N content, resp ectively, over an 85-d period. This has an impact on herbage accumulation under and surrounding dung pats, because most of the N is still in the pats and not available to the plant. Impact on Soil and Water Soil Urine application on pastures ca n increase soil pH by 3 units within approximately 1 d due to hydrolysis of urea to NH4 + (van Groenigen et al., 2005a), although nitrification of the remaining NH4 + to NO3 leads to a decrease in pH over a 2-wk period (Doak, 1952; Haynes and

PAGE 39

39 Williams, 1992; van Groenigen et al., 2005a). Powe ll et al. (1998) found that when urine from rams was applied to pearl millet, the urine no t only increased soil pH from 4.5 to 9.5, but also increased available P, especially during the first week after application. The pH in the top 15 cm of the soil remained elevated for approximate ly 128 d, and Bray I-P appeared to decrease compared to the untreated control, perhaps in dicating movement downward of P. Not only were there decreases in soil ammonium levels in the urine patches 1 d after urine application, but also a large increase in soil nitrate le vels to a depth of 30 to 45 cm. When plots of Nandi setaria ( Setaria sphacelata var. sericea) received 1.4 L of urine (3748 g N m2), the greatest increase in pH in the 0to 0.5-cm soil layer occurred 2 to 6 h after urine application from an initial pH of 5 to a pH of 8, and after the second day there was a linear decline from pH of 8 to a pH of 4.5 (Vallis et al., 1982). Mineralization of urea occurred by 2 h after application, with only 200 g kg-1 of the original amount remaining, and by 14 h after application there was less than 20 g kg-1 present. This caused nitrate to accumulate after the first day, with a maximum NO3 concentration in the 0to 1.5-cm soil layer on the seventh day. Powell et al. (2006) reported feces applied to silt loams generally increased net soil inorganic N (IN), but when feces was applied to a sandy loam soil it caused net IN to decrease for 112 d followed by a rapid increase. This can be attributed to soil texture and chemical properties which had the greates t impact on soil N mi neralization. Another possible reason is when microbial populations are deprived of supp lied organic C input, biol ogically active pools of soil organic matter decline and this leads to a ne t release of soil nutrients. Also, when fed high protein diets cattle feces produced higher net soil IN than feces fr om low protein diets. Cows fed different diets (corn silage-low CP; corn silage -high CP; alfalfa-low CP; alfalfa-high CP) had

PAGE 40

40 different effects on soil IN, howev er, all fecal types applied to sandy loam soil had an initial 112 d of N immobilization followed by a gradual N mineralization to 365 d. Dai (2000) investigated the decomposition of dung over time following deposition by young cows on mixed-species temperate grasslan d. Total N concentration in the dung, after drying at 40oC for 48 h, was greatest (21 g kg-1) on Day 5; the greatest soil-N concentration occurred on Day 6 (16 g kg-1). Total soil N concentration outside the dung patch was 5 g kg-1. Dickenson and Craig (1990) appl ied dung pats to plots which were watered and either covered with transparent plastic sheets (250 x 250 mm) st retched horizontally between four pegs at a height of 150 mm or left uncovered (with no pl astic over the manure pile). Soil N concentration under the uncovered pats increase d initially but then declined. This result could be due to N being removed from the soil by plants at a gr eater rate than it was being added. The P concentrations in the soils in all three treatments were similar and very little P moved out of the dung pats. The final soil P concentrations rema ined below initial levels, possibly due to immobilization by microbes. Water Pastures reduce N loss in sedime nt and runoff water compared with annual crops (Rotz et al., 2005), however, the more nutrien ts added to a system above that which the forage is able to absorb results in build up in the soil and creates risk for runoff and water contamination (Sigua et al., 2004). Low intensity grazing of unfertili zed pasture land seldom causes problems in receiving waters, but pastures heavily fertilized with exogen ous nutrients can have serious impacts (Correll, 1996). In pasture systems in humid climates, subsurface water usually has larger NO3-N concentration and transport than su rface runoff, and has a greater need for evaluation (Owens et al., 1992).

PAGE 41

41 White (1988) reported that NO3 leaching losses from intensively managed grazed pastures can be in the range of 100 to 200 kg N ha-1 yr-1. Jabro et al. (1997) f ound during spring, summer, and fall applications of urine leached more NO3-N below 1 m (19, 15, and 25 g m-2, respectively) than the untreated control (1.7 g m-2). In comparison, when dung was applied in the summer of the first and second year, 2.2 g m-2 and 1.8 g m-2 were leached, respectively. Fate of Phosphorus from Excreta Applied to Grazed Grass Swards Phosphorus Cycle in Grazed Grasslands The P cycle is much more complex than the N cycle because the availability of P depends not only on biologically mediated processes of organic P but on the chemistry of inorganic P (Dubeux et al., 2007). Phosphorus exis ts in various chemical forms, inorganic P and organic P, which vary widely in their behavi or and fate in soils (Fuentes et al., 2006). The main sources of organic P include manure, crop residues, and sl udge. Organic P can be degraded into a long availability form of orthophosphate. Recovery of P in a growing season by plants can be very low; more than 80% may become immobile a nd unavailable for uptake because of adsorption, precipitation, or conversion to the or ganic form (Schachtman et al., 1998). Phosphorus in manure is in various forms but is mostly inorganic, indicating that P availability following application should be high because this P fraction in manure converts to plant-available P in a short period after ap plication (Eghball et al ., 2005). Phosphorus is accumulated at low pH mainly into the alkaliextractable AlP and FeP fractions which represent phosphate adsorbed to soil colloids (Haynes and Williams, 1993). Pathway of Phosphorus Loss from the Agroecosystem Leaching losses and runoff Phosphorus loss is much greater in surface r unoff than subsurface flow and is dependent on the rate, time, and method of P application; fo rm of fertilizer or manure applied; amount and

PAGE 42

42 time of rainfall after a pplication; and vegetati on cover of the land (Shigaki et al., 2006). Longterm application of manures and biosolids typica lly results in soil-P levels in excess of crop needs (Elliot et al., 2002). Phosphorus typically does not leach unless th e concentration in the soil is very high or the soil has very low retention capability. Graetz et al. (1999) measured P accumulation in manure-impacted soils in Florida. Sites were chosen to reflect a wide range of impact and included activ e dairies, dairies abandoned for years, beef cattle pastures, a nd areas not significantly impact ed by human activities (native areas). Pastures consisted of bahiagrass and be rmudagrass. Soil samples were taken of each horizon to a depth of at least 120 cm using a bucke t auger. These soil samples were analyzed for water soluble P (WSP), double-acid extractable P (DAP), and total P (TP). Native areas had less WSP in all horizons (1 mg kg-1) compared to the other areas. The high cattle densities in the abandoned dairies had WSP concentra tions averaging 45, 13, 16, and 8 mg kg-1 for the A, E, Bh, and Bw horizons, respectively. This indicated P m ovement vertically down the soil profile of high cattle density areas, and part of the P exis ted in a form readily removed by water. The DAP concentrations of native areas averaged 4 mg kg-1 compared to high cattle density areas (active and abandoned dairy) that averaged 707, 63, 238, and 72 mg kg-1 for the A, E, Bh, and Bw horizons, respectively. Double acid-extractable c oncentrations for the low cattle density areas averaged 16, 4, 34, and 18 mg kg-1 for A, E, Bh, and Bw horizons, respectively. The P forms (WSP, DAP, TP) in the pastures and forage ar eas tended to accumulate in the spodic horizon, which indicated vertical P movement in the low cattle density soil area, also. Soil pH in native areas was lowest, and it was the highest in the high cattle density areas, which can be related to high dung and soil Ca and Mg concentrations.

PAGE 43

43 Nair et al. (2007) found that comb ined root systems of pine ( Pinus elliotti ) and bahiagrass may absorb soil nutrients more completely than grasses alone in a treeless system. This conclusion was based upon the reduction of WSP in tree-based systems. This superior absorption resulted in greater P uptake in the silvopasture sy stem and less loss of nutrients to surface water. The silvopasture received 6 kg P ha-1 annually and was grazed; the treeless system was planted to bahiagrass and received a single i norganic P application of 6 kg P ha-1 in 2003 and was grazed. Soil samples were taken at a range of dept hs (0-5, 5-15, 15-30, 30-50, 50-75, and 75-100 cm). The treeless pasture had a greater Melich-1 P as depth increased up to 50 cm. This could be due to the silvopasture having a higher soil Al concentration which ranged from 350 to 1040 mg kg-1 compared to the treeless pasture (280-560 mg Al kg-1). This high concentration of Al will adsorb P and form non-available compounds. Phosphorus can move laterally from agricultural fields either in disso lved or particulate (attached to soil particles) fo rms (Elliot et al., 2002). When P is lost through runoff it causes negative environmental effects in surface water, such as eutrophication. The eutrophication of freshwater by increased inputs of P from agricu ltural runoff is a concer n in many areas of the USA, particularly Florida (S arkar and OConnor, 2004). Arthi ngton et al. (2003) found that using three different stocki ng rates (1.49, 2.63, and 3.48 ha cow -1) of pregnant Brahman cows did not affect concentrations or loads of total P or N in runoff from planted summer or mixed winter pastures. They also f ound that control pastures, cont aining no cattle, provided similar amounts of total P and N in runoff water compar ed to pastures containing cattle, although it should be noted that these stocking rates are very low. The summer pastures consistently delivered greater P loads (1.28 kg ha-1) in runoff than the wi nter pastures (0.03 kg ha-1) because

PAGE 44

44 of greater P fertilization for at least 15 to 20 yr, even though P fertilization was discontinued in 1987, 12 yr before the start of this study. In many soils, high P-sorbing oxide compone nts keep leachate P levels well below eutrophication thresholds (0.01 to 0.05 mg L-1); vertical flux of P is potentially significant in areas with shallow ground water and coarse-textured soils with little P-sorbing capacity (Elliot et al., 2002). Phosphorus loss is a concern on sandy, high water table soils with limited P-holding capacity (Mathews et al., 1996), lik e those in Florida. Phosphorus loss increases with input rate or soil test P level (Rotz et al., 2005). Phosphorus soil test values over 330 kg P ha-1 (Mehlich 1) in the upper 20 cm of soil greatly increases the chance for loss from the system, which will affect water quality (Sigua et al., 2004). Mineralization Mineralization of organic P compounds repres ents an important P source for plants, especially in soils with low levels of bioavail able P (Fuentes et al., 2006). Mineralization of organic P in the soil is cata lyzed by various enzymes (phosphat ases), including phytase (Eghball et al., 2005). These enzymes play a fundamental role in the P cycle a llowing orthophosphate to be released from organic and inorganic compound s and increasing the bioavailable P (Fuentes et al, 2006). Mineralization is greatest when soil moistu re is near field capacity and declines with soil drying. For P to become available for pl ants, phosphatase enzymes must breakdown the organic P compound (i.e., phospha te diester-nucleic acids, phosphoprotein, phospholipids, or phosphate monoester-glucose-6-pho sphate, nucleotides) into phosphate (Fuentes et al., 2006). Recovery of Excreta Phosphorus by Grassland Plants It has been reported that dry beef cattle ma nure contains between 2.94 (Iyamuremye et al., 1996) and 4.02 g kg-1 of P (Griffin et al., 2003). Manure cont ains significant amounts of P that can be utilized for crop production (Eghball et al ., 2005). Wilkinson and Lowrey (1973) reported

PAGE 45

45 that P cycling through grazing animals was ineffici ent in the short term because of negligible return of P through urination, the small area of coverage by manure each grazing season, and the low mobility and spatial unavailability of P in manure (Peterson and Gerrish, 1996). There have been experiments showing benefits of excreta application to pastures, but these impacts have not exclusively been linked to P addition. One such study by Dalrymple et al. (1994) reported forage yield increases over a no-excreta control of 220 and 880 kg ha-1 in areas affected by manure piles and urinations, respec tively (Peterson and Gerrish, 1996), although this benefit was not due solely to P addition. Du ring and Weeda (1973; cited by Peterson and Gerrish, 1996) determined that a cattle dung pat affected forage grow th in a zone 5-fold larger than the area covered by the pat. Forage P yields around the 0.25-m2 area increased 23%. Summary The southeastern USA is an important area of livestock and bahiagrass production. Because of increasing cost of fertilizer nutrien ts and growing concern regarding the impact of agricultural systems on the environment, there is a need for greater understanding of nutrient relationships in grazed pasture. There is little information in the literature evaluating the impact of type and number of excreta applications to pasture on leaching of N, changes in soil nutrient concentration, and herbage growth and nutritive value. The objectiv e of the research reported in the chapters that follow is to characterize the e ffects of cattle excreta a pplication on 1) bahiagrass yield, chemical composition, and nutrient recove ry and 2) nitrate leaching to shallow ground water.

PAGE 46

46 CHAPTER 3 BAHIAGRASS HERBAGE DRY MATTER HARV ESTED, NUTRIENT CONCENTRATION, AND NITROGEN RECOVERY FOLLO WING EXCRETA DEPOSITION Introduction Grazed bahiagrass ( Paspalum notatum Flgge) pastures comprise approximately 1.1 million hectares and 75% of the planted perennial pasture resource in Florida (Gates et al., 2004; Mislevy et al., 2005). Bahiagrass is adapted to low soil nutrient levels and limited amounts of fertilizer are applied to grazed swards (Chambli ss and Adjei, 2006). Despite this, these areas are heavily scrutinized for their pote ntial environmental impact. This is due to the large land area planted to this forage and to the sensitivity of Florida ecosystems, especially in terms of nutrient impacts on water quality (Nair et al., 2007). On agri cultural lands, the lateral flow of P to surface water, particularly in Central and South Florida (Nair et al., 2 007), and leaching of nitrates to ground water, more commonly in North Florida, are critical concerns (Woodard et al., 2002). In low-input bahiagrass pastures in Florid a, animal excreta and decaying plant matter (litter) are the major tr ansitory nutrient pools, and the importance of excreta vs. plant litter increases as stocking rate incr eases (Thomas, 1992). Chemical tr ansformations that occur during digestion and excretion of plant nutrients by herbivores cause nutrients in animal excreta to be more available for subsequent uptake by plants a nd more susceptible to loss to the environment than those in plant litte r (Bardgett and Wardle, 2 003; Boddey et al., 2004). Efficient recovery of these nutrients is hindered by the large quantity supplied in a single excreta event and because a disproportionately large number of these events occur in sma ll areas where cattle congregate, e.g., near to shade, water, and supplemental feed sources (Mathews et al., 2004; Sollenberger et al., 2002). It has been suggested that rotational stoc king with short grazing periods, i.e., many paddocks per pasture, decreases the opportunity and tendency of animals to congregate in

PAGE 47

47 lounging areas by intensifying competition for f eed and shortening resi dency periods (Haynes and Williams, 1993). In Florida, it was found that rotationally stocked pastures where grazing periods were short (1 to 7 d) had greater spat ial uniformity in time spent by cattle, excreta deposition, and soil nutrient c oncentration than continuously stocked pastures (Dubeux, 2005). Thus, it is likely that greate r uniformity of excreta deposi tion can be achieved by imposing rotational stocking with short grazing periods. Whether this intensification of grazing management results in more efficient nutrient cycling may depend on the degree to which more uniform excreta deposition enhances nutrient rec overy by grassland plants and avoids excessive nutrient accumulation in soils or nutri ent loss to surface or ground water. There is little information in the literature evaluating the imp act of type and number of excreta applications to pasture on leaching of N, recovery of nutr ients in harvested herbage, and changes in herbage growth and nutritive value. Th e objective of this research was to characterize the effects of cattle excreta application on 1) bahiagrass yield, chemical composition, and nutrient recovery and 2) nitrate leaching to shallow ground water. Materials and Methods Experimental Sites There were two sites used for this study. One location served as the source for excreta and the other site was used for excreta applicati on to bahiagrass. Excret a source pastures were located at the University of Florida Beef Research Unit, northeast of Gainesville, FL (29.72o N latitude, 82.35o W longitude). This site was chosen because well-established Pensacola bahiagrass pastures, fencing, and animals were readily available. Soils at this site were classified as Spodosols of the Pomona (sandy, siliceous, hype rthermic Ultic Alaquods) and Smyrna series (sandy, siliceous, hyperthermic Aeric Alaquods). Average soil pH was 5.5, and Mehlich-I extractable P, K, Mg, and Ca were 17, 29, 50, and 392 mg kg-1, respectively.

PAGE 48

48 A long-term, ungrazed stand of Pensacola bahiag rass at the Plant Science Research Unit, Citra, FL (29.41o N latitude, 82.02o W longitude) was used for a pplication of excreta. The ungrazed area was used for this purpose so that amount and frequency of excreta applications could be controlled. Also, this location was c hosen because the soils are well-drained sands, avoiding potential subsurface late ral flow of water and nutrien ts that could occur under the Spodosols at the Beef Research Unit. This coul d result in mixing of nutrients leaching from adjoining plots and preclude the drawing of meaningful conclusions regarding water quality from the treatments imposed. Also, this site had not be en fertilized for the past 12 yr, so there was minimal carryover effect of previous treatments. Th e soils at the application site are classified as Tavares or Candler fine sands (sandy hypert hermic, uncoated Typic Quartzipsamments). Average soil pH was 5.5, and Mehlich-I extractable P, K, Mg, and Ca were 34, 70, 18, and 164 mg kg-1, respectively. Treatments and Design Treatments were two excreta sources (from he re forward referred to as Average or High pasture management intensity), two types of ex creta (dung and urine), an d three frequencies of excreta application (one, two, or three times pe r year). Average and High management intensity were defined based on N fertilizer amount and stocking rate. Average management pastures received 60 kg N ha-1 yr-1 and were stocked with two yearling heifers ha-1, and High pastures received 120 kg N ha-1yr-1 and were stocked with four yearling heifers ha-1. These two management intensities were selected to repres ent fertilization and st ocking regimes that are common in the Florida livestoc k industry (Average) or repr esent approximately the most intensive management applied to grazed bahiagrass (High). In addition, based on previous work by Stewart et al. (2007) it is exp ected that these treatments will re sult in forage that varies in

PAGE 49

49 nutritive value, especially N concentration, and that these differences likely will affect the composition of dung and urine. The 2 x 2 x 3 factorial resulte d in 12 treatments. In additi on there were three control treatments that received no excreta and we re fertilized with N at 0, 60, or 120 kg ha-1 yr-1. The 15 treatments were replicated three times in a ra ndomized complete block design. Plots were 3 x 3 m in area with a 1-m bahiagrass alley surrounding each plot. Excreta Source Pastures There were two replicates of each management intensity (Average and High) arranged in completely randomized design. Pastures were stoc ked continuously and past ure size was 1 ha for Average and 0.5 ha for High. Each pasture was grazed with two crossbred yearling heifers (Angus x Brahman) with average initial weig ht of 408 kg, and animal care was monitored according to Institutional Animal Care and Us e Committee Protocol Number D655. Heifers were provided with access to water, trace mineral mix (min imum concentrations of Ca, 12; P, 6; NaCl, 19; K, 0.8; Mg, 1.0; S, 0.4; and Fe, 0.4 g [100 g]-1), and artificial shade (3.1 x 3.1 m) on all treatments. All pastures were fertilized with 17 kg P and 66 kg K ha-1 on 15 May 2006. The Average treatment received 60 kg N ha-1 yr-1 in two equal applications of 30 kg ha-1 that were made on 15 May and 15 July 2006. On the same days, the High treatment received 60 kg N ha-1 for a total of 120 kg N ha-1 yr-1. Grazing was initiated on 26 May 2006. Fe rtilization occurre d in advance of initiation of grazing so that bahiagrass herbage would reflect treatment effects by the time cattle first entered the pastures. The pastures were sampled to characteriz e herbage mass and nutritive value during each excreta collection period. Herbage mass was quant ified using double sampli ng with the indirect sampling tool being a disk meter (Stewart et al ., 2005). On each of the four pastures, five double

PAGE 50

50 samples and 20 disk heights were taken at each sampling date. Herbage mass was predicted from disk heights, and a single ca libration equation was used acros s all pastures and the three sampling dates. The equation was herbage mass (kg ha-1) = 340 (disk height [in cm]) 137 (r2 = 0.821). At the same sampling times, hand-plucked samples of the top 10 cm of the forage canopy were taken to represent cattle diets. Samples were collected from 30 representative locations per pasture, composited, dried at 60C, and analyzed for N and P concentration using a microKjeldahl technique (Gallaher et al., 1975) and for in vitro di gestibility using a two-stage procedure (Moore and Mott, 1974). Excreta Collection and Analyses The yearling heifers used for grazing the excr eta source pastures were selected for docile temperament following careful screening of a larg e group of animals of similar age and weight. Initially 12 heifers were chosen. Du ring the winter and spring before the start of the trial, they were fed hay and small amounts of concentrate se veral times each week to develop a routine of close proximity to people. Based on their respon se to this interaction, the group was reduced to eight heifers, the number needed for the grazing study. Stanchions were erected in a corral near the pastures, and animals were moved to the corral several times per week during spring before the start of the trial. Concentrate feed was placed in troughs in front of the stanchions so that heif ers were required to put their heads into the stanchions to gain access to the feed. After se veral weeks of this process, the animals were locked in the stanchions to become accustomed to short periods of restra int. This activity was suspended at least a week before dung and urine collection began. During the trial, animals were brought to th e stanchions for collection of urine and dung. Urine and dung were collected during three peri ods (12-16 June; 26-28 July; and 7-8 Sept. 2006) leading up to the three dates (20 June, 1 Aug., and 13 Sept. 2006) when excreta was applied to

PAGE 51

51 the plots. Application dates were at 6-wk inte rvals. Acquiring suffici ent dung and urine required two, 4-h periods (4 h on each of 2 d) of collection per animal prio r to the first application date, (all plots receiving one, two, or three applications per year rece ived excreta on 20 June). One, 4h period of collection per animal was required prio r to the latter application dates (only plots receiving two or three applica tions per year received excr eta on 1 August and only those receiving three applications were treated on 13 September). Animals were shaded and had access to water during the collection period. For collection, the two pairs of animals fr om a given intensity treatment (i.e., both replicates) were brought from the pasture at 0700 h and locked in the stanchions (two pairs of two animals). To facilitate catching the animal s in the stanchions, the animals were provided with a small amount of grain (~ 200 g per heife r). For the first collection period, when 2 d of collection were required for each group of animal s, 3 d were allowed between collections to ensure that effects of the concen trate fed during the first collecti on were not present in the second excreta collection. To collect excreta, one person was positioned behind each group of two animals. Collection was accomplished using a fishi ng net that was mounted at the end of a 1.5-m wooden rod. A trash compacter bag was inserted in the net to collect the dung or urine. After an excreta event was caught, the dung or urine was immedi ately put into sealed containers and taken to a refrigerator for storage at 4C. Prior to a pplication, all excreta events of a given type (dung or urine) from one intensity treatment (four an imals) were composited across field replicates. Thus, at time of application there were two t ypes of dung and two types of urine (one of each excreta type from both Average and High treatmen ts). Three subsamples were taken from the composited samples of urine and dung and analyz ed to determine excreta chemical composition so that amount of nutrient applie d to plots could be calculated.

PAGE 52

52 During the June excreta collection, samples were taken to ensure that chemical composition did not change markedly during the 8 d of excreta storage prio r to application, i.e., to ascertain that what was appl ied to the plots was similar in composition to what the animal would deposit fresh. This was accomplished by taking two subsamples from each of two separate urine and dung events. One of the two subsamples from each event was analyzed immediately; the other subsamples were stored in the same manner as the urin e and dung that were eventually applied to the plots. At the date of excreta application to plots, these stored subsamples were analyzed to assess changes in chemical com position during the storage period (Table 3-1). Urine samples for analysis were acidified with concentrated sulfuric acid to pH 2 to avoid N volatilization. Dung was analyzed for total N, organic N, NH4-N, P, and K, and urine was analyzed for N, NH4-N, urea-N, P, and K. Dung total N was analyzed using the Kjeldahl methodAOAC 984.13; NH4-N analyzed using Distilla tion AOAC 941.04; organic N by difference (Total N Ammonia-N); and P and K were analyzed using Thermo IRIS Advantage HX Inductively Coupled Plasma Ra dial Spectrometer. Urine total N was analyzed using AOAC 2001.11 Block digestion and Foss 2300 or 2400 Analyzer; urea and ammonia AOAC 941.04, Analysis by Thermo IRIS Advantage HX I nductively Coupled Plasma (ICP) Radial Spectrometer; and for P and K by Thermo IR IS Advantage HX Inductiv ely Coupled Plasma (ICP) Radial Spectrometer.

PAGE 53

53 Table 3.1. Chemical composition of excreta anal yzed immediately after collection (fresh) and after storage (stored) for up to 8 d at 4oC. Each value is the mean of the analysis of two subsamples. Urine Dung Constituent Fresh Stored Fresh Stored ---------------------------------------g kg-1 --------------------------------------Total N 0.41 0.49 2.13 2.07 NH3-N 0 0 0.075 0.070 Urea 0.29 0.26 Organic N 2.05 2.00 Total P 0.0024 0.0029 0.75 0.72 Total K 0.90 0.90 1.42 1.37 Not analyzed for this constitu ent within this excreta type. Plot Management All ungrazed plots to which excreta would subs equently be applied were fertilized on 7 June with 18 kg P and 66 kg K ha-1. Nitrogen application de pended upon excreta-source treatment. Plots that received ex creta from the High management intensity source pastures were fertilized with 120 kg N ha-1 yr-1, split equally in two applications of 60 kg N ha-1 (7 June and 16 Aug. 2006). Plots that received excreta from th e Average management intensity plots were fertilized with 60 kg N ha-1 yr-1, in two applications of 30 kg N ha-1 (on the same dates as High). Following fertilization on 7 June, the entire plot area was irrigated with 25 mm of water because of spring drought and lack of early summer ra infall (rainfall in Marc h through May at this location was 68 mm compared to a 30-yr average of 279 mm). Magnesium sulfate was applied to all plots on 6 July 2006 to address low soil Mg levels. It was applied at 135 kg ha-1 to provide 27 kg Mg ha-1 and 36 kg S ha-1. Excreta Application All plots were staged to a 5-cm stubble on 19 June 2006. All experimental units except the no-excreta controls received dung or urine on 20 June. Subsequent a pplications were on 1 August and 13 September. Only plots receiv ing two or three excreta applications yr-1 received

PAGE 54

54 excreta on 1 August, while on 13 September only those plots receiving three applications yr-1 were treated. Quantity of dung and urine and the area to which they were applied were determined based on values reported in the literature (Haynes and Williams, 1993; Table 2-1) and personal observation of the cattle. Two liters of urine cons tituted one application, a nd it was applied to an area of 0.283 m2, a 60-cm diameter circle, w ith the center of the circle being the center of the plot (Fig. 3-1). Of the 2 L, 1 L was applied to a 30-cm diameter circle with the center being the center of the plot (area of 0.071 m2), and 1 L was applied to the area outside the 30-cm diameter circle but inside the 60cm diameter circle. This was done to reflect the likelihood of greater concentration of urine closer to the center of the affected area The urine was applied using a watering can with a sprinkler head so that runoff from the 60-cm diameter circle was minimal. Dung was applied at 2 kg fresh weight to an area of 0.071 m2 (Fig. 3-1). This was a 30-cm diameter circle with the cente r being the center of the plot. The dung was evenly distributed across this area. Subsequent applications (for th e two and three applicatio ns per year treatments) occurred at the same location in the plot.

PAGE 55

55 Figure 3-1. This drawing (not to scale) shows one plot or expe rimental unit and the areas to which urine and dung were applied. In plots where urine was applied, 1 L was applied to the area labeled 30 (inside a 30-cm diamet er circle) and 1 L to the area labeled 60 (the area outside the 30-cm diameter circle but inside the 60-cm di ameter circle). In plots where dung was applied, 2 kg fresh we ight of dung was applied to the area labeled 30. Forage Harvest and Herbage Analyses Forage was harvested every 28 d following in itial excreta application (20 June), and harvest dates were 18 July, 15 Aug., 12 Sept., and 10 Oct. 2006. To characterize the effect of the excreta application on herbage yield and nutritive value, herbage was clipped at each harvest date to a 5-cm stubble in a 90-cm diameter circle (0.636-m2 area), the center of which was the center of the circles of dung or urine app lication. Harvested herbage was dried at 60oC to a constant weight and weighed. It was subsequently ground to pass 1-mm screen and analyzed for 30 60

PAGE 56

56 N and P concentration using a micro-Kjeldahl te chnique (Gallaher et al ., 1975) and for in vitro digestibility using a two-stage pr ocedure (Moore and Mott, 1974). Lysimeter Placement, Water Sampling, and Water Analyses One lysimeter, designed as described by Woodard et al. (2002), was pos itioned in each plot such that the ceramic cup was directly below the center of each plot and 1.3 m below soil level. The PVC pipe entered the soil at approximately 40 cm from the center of the excreta deposition site (beyond the area to which e ither dung or urine were applied) and angled through the soil so that the ceramic cup was properly placed bene ath the center of the plot and the excreta application site. This approach allowed the sod immediately around the excreta deposit to be undisturbed. This depth for placement of the ly simeters cup was chosen because it is below nearly all grass roots. The intent was that lysimeters be sampled every 14 d or after major rainfall events (> 25 mm). During dry periods when 14-d sampling was not possible because of inadequate soil water, re-initiation of sampling was triggered by major rainfall events (> 25 mm) that resulted in sufficient soil water so that the lysimeters could hold a vacuum. Two days prior to sampling, any water contained in the lysimeter was evacuated, th en a suction (40 to 45 kPa) was placed on it. At sampling, suctions were released. Samples were acidified (pH of < 2) and placed in a cooler within 15 min of extraction. Water sampling between 20 June 2006 and 2007 occurred on 23 and 28 June, 10 and 24 July, 8, 22, and 29 Aug., 15 and 27 Sept., 12 and 19 Oct., 19 Nov., and 29 Dec. 2006, and 5 and 26 Jan., 5 and 19 Feb., and 23 Apr. 2007. The water was analyzed for NO3N and total P. Occasionally throughout the gr owing season, samples were analyzed for NH4-N to insure that there was no water moving down along-side the external wall of the lysimeters and transferring fertilizer N direc tly to the collection cup. At no time were there measurable concentrations of P or NH4-N, so these data are not presented.

PAGE 57

57 Data Presentation and Statistical Analyses Herbage dry matter (DM) harvested is expresse d on a total-season basi s, i.e., the sum of four harvests. Tissue N, P, and in vitro diges tible organic matter (IVDOM) concentrations are the weighted averages across four harvests. Nitr ogen and P harvested were calculated for each harvest date (product of DM ha rvested and nutrient concentrati on in harvested herbage) and summed across the four harvests for a total-seas on measure. Excreta N recovery was calculated by summing N harvested across the four harvests for a given excreta treatment and subtracting from this sum the total seasonal N harvested from the appropriate (Average or High) no-excreta control. This number was then divided by excreta N applied to that plot and the result expressed as a percentage. Data were analyzed using PROC MIXED in SA S (SAS Institute, Inc. 2007). For herbage mass and nutritive value from the excreta sour ce pastures, management intensity was a fixed effect and replicate was a random effect. Da te was considered a repeated measure. For herbage data from the excreta applica tion experiment, an initial analysis was conducted using the 12 factorial treatment combin ations (two management intensities, two excreta types, and three applic ation frequencies). These factor s and their interactions were considered to be fixed effects and replicate a ra ndom effect. Polynomial contrasts were used to determine the nature of the res ponse to application frequency. The excreta type X application frequency in teraction was significant for six of seven herbage response variables for the fa ctorial data set. To more completely explore this interaction and to integrate the no-excreta controls, a fo llow-up analysis was conducted. This approach considered the no-excreta control treatments fo r the Average and High management intensities (60 and 120 kg N ha-1 yr-1) as a zero level of excreta applic ation frequency, providing a fourth level of application frequency (0, 1, 2, and 3) Because of the presence of excreta type X

PAGE 58

58 application frequency interaction in the initial an alysis, these data were analyzed by excreta type and polynomial contrasts were used to determin e the nature of the response to application frequency. Water data were analyzed by sampli ng date using the same statistical approach described for herbage data from th e excreta application experiment. An additional follow-up analysis was carried out for the data from the three control treatments. In this case, the treatment was N fe rtilizer amount, so the effect of N on herbage responses was determined using polynomial contrasts. For all response variable from both experiment s, differences were deemed significant when P 0.10. Data presented are least squares means. Results and Discussion Characteristics of Excreta Source Pastures There were no interactions of sampling da te and management intensity for herbage characteristics of excreta source pastures, but in teraction means are presented so that excreta nutrient concentrations can be considered relative to pasture characteristic s at time of excreta collection (Table 3-2). There were no manageme nt intensity effects for herbage mass (3030 kg ha-1), herbage P concentration (2.0 g kg-1), or herbage IVDOM concentration (520 g kg-1), but herbage N concentration was greater for High th an for Average management intensity (17.1 vs. 14.4 g kg-1). Sampling date affected herbage mass and IVDOM only, although there was a trend ( P = 0.117) toward greater herbage N concentration in July than in the other months. Mass was similar in June (2820 kg ha-1) and September (2810 kg ha-1), but was greatest in July (3470 kg ha-1). The second application of N fertilizer occurred on 19 July, 10 d after the second application of N fertilizer, and this timing along with typical seas onal growth patterns (Stewart et al., 2007) likely explain the greater herbage ma ss in July and the trend toward greater N concentration. Herbage IVDOM was simila r in June and July (558 and 540 g kg-1) but was least

PAGE 59

59 in September (461 g kg-1). When evaluating two intensities of management of bahiagrass pastures, Stewart et al. (2005) reported le sser bahiagrass herbage mass (2.87 vs. 3.42 Mg ha-1) but greater N concentration (18.1 vs. 15.8 g kg-1) on pastures that were fertilized with 120 kg N ha-1 yr-1 and stocked at 2.4 animal units ha-1 than those that received 40 kg N ha-1 yr-1 and were stocked at 1.4 animal units ha-1. These treatments are similar to High and Average intensities in the current study, and herbage N concentration response was sim ilar in the two experiments. Although there was no intensity e ffect on herbage mass in the cu rrent study, herbage mass means were in a similar range to those reported by Stewar t et al. (2005). Excreta Composition Excreta composition for applied dung (Table 33) and urine (Table 3-4) varied across dates. The elevated N concentration in urine dur ing the second collection (Table 3-4) can be attributed in part to the trend ( P = 0.117) toward greater herbag e N concentration during that period (Russelle, 1996), but the large increase in urine N concentration cannot be accounted for by this relatively minor change in diet N. Urine N concentration is highly responsive to diet N, and Valk and Hobbelink (1992; reported in Russe lle, 1996) found that lactating dairy cows fed diets with greater energy concentrations (l esser N concentrations) had decreased urine N concentration.

PAGE 60

60 Table 3-2. Herbage mass, N, P, and in vitro digestible organi c matter (IVDOM) concentrations of bahiagrass herbage during excre ta collection periods in 2006. Herbage mass (kg ha-1) N concentration (g kg-1) P concentration (g kg-1) IVDOM (g kg-1) Management intensity June July Sep. June July Sep. June July Sep. June July Sep. Average 2840 3620 2850 14.5 14.9 13.6 1.84 2.00 2.03 550 547 461 High 2810 3320 2770 16.9 17.6 16.5 1.95 2.08 2.08 567 534 461 P value 0.94 0.42 0.88 0.310.28 0.06 0.79 0.48 0.76 0.480.141.0 SE 259 166 459 0.910.91 0.53 0.24 0.05 0.10 27.9 5.2 7.8 Average refers to pastures fertilized with 60 kg N ha-1 yr-1 and stocked at 2 yearling heifers ha-1, while High refers to pastures fertilized with 120 kg N ha-1 yr-1 and stocked at two yearling heifers ha-1.

PAGE 61

61 Table 3-3. Composition of fresh dung from Aver age and High management intensity source treatments during three collection periods in 2006. Each value reported is the average across three subsamples from a composite dung sample. Dung was composited across replicates within a management intensity treatment, so statistical comparisons of treatment effects are not possible. Dung Application Date Nutrient 20 June 1 August 15 September Average High Average High Average High g kg-1 fresh dung Total N 1.97 2.10 2.03 2.13 2.25 1.85 NH3-N 0.07 0.08 0.08 0.10 0.09 0.08 Organic N 1.89 2.02 1.95 2.03 2.16 1.77 Total P 0.64 0.74 0.71 0.67 0.72 0.57 Total K 1.37 1.49 1.32 1.32 1.46 1.43 Average refers to pastures fertilized with 60 kg N ha-1 yr-1 and stocked at 2 yearling heifers ha-1, while High refers to pastur es fertilized with 120 kg N ha-1 yr-1 and stocked at two yearling heifers ha-1. Table 3-4. Composition of urine from Aver age and High management intensity source treatments during three collection periods of 2006. Each value reported is the average across three subsamples. Urine was co mposited across replicates within a management intensity treatment, so statis tical comparisons of treatment effects are not possible. Urine Application Date Nutrient 20 June 1 August 15 September Average High Average High Average High g kg-1 urine Total N 1.60 1.58 4.04 2.97 0.98 1.88 NH3-N 0.00 0.00 0.00 0.00 0.00 0.00 Urea-N 0.84 0.88 1.82 1.71 0.37 0.86 Total P 0.0033 0.0042 0.0049 0.0047 0.0027 0.0049 Total K 2.20 2.39 3.77 2.76 1.38 2.27 Average refers to pastures fertilized with 60 kg N ha-1 yr-1 and stocked at 2 yearling heifers ha-1, while High refers to pastur es fertilized with 120 kg N ha-1 yr-1 and stocked at two yearling heifers ha-1. Nutrients Applied in Excreta Quantity of N, P, and K ha-1 applied at each of three application frequencies for Average and High management intensities are shown in Tables 3-5 (dung) and 36 (urine). Amount of nutrient applied per unit land area is greater for dung than urine because of the smaller land area to which dung was applied.

PAGE 62

62 Table 3-5. Nutrients applied to dung treatments in bahiagrass swards. Calculations are based on chemical analyses of fresh dung (Table 3-3) and a 2-kg fresh wei ght dung application to a circle of 30-cm diameter. Data are expressed as kg of N, P, and K applied ha-1. Date Source Frequency 20 June 1 August 15 September Total kg N-P-K ha-1 1 557-182-389 0-0-0 0-0-0 557-182-389 2 557-182-389 574-201-373 0-0-0 1131-383-762 Average 3 557-182-389 574-201-373 636-204-415 1767-587-1177 1 594-209-422 0-0-0 0-0-0 594-209-422 2 594-209-422 603-190-373 0-0-0 1197-399-795 High 3 594-209-422 603-190-373 523-164-405 1720-563-1200 Table 3-6. Nutrients applied to urine treatments in bahiagrass swards. Ca lculations are based on the chemical analyses of urine (Table 34) and a 2-L volume of urine applied to a circle of 60-cm diameter. Data are expressed as kg of N, P, and K applied ha-1. Date Source Frequency 20 June 1 August 15 September Total kg N-P-K ha-1 1 113-0.23-267 0-0-0 0-0-0 113-0.23-267 2 113-0.23-267 286-0.35-267 0-0-0 399-0.58-534 Average 3 113-0.23-267 286-0.35-267 69-0.58-98 468-1.16-632 1 112-0.30-169 0-0-0 0-0-0 112-0.30-169 2 112-0.30-169 210-0.33-195 0-0-0 322-0.63-364 High 3 112-0.30-169 210-0.33-195 133-0.35-160 455-0.98-524 Herbage Dry Matter Harvested from Excreta-Treated Plots Total-season herbage DM harvested wa s affected by management intensity ( P < 0.001) and excreta type ( P < 0.001) main effects and by the excreta t ype X application fr equency interaction ( P = 0.063) (Table A-1). For plots receiving excreta, as manageme nt intensity increased from Average to High, forage DM harvested increased from 3230 to 3850 kg ha-1. Deenen and Middelkoop (1992) found the extent to which ur ine affected herbage growth of perennial ryegrass was dependent upon the level of N fertili zer applied. There was no effect of urine in swards receiving 400 kg N ha-1, but positive effects on yield were noted in swards receiving 250 kg N ha-1. In the current study, there were no interact ions involving management intensity (i.e.,

PAGE 63

63 N fertilizer level), so the response to excreta was consistent across levels of management intensity. One reason for the lack of interaction in the curren t study may be that fertilizer N levels applied were much less than thos e used by Deenen and Middelkoop (1992). For the three control treatments to which no excreta was applied in the current study, the DM harvested response to N application had lin ear and quadratic terms, increasing from to 1310 to 2450 to 3060 kg ha-1 as N fertilizer applied increas ed from 0 to 60 to 120 kg ha-1 (Table 3-7). Responses to management intensity in both excr eta-treated and no-excret a plots were primarily due to the differences in N fertilization and ar e consistent with bahiagrass responses to N fertilizer observed in the liter ature (Burton et al., 1997; Twidwe ll et al., 1998). Total-season DM harvested in the current study was less than repo rted in many previous studies including those by Burton et al. (1997), Twidwell et al. (1998), and Interrante et al (2007). The two primary reasons for low yields in the current research are rainfa ll and previous management at the plot site. Rainfall totaled 722 mm during March through October 2006 compar ed to a 30-yr average of 962 mm (Table 3-8). In addition, the research site had not been fert ilized for at least 12 yr, so overall stand vigor was relatively lo w at the beginning of the trial a nd this affected first-year DM harvested. There was excreta type X application frequenc y interaction for bahi agrass DM harvested. Interaction occurred b ecause there was no effect of dung appl ication frequency on DM harvested (Fig. 3-2), but there were linear and quadratic effects of urine application frequency. The response to urine incr eased from 2760 kg ha-1 at zero applications to 4670 kg ha-1 with three applications. The quadratic effect was significan t because the greatest increase in DM harvested was with the first urine application, while addi tional urine applications had less effect on DM

PAGE 64

64 harvested. For each level of excreta application frequency from one through three, urine-treated plots outyielded dung-treated plots ( P < 0.003; Fig. 3-2). Day and Detling (1990) applied simulated ur ine to a mixture of little bluestem [ Schizachyrium scoparium (Michx.) Nash] and Kentucky bluegrass ( Poa pratensis L.). Urinetreated areas outyielded the nourine control 3870 to 2720 kg ha-1. Norman and Green (1958) applied urine to a mixture of cool-season grasse s, legumes, and forbs in spring. One month after application, herbage DM harvested was 1870 kg ha-1 for treated plots compared to 1160 kg ha-1 for the untreated control and 3 mo after application the difference was 2050 vs. 1160 kg ha-1, respectively. Thus, urine applicati on is associated with yield increas es that can carry over for at least several months. One reason for lack of response to dung applic ation was the interference effect on herbage under the pat. Reduction in DM harvested under th e dung pat (circle of 15 -cm radius) compared with the ring extending 15 cm beyond the e dge of the dung pat was greater than 500 kg ha-1 averaged across application freque ncies (Chapter 4). Hi rata et al. (1991) de scribed interference of dung pats on herbage beneath them or even death of herbage under th e pat, resulting in reduced yields (Hirata et al., 1991). MacDiarmid and Watkin (1971) report ed that 75% of grass tillers and rooted nodes of clover ( Trifolium spp.) stolons under the dung pat were dead within 15 d of application. This resulted in a significant redu ction in yield from the area of the pat. Another reason for the lack of a positive yield response in the dung-treated plots is the high proportion of organic N in dung relative to the hi gh proportion of total urine N that is urea N (Table 3-3 and 34). Norman and Green (1958) suggested that chemical differences between dung and urine accounted for lesser response to dung. Additionally in the current study, there was relatively little dung beetle ( Scarabaeidae ) activity apparent as well as drier than normal weather, both of which

PAGE 65

65 can result in slower nut rient release (Holter, 1979; Dickenson and Craig, 1990; Yokoyama et al., 1991). Table 3-7. Effect of N fertiliza tion on bahiagrass herbage res ponses for treatments to which no excreta was applied. Nitrogen fertilizer applied (kg ha-1) Response variable 0 60 120 Polynomial contrast Standard Error DM harvested (kg ha-1) 1310 2450 3060 L**, Q 125 N concentration (g kg-1) 13.6 12.7 13.9 NS 0.46 P concentration (g kg-1) 3.87 3.35 3.54 NS 0.22 IVDOM (g kg-1) 559 567 572 L** 7.3 N harvested (kg ha-1) 17.6 31.1 42.4 L** 1.4 P harvested (kg ha-1) 5.0 9.6 11.1 L** 0.6 L = linear, Q = quadratic; **, P 0.01; *, NS, P > 0.10; Letter followed by no symbol, P 0.10 Dry matter (DM) harvested; in vitr o digestible organic matter (IVDOM) Table 3-8. Monthly rainfall totals for 2006 for th e research location and the 30-yr average for Island Grove, FL. Island Grove is located 10 km from the research location and is the nearest site for which 30-yr data exist. Month 2006 30-yr Average -------------------------------------------mm----------------------------------January 22 107 February 147 83 March 4 85 April 52 75 May 12 119 June 130 180 July 233 167 August 159 172 September 93 115 October 39 49 November 47 61 December 93 117 Annual 1031 1330

PAGE 66

66 0 1500 3000 4500 6000 0123 Application Frequency (Number yr-1)DM Harvested (kg ha-1) Urine Dung Figure 3-2. Excreta type X excr eta application freque ncy interaction (P = 0.063) for bahiagrass DM harvested during 2006. There was no e ffect of dung application frequency on DM harvested (P > 0.375), but there were linear (P < 0.001) and quadratic (P = 0.097) effects of urine application frequency on th e response. Standard error of a treatment mean was 103 kg ha-1. Dry matter harvested was greater for urinethan dung-treated plots for Application Frequencies 1 (P = 0.002), 2 (P = 0.003), and 3 (P < 0.001).

PAGE 67

67 Herbage Nitrogen Concentration Herbage N concentration was affected by management intensity ( P = 0.039), application frequency ( P < 0.001), excreta type ( P < 0.001), and application fr equency X excreta type interaction ( P < 0.001) (Table A-1). High management intensity plots had greater N concentration than Average intensity (14.7 and 14.2 g kg-1, respectively). For control plots, N concentration ranged only from 12.7 to 13.9 g kg-1 and was not affected by amount of N fertilizer applied (Table 3-7). Pensacola bahiagrass herbage N concentra tion increased from 9.9 to 13.8 g kg-1 when N fertilizer amount increased from 0 to 224 kg ha-1 (Beaty et al., 1975). In Louisiana, Pensacola bahiagrass N concentration increased from 16.8 to 23.0 g kg-1 as N fertilizer amount increased from 0 to 450 kg ha-1 (Twidwell et al., 1998), while in Georgia it increased from 10.6 to 16.8 g kg-1 as N fertilizer increased from 56 to 450 kg ha-1 (Burton et al., 1997). Thus, bahiagrass is responsive to N fertilizer, but the range in N applied was relatively small in the current study. There was excreta type X application freque ncy interaction for he rbage N concentration. Interaction occurred because there was no effect of frequency of dung application on N concentration ( P > 0.246), but there were linear ( P < 0.001) and cubic effects ( P < 0.001) for response to urine application fr equency (Fig. 3-3). The linear e ffect was associated with an increase from 13.3 to 16.2 g kg-1 as application frequency increased from zero to three. The cubic effect occurred because there was little cha nge in N concentrati on between zero and one application, a large increase be tween one and two applications, and little change between two and three applications. Ur ine-treated plots had greater herbag e N concentration than dung-treated plots for Application Frequencies 2 and 3 ( P < 0.001), but there was no excreta type effect for a single application ( P = 0.495; Fig. 3-3).

PAGE 68

68 Jaramillo and Detling (1992) reported that herbag e N concentration in urine-affected areas of western wheatgrass [ Pascopyrum smithii (Rydb. Lve)] was greater than in control patches. Herbage N concentr ation averaged 30 and 16 g kg-1 for affected and unaffected areas. Ledgard et al. (1982) found similar effects of urine on N concentration in a perennial ryegrass ( Lolium perenne L.)-white clover ( Trifolium repens L.) mixture. Ryegrass herbage N concentration in urin e patches was 47 g kg-1 compared to 30 g kg-1 for unaffected areas. The range in N concentration in dung-t reated plots was only 13.2 to 13.6 g kg-1 (P > 0.246). Lack of N concentration re sponse to dung application fre quency can be attributed to reasons similar to those for absence of a DM harvested response to dung. Others have found some impact of dung on herbage N. Perennial ry egrass herbage from dung-affected areas had N concentration of 12.3 compared to 10.8 g kg-1 for unaffected herbage (Jorgensen and Jensen, 1997). Similarly, Greenhalgh and Reid (1968) stated that N concentration of perennial ryegrass herbage near a dung deposit was greater than simila r herbage growing in harvested fields (32 vs. 27 g kg-1, respectively).

PAGE 69

69 0 5 10 15 20 0123 Application Frequency (Number yr-1)N Concentration (g kg-1) Urine Dung Figure 3-3. Excreta type X excreta application frequency interaction ( P < 0.001) for bahiagrass herbage N concentration during the 2006 growing season. There was no effect of dung application frequency on N concentration ( P > 0.246), but there were linear ( P < 0.001) and cubic ( P < 0.001) effects of urine appli cation frequency on the response. Standard error of a treatment mean was 1.5 g kg-1. Herbage N concentration was greater for urinethan dung-treated plot s for Application Frequencies 2 and 3 ( P < 0.001), but there was no excreta type effect for a single application ( P = 0.495).

PAGE 70

70 Herbage Phosphorus Concentration Herbage P concentration was affected by application frequency ( P = 0.003) and the application frequency X excreta t ype interaction (Table A-1). Inte raction occurred because there was no effect of dung application fre quency on herbage P concentration ( P > 0.583), but there were linear ( P = 0.025) and quadratic ( P = 0.002) effects on herbage P concentration in urinetreated plots (Fig. 3-4). At A pplication Frequencies 1 and 3, P concentration in herbage from urine-treated plots ha d lesser and greater P concentrations ( P = 0.004 and 0.035), respectively, than herbage from dung-treated plots (Fig. 3-4). There was no effect of excreta type at Application Frequency 2 ( P = 0.752). Herbage P concentration in dung-treated plots remained in a narrow range between 3.33 and 3.45 g kg-1. Absence of bahiagrass DM harvested and herbage P concentration response to dung ap plication frequency i ndicate that almost none of the large amounts of P applied (up to nearly 600 kg P ha-1 for the three applications yr-1 treatment) were taken up by the plant. The strong quadratic effect of urine app lication frequency on tissue P concentration occurred because of a d ecline from 3.45 with no ur ine applied to 2.83 g P kg-1 with one urine application. The initial decline is likely relate d to greater herbage DM accumulation following one urine application with essentiall y no additional P a pplied (0.2-0.3 kg P ha-1; Table 3-6). Powell et al. (1998) reported that when urin e from rams was applied to pearl millet [ Pennisetum glaucum (L.) Br.], the urine not only increased pH, but also increased available P, especially during the first week after appli cation. In the current study, as application frequency increased from one to three yr-1, herbage P concentration in creased from 2.83 to 3.87 g P kg-1. This large increase in P concentration is not due to gr eater P application because those amounts changed only from 0.23 (Average) or 0.30 (High) kg P ha-1 for one urine application to 1.16 (Average) or 0.98 (High) kg P ha-1 for three applications (Table 3-6). T hus, the response may be due to greater

PAGE 71

71 plant vigor associated with grea ter N application as frequency of urine applic ation increased. This likely would lead to greater root growth and soil volume explored for P. We currently do not have root-mass data to support this argumen t, but Beaty et al. ( 1975) reported greater P concentrations in bahiagrass herbage associated w ith greater N rates and e xplained this response based on greater root mass and exploration of the soil. 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4 0123 Application Frequency (Number yr-1)P concentration (g kg-1) Dung Urine Figure 3-4. Excreta type X excreta application frequency interaction ( P = 0.002) for bahiagrass herbage P concentration during the 2006 gr owing season. There was no effect of dung application frequency on P concentration ( P > 0.583), but there were linear ( P = 0.025) and quadratic ( P = 0.002) effects of urine a pplication frequency on the response. Standard error of a treatment mean was 0.07 g kg-1. Herbage P concentration was affected by excreta type for Application Frequencies 1 and 3 ( P = 0.004 and 0.035, respectively), but there was no e ffect of excreta type at Frequency 2 ( P = 0.752).

PAGE 72

72 Herbage In Vitro Digest ible Organic Matter Herbage IVDOM was affected by management intensity (P = 0.001), application frequency ( P = 0.005), excreta type ( P = 0.001), and application frequency X excreta type interaction ( P = 0.099). Herbage in High management intensity pl ots was greater in IVDOM than in Average plots (577 vs. 567 g kg-1). In no-excreta plots, IVDOM increased linearly ( P = 0.007) from 559 to 572 g kg-1 as N rate increased from zero to 120 kg ha-1 (Table 3-7). Coleman et al. (2004) suggested that N fertilization has shown no consis tent effect on herbage digestibility, however a number of recent studies have shown greater IV DOM of bahiagrass herbage grown at greater N fertilization (Newman et al., 2006; Stewar t et al., 2007). Stewar t reported IVDOM of Pensacola bahiagrass increased from 459 to 479 to 505 g kg-1 as N rate increased from 40 to 120 to 360 kg ha-1. Similarly, Newman et al. (2006) reporte d increases in IVDOM from 443 to 487 g kg-1 when N fertilizer amount increased from 80 to 320 kg ha-1. One factor that works against increasing IVDOM with greater N fertilization is increasing st em development at greater N amounts (Coleman et al., 2004). Bahiagrass, ho wever, has limited stem elongation, thus its digestibility is less likely than most C4 grasses to be affected negatively by N fertilization. Excreta application frequency X excreta type interaction occurred because there was no effect of dung application frequency on bahiag rass IVDOM while urine application frequency did have an effect (Fig. 35). Herbage IVDOM changed li ttle following a single urine application, but subsequent applicati ons increased IVDOM from 565 to 588 g kg-1 (Fig. 3-5). With the exception of the lack of IVDOM respons e to the first applica tion of urine, these IVDOM responses to urine (added N) correspond to those already described in the current study for management intensity treatments and the no-exc reta control treatments. Excreta type affected IVDOM only for Application Frequency 3 when he rbage from urine-treated plots had greater IVDOM than from dung-treated plots ( P = 0.006).

PAGE 73

73 There are limited data describing digestibility of herbage in studies of excreta deposition. Greenhalgh and Reid (1968) stated in a general way that digestibility of perennial ryegrass herbage near a dung deposit may be greater than similar herbage growing in mechanically harvested fields. 545 550 555 560 565 570 575 580 585 590 595 0123 Application Frequency (Number yr-1)IVDOM (g kg-1) Urine Dung Figure 3-5. Excreta type X excreta application frequency interaction ( P = 0.099) for bahiagrass herbage in vitro digestible organic ma tter (IVDOM) concentr ation during the 2006 growing season. There was no effect of dung application frequency on IVDOM ( P > 0.300), but there was a linear ( P = 0.004) effect of urine application frequency on the response. Standard error of a treatment mean was 3.1 g kg-1. Herbage IVDOM concentration is different for Frequency 3 ( P = 0.006) but not different for Frequencies 2 ( P = 0.132) and 1 ( P = 0.445). Total Nitrogen Harvested Total N harvested was affected by management intensity ( P < 0.001), excreta application frequency ( P = 0.003), excreta type ( P < 0.001), and application fr equency X excreta type interaction ( P < 0.001) (Table A-1). High management in tensity plots had greater N harvested

PAGE 74

74 than Average (57 vs. 47 kg ha-1), and for no excreta plots the re sponse to N fertilizer amount was linear, increasing from 18 to 42 kg ha-1 as N amount increased from 0 to 120 kg ha-1 (Table 3-7). Frequency X type interaction occurred because N harvested increased linearly (P < 0.001) from 37 to 75 kg ha-1 as urine application frequency incr eased, but there was no effect of dung application frequency on the response (Fig. 3-6). Excreta type aff ected N harvested at Application Frequencies 1 through 3 ( P < 0.0001) (Fig. 3-6). Deenen and Middelkoop (1992) fertilized pe rennial ryegrass with either 250 or 400 kg N ha-1 yr-1 and applied a single application of dung at th ree different dates. As in the current study, N harvested in the grass was not si gnificantly different than in th e controls which did not receive dung. Ma et al. (2007) applied 6 kg of sheep dung uniformly to a mixed bunchgrass sward in Inner Mongolia. The area was harvested 32 and 65 d after dung applica tion. Nitrogen harvested 32 d after application was 48 and 35 kg N ha-1 for dung-treated and cont rol plots, respectively. Sixty-five days after dung applicati on, herbage N harvested was 32 and 25 kg ha-1 for these two treatments, respectively. Greater impact of dung in the study by Ma et al. (2007) than in the current study can be attributed to dung being uniformly spread over the plots and the general form of sheep dung compared to that of cattle. Th us, there are a range of N harvested responses to dung in the literature with amount of N fertilizer, amount of dung, and type of dung affecting the response. Cuttle and Bourne (1993) made si ngle urine applications (3.5 L m-2) to different perennial ryegrass plots at five dates between August and November. Nitrogen harvested ranged from 70 for early applications to 4 kg ha-1 for late applications. For untreated controls, N harvested was much less, ranging from a low of less than 1 kg ha-1 to a high of 13 kg ha-1. Similar results were also noted by Ball et al. (1979) using perennial ryegrass. They found that N harvested increased

PAGE 75

75 as N treatment increased from 0 N and no ur ine applied, to 300 or 600 kg fertilizer N ha-1 plus a single urine application. Herbage N harvested was 143, 188, and 202 kg N ha-1 for the three treatments, but absence a no-urin e control for each fertilizer le vel makes it impossible to draw conclusions about the impact of urine. 0 10 20 30 40 50 60 70 80 0123 Application Frequency (Number yr-1)N Harvested (kg ha-1) Urine Dung Figure 3-6. Excreta type X excreta application frequency interaction ( P < 0.001) for N harvested in bahiagrass herbage du ring the 2006 growing season. There was no effect of dung application frequency on N harvested ( P > 0.404), but there was a linear ( P < 0.001) effect of urine application frequency on th e response. Standard error of a treatment mean was 1.5 kg ha-1. Herbage N harvested was greater for urinethan dungtreated plots for Frequencies 1 through 3 ( P < 0.001).

PAGE 76

76 Total Phosphorus Harvested Total-season P harvested was affected by main effects of management intensity, application frequency, and excreta type ( P < 0.001) and the interactions of intensity X excreta type ( P = 0.033) and application fr equency X excreta type ( P < 0.001) (Table A-1). The management intensity X excreta type interact ion occurred because for dung-treated plots the magnitude of the advantage of High over Average management intensity (10.9 vs. 8.0 kg ha-1; P < 0.001) was greater than the advantage in urine-treated plots (13.7 vs. 12.4 kg ha-1; P = 0.054). Phosphorus harvested increased linearly from 5.0 to 11.1 as N fertilizer increased from 0 to 120 kg ha-1 for the no-excreta controls ( P = 0.002) (Table 3-7). Interaction of application frequency X excreta type occurred because there was no effect of dung application frequency on P harvested ( P > 0.570), but there were linear ( P < 0.001) and quadratic effects ( P = 0.052) of urine application fre quency on the response (Fig 3-7). Phosphorus harvested increased from 9 to 18 kg ha-1 as urine application frequency increased from zero to three. Herbage P harvested was greater for urin ethan dung-treated plots for Application Frequencies 2 ( P = 0.007) and 3 ( P < 0.001) and tended to be greater for Frequency 1 ( P = 0.115). Newman et al. (2005) harvested bahiagrass hay th at was treated with different levels of N fertilizer (0, 50, 67, and 100 kg N ha-1 harvest-1). Phosphorus harvested was 32, 39, 53, and 55 kg ha-1 for these four treatments, respectively. The response was primarily yield driven as DM harvested increased from 9.7 to 20.5 Mg ha-1, but herbage P concentration varied only from 3.7 (at the zero N rate) to 3.0 g kg-1 (at the 100 N rate). The values reported in the current study are much lower than those reported by Newman et al. (2005) due to lower N rates in some cases and likely greater N losses due to volatilization for othe rs, especially multiple applications of urine. There are few studies in the lite rature that measure tissue P re sponse or P removal under excreta

PAGE 77

77 application conditions similar to those in this experiment. There is a large body of literature describing P removal responses to uniform applicati ons of animal excreta or N fertilizer to hay fields, but these data have limited relevance to excreta applications like those in the current study. 6 8 10 12 14 16 18 20 0123 Application Frequency (Number yr-1)P Harvested (kg ha-1) Urine Dung Figure 3-7. Excreta type X excreta application frequency interaction ( P < 0.001) for P harvested in bahiagrass herbage du ring the 2006 growing season. There was no effect of dung application frequency on P harvested ( P > 0.570), but there were linear ( P < 0.001) and quadratic effects ( P = 0.052) of urine application frequency on the response. Standard error of a treatment mean was 0.35 kg ha-1. Herbage P harvested was greater for urinethan dung-treated plots for Application Frequencies 2 ( P = 0.007) and 3 ( P < 0.001) and tended to be greater for Frequency 1 ( P = 0.115).

PAGE 78

78 Excreta Nitrogen Recovery Total-season excreta N recovery was affect ed by the main effects of excreta type ( P < 0.001) and a linear effect of application frequency ( P = 0.0638) (Table A-1). There were no interactions. Across the range of excreta application freque ncies in this study, excreta N recovery from urine was consistently greater than from dung a nd N recovery decreased as number of excreta applications increased (Fig. 3-8). Over th e 2006 season, the breakdown of dung pats was less than expected, possibly associated with minimal dung beetle activity leading to reduced recovery of N. Perennial ryegrass fertili zed with either 250 or 400 kg N ha-1 yr-1 and receiving a single application of dung, had N recovery over a 3-mo period of 1.9 and 0.9% for the two N rates, respectively (Deenen and Middelkoop, 1992). These re sults are similar to those of the current study. In a study in which a single ur ine application was made to perennial ryegrass-white clover mixtures fertilized with 300 and 600 kg ha-1, Ball and Keeny (1981) found that total N recovery (not only urine N) was 37 and 23% for the fe rtilization treatments of 300 and 600 kg N ha-1. The N application rates were likely in excess of plan t requirements, increasing the risk for loss. Greatly exceeding plant nutrient needs may explai n in part the lesser N recovery with multiple urine applications in the current study. Cuttle an d Bourne (1993) determined urine-N recovery in perennial ryegrass for a single ur ine application at five date s from August through November. The cumulative recovery in herbag e ranged from 40% of N from the first application to 1% of N from the next to last application. They noted th at the seasonal pattern of herbage production was the dominant factor determining N harvested imme diately following the urine treatments. In the current study, greater recovery for lesser application frequencies was li kely due in part to the fact

PAGE 79

79 that these applications occurred earlier in the growing season, allowing more time for nutrient uptake by the bahiagrass. Another reason for low capture of N could be growth patterns of ba hiagrass. Blue (1973) noted that fertilizer N recovery by bahiagrass wa s low early in stand life due to large amounts of N stored in the rhizome-root system. Although our plots were long-time established stands, visual observations suggest that low soil fertility or some other factor was limiting overall stand vigor at the start of the experiment. Thus our plots may have responded somewhat like establishing stands in terms of building up the rhizome-root system and capturing significant amounts of N in storage organs. Core samples will be taken following completion of the second year of the study to assess the impact of ex creta and control treatme nts on storage organ mass and N concentration.

PAGE 80

80 0 5 10 15 20 25 30 123Application Frequency (Number yr-1)Nitrogen Recovery (%) Urine Dung Figure 3-8. Excreta type main effect ( P < 0.001) for excreta N recove ry in harvested bahiagrass herbage during the 2006 growing season. Th ere was no excreta type X excreta application frequency interaction ( P = 0.543), but there was a linear effect ( P = 0.064) of excreta application frequency on N recover y. Standard error of a treatment mean was 1.8%.

PAGE 81

81 Nutrient Concentration in Shallow Groundwater Much drier than average weather preceded the start of the stu dy in spring and early summer 2006, and this in combination with less than average rainfall in all remaining months of the year except July (Table 3-8) limited the occasions when soil water was sufficient for lysimeter sampling to occur. There were however, at least two sampling dates mo-1 except for November 2006 and March, April, and May 2007. The only significant effects of treatments we re observed on 15 Sept. 2006 when there was excreta type X applicatio n number interaction ( P = 0.055) for NO3-N concentration (Fig. 3-9). Interaction occurred b ecause there was a linear effect of application frequency for dung-treated plots ( P = 0.039), but the effect in urin e-treated plots was quadratic ( P = 0.057; Fig. 3-9). In dung-treated plots, the increase in water NO3-N concentration was only 0.07 to 0.15 mg L-1 as frequency increased from 0 to 3, and in urin e-treated plots concentration increased from 0.02 (zero applications) to 0.10 (two applications) before decreasing ag ain to 0.02 (three applications; Fig. 3-9). There were isolat ed elevated numbers for NO3-N concentration, incl uding values from two lysimeters that exceeded 10 mg L-1 on 22 Aug. 2006 (High management intensity, urine, 1 and 2 applications yr-1) and one value of 19 mg L-1 on 29 Aug. 2006 (High management intensity, urine, 2 applications yr-1). On 22 Aug. 2006, there was one other sample greater than 5 mg L-1, and on 29 Aug. 2006 there were two ot her samples greater than 5 mg L-1. At other sampling dates, maximum values for NO3-N concentration never exceeded 5 mg L-1 and treatment means rarely exceeded 2 mg L-1. Wachendorf et al. (2005) applied urine (equivalent to 1030 kg N ha-1) and dung (equivalent to 1050 kg N ha-1) labeled with 15N to perennial ryegrass plots with free-drainage lysimeters. The urine-treated plots had the greatest loss of NO3, which occurred within 100 d of the application

PAGE 82

82 (120 kg N ha-1), compared to a no-excreta control and dung which had losses of approximately 0 and 10 kg N ha-1. The majority of NO3 under urine patches was leache d within a 60-d period. Low levels of NO3-N concentration in the current study may be due in part to lower than normal rainfall because July was the only mont h during 2006 when rainfall exceeded the 30-yr mean. It is likely that there were significant losses of urine-N due to volatilization because of excreta application during summe r (Russelle, 1996). In addition, bahiagrass has been shown to be a relatively efficient scrubber of N below gr ound, resulting in accumulation of N in rhizomes and roots (Blue, 1973). The plots in the current study will be sampled at the end of Year 2 to quantify differences in rhizome-root mass and N content. So, despite high application rates and relatively poor recoveries of excreta N in harvested herbage, the lack of significant NO3-N concentrations in water is likely due to lower th an average rainfall, volat ilization losses, and N capture and subsequent storage in rhizomes and r oots. Hack-ten Broeck et al. (1996) noted that dung patches rarely affect nitrate leaching, because of high proportions of N excreted in the urine and the preponderance of organic N in dung that slowly degrades.

PAGE 83

83 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0123 Application Frequency (Number yr-1)NO3-N Concentration (mg L-1) Dung Urine Figure 3-9. Excreta type X applicat ion frequency interaction for NO3-N concentration in shallow soil water on 15 Sept. 2006. There was a qua dratic effect of urine application frequency on NO3-N concentration ( P = 0.057), and there was a linear effect ( P = 0.0394) of dung application frequency on the re sponse. There was no effect of excreta type at any level of application frequency ( P > 0.181). Standard e rror of a treatment mean was 0.03 mg L-1.

PAGE 84

84 Summary and Conclusions The objective of this research was to charac terize the effects of cattle excreta type and application frequency on 1) he rbage yield, chemical compositi on, and nutrient recovery and 2) nitrate leaching to shallow ground water under bahiagrass swards managed at two different intensities. The study was car ried out during the period fr om June 2006 through June 2007. The High and Average management intensity tr eatments imposed on the pastures grazed by livestock (excreta source pastures) had limite d impact on herbage characteristics of grazed bahiagrass. Herbage N concentration was greater for High than Average, but herbage mass, P concentration, and IVDOM were not affected. There were sampling date effects on pasture herbage mass with greatest mass o ccurring in July compared to June and September, and herbage CP tended to be greater in July than other dates. Dung nutrient c oncentration was quite consistent between treatments and across the three collect ion periods. Urine nutrient concentration was similar between treatments in June (not analyzed statistically), but it was quite high for the Average treatment relative to the High treatmen t in July and for High relative to Average in September. Herbage characteristics did not vary s easonally to the same degree as urine chemical composition, so the seasonal changes in urin e composition, especially the high nutrient concentrations in July, are not understood. In the plots to which excreta was applied, ther e were interactions of excreta application frequency X excreta type for all herbage respons es except for excreta N recovery. Interaction occurred because dung applicati on had no effect on herbage res ponses, whereas responses to urine were consistently signifi cant. Greater herbage response to urine than dung was expected because of the high proportion of dung N that is in an organic form and the greater availability to plants of nutrients in urine. The general abse nce of response to dung was not expected and could be attributed to a number of factors includi ng physical interference of the dung pat, the high

PAGE 85

85 concentration of organic N in dung as a proportion of total N, limited a pparent activity of dung beetles, and a drier than normal year l eading to rapid drying and crusting of dung. High management intensity applied to the ungrazed bahiagrass plot s increased herbage response over Average for all response variable s except P concentration and N recovery. With the exception of excreta type X management intensity interacti on for P harvested, there were no interactions of other treatments with management intensity. This i ndicates that both excreta types and the three excreta ap plication frequencies evaluated in this study had similar effects on response variables across the range of management intensities tested. Excreta N recovery was greater for urine than dung averaging 22 and 2%, respectively. Recovery was also affected by excreta applica tion frequency, decreasing from 28 to 18% as urine application frequency increased from one to three yr-1, and from 4 to less than 1% as dung application frequency increased from one to three. These recoveries are in the lower part of the range reported in the literature, but it does not ap pear that leaching losses explain this response. Significant responses of shallow soil water NO3-N concentrations occurred at only one sampling date (15 Sept. 2006), and these values were less than 0.15 mg L-1. Greater concentrations occurred in individual wells at two other dates in A ugust, but these were not consistent across replicates within a trea tment. Warm, dry weather may have increased volatilization losses of urine N (Russelle, 1996), and bahiagrass has been reported to store larg e quantities of N in rhizomes and roots (Blue, 1973). Low values for dung recovery and absence of herbage response to dung suggest limited mineralization of nutrien ts in dung during the c ourse of the growing season. Greater percent recovery of nutrients in excreta occurred w ith single excreta applications suggesting that grazing management practices which increase uniformity of excreta deposition will likely increase efficiency of nut rient cycling in grazed grasslands.

PAGE 86

86 CHAPTER 4 SPATIAL PATTERNS OF BAHIAGRASS HERBAGE ACCUMULA TION AND NUTRIENT CONCENTRATION RESPONSES TO TYPE AND FREQUENCY OF EXCRETA DEPOSITION Introduction Planted grasslands and non-fo rested rangeland comprise near ly 30% of the USA land area (Barnes and Nelson, 2003) and occupy more than 4 million ha in Florida (Dubeux et al., 2007). Most grasslands in Florida are managed extensively with limite d fertilizer input, but because there is such a large area covere d by grasslands, the fate of nutrients can have a major impact on ecosystem function (Nair et al., 2007). When livestock graze grassland, a large proportion of the nutrients consumed in forage are returned to the sward in excreta (Sollenberger et al., 2002). A single urina tion from mature cattle may provide the equivalent of 5 mm of rain and 400 to 500 kg N ha-1, while dung may supply the equivalent of 110 kg P and 220 kg of K ha-1 along with other nutrien ts (Haynes and Williams, 1993). Nutrients recycling through animal excreta ha ve long been considered beneficial to the fertility of grazed pastures (Bal l et al., 1979). For example, urine patches in a mixed grassland contained 112 g m-2 more above-ground biomass and 2.5 g m-2 more plant N than unaffected areas (Day and Detling, 1990). In a Colorado st udy, urine patches affected only 2% of the pasture surface, but they contributed 7 to 14% of consumed forage (Day and Detling, 1990). The areas covered by a single dung or urine application by cattle have been estimated at ~ 0.1 and 0.4-m2, respectively (Haynes and Williams, 1993). To effectively define the impact of dung and urine application on grasslands, the spatia l pattern of plant responses around an excreta deposit needs to be described. Lotero et al. (1966) observed cattle grazing tall fescue ( Festuca arundinacea Schreb.) and reported that urine affect ed plant response in an area of ~ 1.02 m2. With dairy cows, Lantinga et al. (1987) reported that urine affected plant growth in an area of

PAGE 87

87 0.68 m2. Based on estimates from Haynes and William s (1993) and de Klein (2001), an affected area for urine is ~ 0.75 m2. In a New Zealand pasture of ryegrasses ( Lolium spp.) and white clover ( Trifolium repens L.), dung pats killed 75% of grass tillers and ro oted nodes of clover stolons under the pat within 15 d of application, resulting in significant yield reduction (MacDiar mid and Watkin, 1971). They noted that yield increased in response to dung from the edge of the pat to 45 cm beyond the edge, but in a second study the increase was lim ited to 15 cm. In related research, Deenen and Middelkoop (1992) applied dung pats to circles of radius 15 cm in perennial ryegrass ( Lolium perenne L.). They found that dung aff ected plant responses in an area that extended 15 cm from the edge of the dung patch. More detailed description of the spatial pa tterns in herbage accumulation and chemical composition around dung and urine deposits woul d aid assessments of excreta impact on grasslands. Measuring the effect of a range of a pplication frequencies would also be valuable as some areas of grazed grassland receive no ex creta while other areas may receive multiple deposits in a given year (Mathews et al., 2004). The objective of th e research reported in Chapter 4 is to characterize the spa tial patterns of bahiagrass ( Paspalum notatum Flgge) herbage accumulation, chemical composition, and nutrient removal following application of dung and urine at a range of application frequencies. Materials and Methods There were two sites used for this study. At one location, pastures were grazed by yearling beef heifers and these pastures served as the source for excreta. The other site was used for excreta application to ungrazed bahiagrass plots. Excreta source pastures were located at the University of Florida Beef Research Un it, northeast of Gainesville, FL (29.72o N latitude, 82.35o W longitude). A long-term, ungrazed stand of Pensacola bahiagrass at the Plant Science

PAGE 88

88 Research Unit, Citra, FL (29.41o N latitude, 82.02o W longitude)was used for applications of excreta. Site characteristics a nd the rationale for choosing thes e locations were described in Chapter 3. Treatments and Design Treatments applied to ungrazed bahiagrass pl ots were two management intensities (the same N amounts as were applied to excreta sour ce pastures, i.e., 60 [Ave rage] and 120 [High] kg ha-1 yr-1), two types of excreta (dung and urine), a nd three frequencies of excreta application (one, two, or three times per year). The 2 x 2 x 3 factorial accounted for 12 treatments. In addition, there were three control treatments that received no excreta and we re fertilized with N at 0, 60 (no excreta control for Average management intensity plots), or 120 kg ha-1 yr-1 (no excreta control for High management intensity pl ots). The 15 treatments were replicated three times in a randomized complete block design. Plot s were 3 x 3 m in area with a 1-m bahiagrass alley surrounding each plot. Average and High management intensity so urce pastures were defined based on N fertilizer amount and stocking rate. Averag e management pastures received 60 kg N ha-1 yr-1 and were stocked with two yearling heifers ha-1, and High pastures received 120 kg N ha-1 yr-1 and were stocked with four yearling heifers ha-1. These two management intensities were selected to represent fertilization and stoc king regimes that are common in the Florida livestock industry (Average) or represent approximately the most intensive management applied to grazed bahiagrass (High). In addition, based on previous wo rk by Stewart et al. (2007) it is expected that these treatments will result in fo rage that varies in nutritive va lue, especially N concentration, and that these differences could aff ect the composition of dung and urine.

PAGE 89

89 Excreta Source Pastures There were two replicates of each source treatment (Average and High management intensities) arranged in a comp letely randomized design. Pastures were stocked continuously and pasture size was 1 ha for Av erage and 0.5 ha for High. Each pasture was grazed with two crossbred yearling heifers (Angus x Brahman) with average initial weight of 408 kg. Grazing was initiated on 26 May 2006. Details of animal management, pasture fertilization, and pasture sampling were provided in Chapter 3. Excreta Collection and Analyses Animal selection and training we re described in Chapter 3. During the trial, animals were brought to stanchions for collection of urin e and dung. Urine and dung were collected during three periods (12-16 June; 26-28 July; and 7-8 Sept 2006) leading up to the three dates (20 June, 1 Aug., and 13 Sept. 2006) when excreta was applied to the plots. Excreta application dates were at 6-wk intervals. The process used for excreta collection is described in Chapter 3. After urine and dung were collected separately, they were immediatel y put into sealed containers and taken to a refrigerator for storage at 4C. Prior to application, a ll excreta events of a given type (dung or urine) from one intensity treatment (four animal s) were composited across field replicates. Thus, at time of application there we re two types of dung and two types of urine (one of each excreta type from both Average and High treatments) Three subsamples were taken from the composited samples of urine and dung and analyz ed to determine excreta chemical composition so that amount of nutrients appl ied to plots could be calculated. During the June excreta collection, samples were taken to ensure that chemical composition did not change markedly during the 8 d of excreta storage prio r to application, i.e., to ascertain that what was appl ied to the plots was similar in composition to what the animal

PAGE 90

90 would deposit fresh. This was accomplished by taking two subsamples from each of two separate urine and dung events. One of the two subsamples from each event was analyzed immediately; the other subsamples were stored in the same manner as the urin e and dung that were eventually applied to the plots. At the date of excreta application to plots, these stored subsamples were analyzed to assess changes in chemical compos ition during the storage pe riod. Results of these analyses were reported previously (Table 3-1). Urine samples for analysis were acidified with concentrated sulfuric acid to pH 2 to avoid N volatilization. Dung was analyzed for total N, organic N, NH4-N, P, and K, and urine was analyzed for N, NH4-N, urea-N, P, and K. Methods of urine and dung analyses, nutrient concentrations of both (Tables 33 and 3-4), and quantity of N, P, and K applied to each plot (Tables 3-5 and 3-6) were reported in Chapter 3. Plot Management All plots to which excreta would subsequently be applied were fertilized on 7 June with 18 kg P and 66 kg K ha-1. Nitrogen application depended upon ex creta-source treatment. Plots that received excreta from the High management intens ity source pastures were fertilized with 120 kg N ha-1 yr-1, split equally in two applications of 60 kg N ha-1 (7 June and 16 Aug. 2006). Plots that received excreta from the Average management intensity plots were fertilized with 60 kg N ha-1 yr-1, in two applications of 30 kg N ha-1 (on the same dates as High). Following fertilization on 7 June, the entire plot area was irrigated with 25 mm of water because of spring drought and lack of early summer rainfall (rainfall in March thro ugh May at this location was 68 mm compared to a 30-yr average of 279 mm). Magnesium sulfate was applied to all plots on 6 July 2006 to address low soil Mg levels. It was applied at 135 kg ha-1 to provide 27 kg Mg ha-1 and 36 kg S ha-1.

PAGE 91

91 Excreta Application Plots were staged to a 5-cm stubble height on 19 June 2006. All experimental units except the no-excreta controls received dung or urine on 20 June. Subsequent applications were on 1 August and 13 September. Only plots receiv ing two or three excreta applications yr-1 received excreta on 1 August, while on 13 September only those plots receiving three applications yr-1 were treated. Quantity of dung and urine and the area to which they were applied were determined based on values reported for cattle in the literature (Haynes and Williams, 1993) and the authors personal observation of the cattle used in this study. Haynes and Williams (1993) indicated that surface area ranges from 0.16 to 0.49 m2 for urine and 0.05 to 0.09 m2 for dung. Urine volume is said to range between 1.6 and 2.2 L and fresh dung mass from 1.5 to 2.7 kg (Haynes and Williams, 1993). In the current study, 2 L of urine constituted one application, and it was applied to an area of 0.283 m2, a 30-cm radius circle, with the center being the center of the plot (Fig. 4-1). Of the 2 L of urine, 1 L was applied to a 15-cm radius circ le with the center being the center of the plot (area of 0.071 m2), and 1 L was applied to the area outside the 15-cm radius circle but inside a circle of 30-cm radius. This was done to reflect the likelihood of greater concentration of urine closer to the center of the affect ed area in a natural urine deposit. The urine was applied using a sprinkler head on a watering can, an d rate of application was cont rolled so that runoff from the application area was minimal. Dung was applie d at 2 kg fresh weight to an area of 0.071 m2, based on the range in surface area for dung applications (0.05 to 0.09 m2) in the review by Haynes and Williams (1993). This was a circle of 15-cm radius with the midpoint being the center of the plot (Fig. 4-1). The dung was even ly distributed across th is area. Subsequent

PAGE 92

92 applications (for the two and three applicati ons per year treatments) occurred at the same location. Forage Harvest and Laboratory Analyses Forage was harvested every 28 d following in itial excreta application (20 June), and harvest dates were 18 July, 15 Aug., 12 Sept., and 10 Oct. 2006. To characterize the spatial response of herbage DM harveste d and nutritive value, herbage wa s harvested beginning with a circle of radius 15 cm that was centered on the midpoint of the excreta application. Thereafter, concentric rings were harvested sequentially. Ri ngs are defined based on their radius from the center of the excreta application (Fig. 4-1). Rings 1 through 3 were harvested for dung-trea ted plots and Rings 1 through 4 for urinetreated plots. For the urine-treated plots, R4 wa s harvested because of th e greater area of urine vs. dung applications and because urine typically a ffects an area ~2.3 times the area of the urine patch (de Klein, 2001). Throughout this chapter, the rings will be referred to as R1 through R4, as defined above. Following harvest, the herbage from each ring was dried separately at 60oC, weighed, and ground to pass 1-mm screen. Herbage analyses for N and P were conducted at the Forage Evaluation Support Laboratory using a micro-Kj eldahl technique followed by semi-automated colorimetric analysis of the digestate (Gallaher et al., 1975; Hambleton, 1977). Nitrogen and P harvested in each ring were calculated by multiplying DM harvested times nutrient concentration. In vitro digestible organic matte r concentration (IVDOM) was determined using a two-stage procedure (Moore and Mott, 1974).

PAGE 93

93 Figure 4-1. Diagram of harvested rings to quantif y spatial pattern of re sponse to dung and urine application (R1, circle of 15-cm radius; R 2, 15to 30-cm radius; R3, 30to 45-cm radius; R4, 45to 60-cm radius). Statistical Analysis Herbage DM harvested, herbage N, P, and IVDOM concentrations, and herbage N and P harvested were analyzed using analysis of va riance in PROC MIXED of SAS (SAS Institute, Inc., 2007). Herbage DM, N, and P harvested ar e the sums of four harvests, and nutrient concentrations and digestibility are weighted averages across the four harvests. Data were analyzed by excreta type because of the greater number of rings harvested for urinethan dungtreated plots. The no excreta, 60 kg N ha-1 treatment served as a zero excreta application frequency for the Average management intens ity treatments (both dung and urine), and the no excreta, 120 kg N ha-1 treatment served as a zero excret a application frequency for the High management intensity treatments (both dung and urine). Fixed effects in the models were management intensity, excreta application fr equency, ring number, and their interactions. R1 R2 R3 R4

PAGE 94

94 Replicate was a random effect. Polynomial contra sts were used to assess the response to ring number and excreta applicati on frequency. Differences were considered significant when P 0.10. Data presented are least squares means. In this chapter, only ring main effects and in teractions with ring will be reported. This is done for two reasons. First, our objective in this ch apter is to assess spatia l patterns of response, and differences among rings are how this is charac terized. Also, because of the concentric circle sampling approach used, the amount of land area sampled was different for each ring (i.e., R1 through R4). Thus, when SAS calculates a main eff ect mean for the other treatment factors (i.e., excreta application frequency and management intensity) and the frequency x intensity interaction means, the value is a mean across levels of ring numb er. When calculated in this manner, these means are not weighted to account fo r the different areas of the rings. In Chapter 3, all data presented were from the entire 0.636-m2 circular area around the excreta deposit encompassed by Rings 1 through 3, and they were calculated by weighting the responses for the specific areas of each of those three rings. T hus, the best assessment of the effects of management intensity and excreta application fre quency main effects and the interaction of these two factors are the data presented in Chapter 3. Results and Discussion Dry Matter Harvested There was a main effect ( P < 0.001) of ring number on DM harvested for dung-treated plots, but there also was applicati on frequency X ring number interaction ( P = 0.027) (Table A2). Likewise for urine-treated pl ots, there was a main effect ( P < 0.001) of ring number and an excreta application frequency X ring number interaction ( P < 0.001) (Table A-3). Application frequency X ring number interacti on occurred for urinetreated plots because DM harvested decreased linearly ( P < 0.001) as ring number increased for Application

PAGE 95

95 Frequencies 2 and 3 (Fig. 4-2). The response also decreased with increa sing ring number for one urine application [linear ( P < 0.001) and quadratic ( P = 0.058)], but there was no effect ( P = 0.352) of ring number for control plots. Lotero et al. (1966) quan tified tall fescue DM harvested in concentric circles around natu rally occurring urine deposits. Inner and outer radii of areas sampled were 0 to 15 cm, 15 to 25 cm, 25 to 35 cm, 35 to 45 cm, 45 to 55 cm, and 55 to 65 cm. Similar to our results, they reporte d that the effect of urine on forage growth is most pronounced at the center of deposition and decreases with increasing distance from that point. A question of interest is how far beyond the area of the urine a pplication is herbage production affected. The DM harvested in R4 for all applic ation frequency treatments was ~ 3000 kg ha-1, but for R1 the DM harvested ranged from ~ 3000 kg ha-1 for the zero urine application frequency to 6000 kg ha-1 for an application frequency of three (Fig. 4-2). This pattern of response is reflected in polynomial contrasts whic h showed that DM harvested increased with increasing excreta app lication frequency for R1 (linear [ P = 0.001] and quadratic [ P = 0.0192]), R2 (linear [ P = 0.056] and quadratic [ P < 0.001]), and R3 (linear [ P = 0.002]), but for R4 there was no effect of application frequency ( P > 0.153). Lotero et al. (1966) observed cattle grazing ta ll fescue, and immediately following a urine event a cage was placed around the deposit. Subseque nt harvests of forage around the urine spot indicated that urine affected pl ant growth in an area of ~ 1.02 m2. In our study, the absence of urine application frequency effect for R4 s uggests that urine impact on DM harvested was limited to a circle of radius 45 cm from the center of application, an area of 0.64 m2. One reason why the area affected may be smaller in the curre nt study than in Lotero et al. (1966) is because they sampled spots where urine was deposited na turally by cows. In the current study, urine was applied using watering cans with sprinkler he ads to ensure uniform application. Visual

PAGE 96

96 observations in the field suggest that rate of flow of natural urine deposition is more rapid and may be associated with greater lateral movement of urine across the soil surface before it soaks in. Despite this difference, data from the cu rrent study are comparable with several other experiments. With dairy cows, Lantinga et al. (1987) re ported that urine affected growth in a 0.68-m2 area. Haynes and Williams (1993) indicated that the area covered by a urine event ranges from 0.16 to 0.49 m2, and de Klein (2001) reported that the area affected is 2.3 times the area to which urine is applied. Using the mid-point of the range proposed by Haynes and Williams (1993) and the factor suggested by de Kl ein (2001) results in a calcu lated affected area of 0.75 m2, comparable to that measured in the current study. When artificial urine wa s applied to perennial ryegrass that was fertilized with 250 kg N ha-1 yr-1, effects were measured up to only 15 cm from the edge of the urine patch (Deenen and Midde lkoop, 1992). This corresponds to the results of the current study because urine was applied to a ci rcle of 30-cm radius and herbage accumulation was affected only through a 45-cm radius. As in urine-treated plots, there was app lication frequency X ring number interaction in dung-treated plots, but the response was very different than for urine. The primary factor driving the interaction was the negative impact of dung app lication, particularly multiple applications, on DM harvested in Ring 1 (Fig. 4-3). This resu lted in quadratic effects of ring number on DM harvested for three ( P = 0.025) dung applications, linear ( P < 0.001) and quadratic ( P = 0.008) effects for two applications, and no effect for one ( P > 0.370) or for zero applications ( P > 0.100) (Fig. 4-3). Reduction in DM harvested in Ri ng 1, where the dung was applied, ranged from 500 to more than 1000 kg ha-1 for plots to which multiple applications of dung were made.

PAGE 97

97 Focusing on individual rings, there was an e ffect of number of dung applications on DM harvested only in Ring 1. This response had significant linear ( P = 0.047) and cubic ( P = 0.041) terms. The linear effect reflects greater depr ession in DM harvested with increasing dung application frequency. The cubic effect is due to the greater negative im pact of two vs. three dung applications, a response for which there is no apparent explanation. In Rings 2 ( P > 0.401) and 3 ( P > 0.137), there was no effect of dung applic ation frequency on DM harvested. This indicates that effect of dung on th is response was limited to the circle of 15-cm radius that was the initial area of application. In a New Zealand pasture that was primarily ryegrass and white clover, dung pats of 1.8 kg fresh weight were applied to a circle of 15cm radius (MacDiarmid and Watkin, 1971). They reported that 75% of grass til lers and rooted nodes of clove r stolons under the dung pat were dead within 15 d of application. This resulted in a si gnificant reduction in yi eld from the area of the pat as was observed in the current study. They noted that in one expe riment, yield increased in response to dung to a radius of 61 cm from th e center of application, bu t in a second study the increase was limited to a radius of 30 cm. In related research, Deenen and Middelkoop (1992) applied a single dung pat (2.5 kg fres h weight) to a circle with ra dius of 15 cm in perennial ryegrass plots. They measured DM harvested in five concentric rings or bands around the dung pat. The radii of the areas harvested were 0 to 15, 15 to 30, 30 to 45, 45 to 60, and 60 to 75 cm. They found that the dung-affected area was confined to 15 cm fr om the edge of the dung patch. Thus, results of the current study differ somewhat from previous work. In this experiment, there were no measurable effects of dung on DM harves ted outside the area of dung application, while several studies in the literature report eff ects extending at least 15 cm beyond the dung pat.

PAGE 98

98 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 1234 Ring Dry Matter Harvested (kg ha-1) Urine-3 Urine-2 Urine-1 Urine-0 Figure 4-2. Ring number by urine application fr equency interaction ( P < 0.001) on herbage DM harvested during 2006. Ring number effects were linear for three ( P < 0.001) and two applications ( P < 0.001), linear ( P < 0.001) and quadratic ( P = 0.058) for one application, and not significant ( P > 0.352) for zero applications. Standard error of a treatment mean was 76 kg ha-1. Ring Numbers 1 through 4 re fer to sampling areas 0 to 15, 15 to 30, 30 to 45, and 45 to 60 cm fr om the center of th e urine application, respectively.

PAGE 99

99 1500 1700 1900 2100 2300 2500 2700 2900 3100 3300 123 RingDry Matter Harvested (kg ha-1) Dung-0 Dung-1 Dung-3 Dung-2 Figure 4-3. Ring number by dung app lication frequency interaction ( P < 0.027) on herbage DM harvested during 2006. Ring number e ffects were quadratic for three ( P = 0.025) applications, linear ( P < 0.001) and quadratic ( P = 0.008) for two applications, and not significant for one ( P > 0.370) or for zero applications ( P > 0.100). Standard error of a treatment mean was 106 kg ha-1. Ring Numbers 1 through 3 refer to sampling areas 0 to 15, 15 to 30, and 30 to 45 cm from the center of the dung application, respectively.

PAGE 100

100 Nitrogen Concentration For dung-treated plots, there was no effect of ring ( P > 0.160) on herbage N concentration nor were there interactions of other treatments with ring ( P > 0.686) (Table A-2). Means for Rings 1 through 3 were 13.5, 13.3, and 13.6 g kg-1, respectively. There has been limited research investigating the spatia l impact of dung on herbage N con centration. Jorgen sen and Jensen (1997) applied 2.1 kg of sheep feces, the equivalent of 960 kg N ha-1, in July to a 25-cm diameter circle in a mixture of perenni al ryegrass cv. Sisu and white clover cv. Milkanova. The herbage was harvested in concentric circles around the app lication area such that the first extended 0 to 15 cm beyond the edge of that area and the se cond extended 15 to 30 cm beyond the edge. In the 0to 15-cm zone, neither grass nor clover herbag e N concentrations were affected by feces in October or the following June. By August, 13 mo after feces application, grass N concentration was greater in the 0 to 15 cm zone than for control plots (25.2 vs. 20.5 g kg-1). These authors noted that less than 2 g kg-1 of dung N was recovered in harv ested herbage by October following July application, and after 13 mo only 35 g kg-1 of dung N was recovered in harvested herbage. The application frequency x ring number interaction ( P < 0.001) on urine-treated plots occurred because there were linear effects of ring for two and thr ee urine applications, but there was no ring effect for the control (means of 13.3, 13.1, 13.5, and 13.3 g kg-1 for R1 through R4, respectively) or for one urine application (14.1, 13.6, 13.6, and 13.6 g kg-1 for R1 through R4, respectively) (Fig. 4-4). Moving from the cente r of the urine deposit outward, N concentration decreased linearly ( P < 0.001) for two (17.8, 16.8, 15.3, and 14.4 g kg-1) and for three urine applications (17.8, 16.4, 15.6, and 14.0 g kg-1, respectively). Lotero et al. (1966) evaluated urine impact s on tall fescue pastures. They measured herbage N concentration at in creasing distances from a urin e deposit. They found a marked decrease in N concentration moving away from the point of impact, but it was apparent only at

PAGE 101

101 the first cutting following applic ations of urine in spring, summe r, and autumn. In the current study, the lack of significance for th e single application treatment likel y is due to the fact that it occurred at the beginning of the experimental period and because the herbage N concentrations reported are weighted averages across four harv ests that occurred over a 112-d period following that application. Application Frequencies 2 and 3 received urine at Day 42 and Days 42 and 84 in this period, respectively, and thei r effect was measurable acro ss the time period of the study. The effect of application frequency within each ring number was explored to determine how far from the center of urin e application an effect on herb age N concentration could be detected. This response had strong linear ( P < 0.001) and cubic ( P < 0.001) effects of application frequency for Rings 1 through 3. With increa sing urine application frequency, herbage N concentration increased, but the effect was mi nimal between zero and one application and likewise between two and three a pplications (Fig. 4-4). For Ring 4, the data began to converge markedly, but the linear effect (P = 0.029) of application frequency remained. This indicates that the effect of urine application on herbage N concentration extended 30 cm beyond the edge of application and 15 cm further than th e effect on DM harvested (Fig. 4-2).

PAGE 102

102 10 11 12 13 14 15 16 17 18 19 1234 RingN Concentration (g kg-1) Urine-3 Urine-2 Urine-1 Urine-0 Figure 4-4. Ring number by urine application frequency interac tion (P < 0.001) on total-season herbage N concentration during 2006. Ring numbe r effects were linear for three (P < 0.001) and for two (P < 0.001) applications, but not significant for one (P > 0.171) and zero (P > 0.128) urine applications. Sta ndard error of a treatment mean was 0.25 g kg-1. Ring Numbers 1 through 4 refer to sampling areas 0 to 15, 15 to 30, 30 to 45, and 45 to 60 cm from the center of the urine applic ation, respectively.

PAGE 103

103 Herbage Phosphorus Concentration Total-season P concentration was not aff ected by ring number in either the dung( P > 0.498) or urine-treated plots (P >0.342) (Tab les A-2 and A-3, respectively). There were no interactions involving ring number for either dung ( P > 0.376) or urine ( P > 0.146). Herbage P concentrations were 3.4, 3.3, and 3.4 g kg-1 in Rings 1 through 3, respectively, in dung-treated plots. In spite of large amounts of total P app lied in dung, there was little apparent uptake of P from dung (Chapter 3) and as such there was littl e impact on spatial characteristics of herbage P. In urine-treated plots, P concen trations were 3.3, 3.4, 3.3, and 3.6 g kg-1 for Rings 1 through 4, respectively. The primary effect of urine on P concentration was associ ated with application frequency (Chapter 3) and likely due to greater soil exploration by roots of more vigorous plants that had received more urine. Spatial variability (i.e., ring effect) in P concentration was minimal across the area sampled regardless of applic ation frequency or management intensity. Adeli and Varco (2001) evaluated the us e of swine lagoon effluent on bermudagrass [ Cynodon dactylon (L.) Pers.] and johnsongrass [ Sorghum halepense (L.) Pers.]. They suggested that total P accumulation by forage grasses is mo re closely related to DM production rather than tissue P concentration, which vari ed little. Newman et al. (2005) assessed the effect of N fertilization on tissue P concentration and remova l. Bahiagrass tissue P concentration decreased from 3.7 for the no-N control to 3.1 g kg-1 when N was applied at 45 kg ha-1 per harvest, but there was no change in the response as N fertilization increased to 60 and 90 kg ha-1 per harvest. Herbage DM harvested more than doubled across th is range of treatments. In urine-treated plots in the current study, DM harvested and herbage N concentration were greatest closest to the center of urine applications, and for most applica tion frequencies decreased as distance from the center increased. However, greater DM harves ted and herbage N concentration were not accompanied by decreasing herbage P concentration as occurred in the Newman et al. (2005)

PAGE 104

104 study. In results of the current study reported in Chapter 3, herbage P concentration was greater for the control than when a single applic ation of urine was made (3.45 vs. 2.83 g P kg-1, respectively), similar to the Newman et al. (2005) data. Herbage IVDOM For dung-treated plots, there were no interactions involving ring number ( P > 0.249; Table A-2), but there were linear ( P = 0.011; Table A-2) effects of ring number on IVDOM. As ring number increased from R1 to R3 in dung-treate d plots, IVDOM decreased from 570 to 560 g kg-1. The biological implications of this small change are not likely to be great. Greater bahiagrass herbage IVDOM has been reported with greater N fertilization by Newman et al. (2006) and Stewart et al. (2007), however in thes e studies it was accompanied by greater herbage N concentration. In the current study, ther e was no ring number effect on herbage N concentration, thus the reason for greater IVDOM in R1 and R2 is not clear. Similar results to those in the current study we re obtained for dung-affected areas of perennial ryegrass (Greenhalgh and Reid, 1968). They observed cattle in pastures and marked fouled areas and paired clean areas. Averaged acro ss two grazing intensities, they found that the fouled areas had an in vitro digestibility of 757 g kg-1, which was greater than 740 g kg-1 in the clean areas. Total-season IVDOM for urine-treated plots was affected by an application frequency X ring number interaction ( P = 0.054), and the ring number main effect was also significant ( P = 0.083) (Table A-3). Interaction occurr ed because there was a linear ( P < 0.001) decline in IVDOM (from 584 to 562 g kg-1) with increasing ring number for Application Frequency 2, but there was no effect for Application Frequencies 0 and 1 ( P > 0.272) (Fig. 4-5). There was a trend ( P = 0.145) toward decreasing IVDOM with increas ing ring number for th e three applications per year treatment. In general this follows th e pattern of response reported by Newman et al. (2006) and Stewart et al. (2007) where greater bahiagrass herbage N concentration was

PAGE 105

105 associated with greater IVDOM. This was true for Frequency 2 where both N concentration and IVDOM decreased with increasing ring number and this was also the tendency for Frequency 3. For frequencies 0 and 1, neither N concen tration nor IVDOM decreased as ring number increased. The departure from the expected trend occurred with Frequency 3, and in particular Ring 3 within that frequency treatment (Fig. 4-5) No other papers were found in the literature that reported digestibility re sponses to urine application. 550 555 560 565 570 575 580 585 590 595 600 1234 RingIVDOM (g kg-1) Urine-3 Urine-2 Urine-1 Urine-0 Figure 4-5. Ring number X urine ap plication frequency interaction ( P = 0.054) for in vitro digestible organic matter (IVDOM) concen tration during 2006. Ring number effects were linear ( P <0.0001) for two applications of ur ine and not significant for three ( P > 0.144), one, and zero applications ( P > 0.272). Standard error of a treatment mean was 2.4 g kg-1. Ring Numbers 1 through 4 refer to sampling areas 0 to 15, 15 to 30, 30 to 45, and 45 to 60 cm from the center of the urine applic ation, respectively.

PAGE 106

106 Herbage Nitrogen Harvested Nitrogen harvested in bahiagrass herbage wa s affected by application frequency X ring number interaction for urine( P < 0.001; Table A-3) a nd for dung-treated plots ( P = 0.019; Table A-2). The ring number X app lication frequency interaction fo r urine-treated plots occurred because there was a linear decline in N harveste d with increasing ring number for Application Frequencies 1, 2, and 3, but there was no effect of ring number for the control (Fig. 4-6). Ring number means for the control ranged only from 35 to 39 kg ha-1, while for Frequencies 1 through 3, N harvested decreased from 64 to 37 kg ha-1, 98 to 42 kg ha-1, and 106 to 45 kg ha-1, respectively, as ring number increased from one to four (Fig. 4-6). Nitrogen harvested showed very similar patterns to those for herbage DM harvested and was driven primarily by the DM harvested response as opposed to herbage N concentration. The effect of application frequency within each ring number was also assessed for N harvested to determine how far from the center of application an effect of urine could be detected. This respon se showed linear ( P < 0.001) and quadratic ( P < 0.069) effects of application frequency for Rings 1 and 2 and a linear effect ( P < 0.001) for Ring 3. Thus, for Rings 1 through 3, increasing urine application fre quency resulted in greate r herbage N harvested (Fig. 4-6). As with herbage N concentration, the data began to converge markedly in Ring 4, but the linear effect (P = 0.030) remained. This i ndicates that the effect of urine application on herbage N harvested extended to 60 cm from th e center of the urine ev ent and 15 cm beyond the effect of urine on DM harvested (Fig. 4-2). For dung-treated plots, total-season N harvested was affected by a ring number main effect ( P < 0.001) and an interaction of app lication frequency X ring number ( P = 0.019). The interaction occurred because th ere was no effect of ring number on N harvested for Application Frequencies 0 ( P > 0.167) and 1 ( P > 0.288), but there were linear ( P < 0.001) and quadratic ( P =

PAGE 107

107 0.007) effects for two dung applications and quadratic ( P = 0.048) effects for three applications (Fig. 4-7). This response is similar to that fo r DM harvested because there was no effect of dung on herbage N concentration. Thus, as observed for DM harvested, an importa nt factor affecting the N harvested response appears to be physical interference of dung in Ri ng 1 that reduced both DM and N harvested (Figs. 4-3 and 4-7). Th is effect was most pronounced when dung was applied two or three times per year. The effect of application frequency with in a level of ring number was assessed to determine how far from the cente r of application dung affected N harvested. There were a linear ( P = 0.055) and cubic ( P = 0.047) effects of application freq uency for Ring 1. Linear effects reveal the general pattern of decreasing N ha rvested with increasing numbers of excreta applications. The cubic effect was significant b ecause two dung applicati ons actually depressed N harvested more than three applications. The bi ological significance of th e cubic effect is not clear. There were no effects of a pplication frequency for Rings 2 ( P > 0.441) or 3 ( P = 0.194) indicating that dung application had no effect on the N harveste d response outside of the immediate 15-cm radius circle to which it was applied. The minimal positive impact of dung, despite containing high amounts of N, is attrib uted to the slow physical breakdown of the dung pats and low mineralization rates of orga nic N in dung (Deenen and Middelkoop, 1992).

PAGE 108

108 0 20 40 60 80 100 120 1234 RingN Harvested (kg ha-1) Urine-3 Urine-2 Urine-1 Urine-0 Figure 4-6. Ring number X urine app lication frequency interaction ( P < 0.001) for bahiagrass herbage N harvested during 2006. The ring number effect was linear for three ( P < 0.001), two ( P < 0.001), and one ur ine application ( P < 0.001) and there was no effect of ring number ( P >0.167) for the no urine control. Standard error of a treatment mean was 1.2 kg ha-1. Ring Numbers 1 through 4 refer to sampling areas 0 to 15, 15 to 30, 30 to 45, and 45 to 60 cm from the cente r of the urine applic ation, respectively.

PAGE 109

109 20 25 30 35 40 45 123 RingN Harvested (kg ha-1) Dung-0 Dung-1 Dung-2 Dung-3 Figure 4-7. Ring number X dung applic ation frequency interaction ( P = 0.019) for bahiagrass herbage N harvested during 2006. The ri ng number effect was quadratic ( P = 0.048) for three applications and linear ( P < 0.001) and quadratic ( P = 0.007) for two dung applications, but was not significant ( P > 0.288) for one and no dung applications ( P > 0.167). Standard error of a treatment mean was 1.5 kg ha-1. Ring Numbers 1 through 3 refer to sampling areas 0 to 15, 15 to 30, and 30 to 45 cm from the center of the dung applicati on, respectively. Herbage Phosphorus Harvested For both dungand urine-treated plots there was a ring number main effect ( P 0.003) on herbage P harvested and an applicati on frequency X ring number interaction ( P = 0.0584 for dung and P < 0.001 for urine; Tables A-2 and A-3). The application X ring number interaction in urine-treated plots occurred because P harvested decreased with incr easing ring number for Applicati on Frequencies 1 through 3, but there was no effect of ring number for the zero application frequency (Fig. 4-8). This pattern of response is nearly identical to DM harveste d because there was no effect of ring number on

PAGE 110

110 herbage P concentration in urine-tr eated plots (average of 3.4 g kg-1). These results support the conclusions of Adeli and Varco ( 2001) that total P removed in fo rage grasses is more closely related to DM production rather than tissu e P concentration, whic h varies much less. The effect of urine application frequenc y on P harvested within a ring number was significant for R1 (linear, P <0.001), R2 (linear, P < 0.001; quadratic, P =0.051), and R3 (linear, P < 0.001), but there was no effect of application fr equency in R4 (Fig. 4-8). Thus like for DM harvested, the impact of urine on P harvested extended up to 45 cm from the center of the application or 15 cm beyond the edge of the applicati on (Fig. 4-8). For dung-treated plots, there was an applica tion frequency X ring number interaction for P harvested that occurred primarily because of physical interference of dung in R1 (Fig. 4-9). This led to lower P harvested for tw o and three dung applications yr-1 in R1 compared to one or zero applications (application fr equency effect linear, P = 0.047; quadratic, P = 0.041), while in Rings 2 and 3 there was no effect of a pplication frequency on P harvested ( P > 0.401 for Ring 2 and P > 0.137 for Ring 3). This response is reflec ted in quadratic ring number effects ( P = 0.073) on P harvested for three dung applications, linear ( P = 0.004) and quadratic ( P = 0.046) effects for two applications, and no effect for one ( P > 0.257) and zero applications ( P > 0.289). Similar to the responses for N and DM harvested, there was no effect of dung application on P harvested beyond R1, the area actually covered by the dung pat. Lack of an effect of dung on P harvested is likely associated with the slow breakdown of dung pats in the curre nt study, perhaps due to drier than normal weather and to little apparent ac tivity of dung beetles whic h incorporate dung into the soil and enhance nutrient releas e (Williams, 1950; Hughes et al., 1975).

PAGE 111

111 0 5 10 15 20 25 1234 RingP Harvested (kg ha-1) Urine-3 Urine-2 Urine-1 Urine-0 Figure 4-8. Ring number X urine ap plication frequency interaction ( P < 0.001) for bahiagrass herbage P harvested during 2006. Ring number effects were linear ( P < 0.001) and cubic ( P = 0.0419) for three urine app lications, linear for two ( P < 0.001) and one application ( P = 0.019), and not significant ( P > 0.194) for zero applications. Standard error of a treatment mean was 0.4 kg ha-1. Ring Numbers 1 through 4 refer to sampling areas 0 to 15, 15 to 30, 30 to 45, and 45 to 60 cm from the center of the urine application, respectively.

PAGE 112

112 4 5 6 7 8 9 10 11 123 RingP Harvested (kg ha-1) Dung-3 Dung-2 Dung-1 Dung-0 4-9. Ring number X dung application frequency inte raction (P = 0.058) for bahiagrass herbage P harvested during 2006. Ring number effects we re quadratic (P = 0.073) for three dung applications, linear (P = 0.004) and quadra tic (P = 0.046) for two applications, and not significant for one (P > 0.257) or for zer o applications (P > 0.289). Standard error of a treatment mean was 0.4 kg ha-1. Ring Numbers 1 through 3 refer to sampling areas 0 to 15, 15 to 30, and 30 to 45 cm from the center of the dung application, respectively. Summary and Conclusions The objectives of this research were to determine the effects of excreta type and application frequency to bahiagrass swards on spa tial patterns of herbage DM harvested, herbage N and P concentration, and N and P harvested re sponses. Spatial patterns of response were determined by sampling swards in concentric ri ngs (termed ring numbers, with Ring 1 being the area closest to the center of the de posit) surrounding urine and dung deposits. Responses to urine and dung were distinctly di fferent. Urine application typically resulted in greater DM harvested, N and P concentra tion, and N and P harvested. In contrast, dung

PAGE 113

113 application had no measurable im pact on herbage N and P concentrations and decreased herbage DM harvested because of physical interference in the area to which dung was applied. Spatial variation in herbage responses reflected these trends. For urine-treated plots, greatest DM harvested and N and P concentrations occurred generally in Ring 1, enco mpassing the center of the area of urine application. He rbage responses typically decreas ed as distance from the center of urine application increased. Significant eff ects of urine on DM harvested and herbage P harvested extended only 15 cm beyond the edge of the area to which urine was applied, while this distance was 30 cm for he rbage N concentration and N ha rvested. For dung-treated plots, especially those receiving multiple dung applicat ions, physical interference reduced herbage DM harvested in Ring 1 by more than 500 (three dung applications yr-1) or 1100 kg ha-1 (two dung applications). Herbage DM harvested and herbag e N and P concentrations in areas outside of Ring 1 were similar for plots treated with dung a nd the no dung control. T hus, unlike urine there was no measurable effect of dung beyond the immedi ate area to which it was applied. Lack of positive response to dung was due to physical inte rference, as already mentioned, but it was accentuated by limited apparent dung beetle activity, the drier than normal weather that may have reduced breakdown of dung, and the high proporti on of N in dung that is in an organic form and only slowly released for plant growth. These data provide rationale for efforts to increase uniformity of excreta deposition on pastures. By avoiding repeated dung applications in the same area, there is opportunity to avoid much of the negative impact of physical interference. Responses to single applications of urine are often significant, especially in terms of DM harvested, and additional a pplications have lesser impact (Fig. 4-2). Thus management that increa ses the uniformity of urine distribution should

PAGE 114

114 also increase overall pasture herbage accumu lation and, based on results from Chapter 3, increase the efficiency of N recovery.

PAGE 115

115 CHAPTER 5 SUMMARY AND CONCLUSIONS Pasture-based, forage-livestock systems in Florida are planted primarily to bahiagrass ( Paspalum notatum Flgge) because of its tolerance to close grazing and adaptation to a wide range of soil conditions including low fertili ty, drought, and short-term flooding. Bahiagrass pastures support approximately 1 million head of beef cattle in Florida. Even though these pastures typically receive low amounts of fertilizer and are stocked at low to moderate rates, the quantity of land area occupied and the number of animals supported statewide on bahiagrass pasture make nutrient management a key issue. The importance of nutrient management also is related to the sensitivity of Florida ecosystems, especially in terms of nutrient impacts on water quality (Nair et al., 2007). On agricultural la nds, the lateral flow of P to surface water, particularly in Central and Sout h Florida (Nair et al., 2007), and leaching of nitrates to ground water, more commonly in North Florida, ar e critical concerns (W oodard et al., 2002). One of the many challenges faced by livesto ck producers utilizing pasture as a feed resource is avoidance of over accu mulation of livestock wastes in certain areas of the pasture. Uneven distribution of soil nutrients, due to non-uni form excreta deposition, is thought to lead to leaching and volatilization of N and runoff of P a nd other nutrients from so-called nutrient hot spots in pastures, yet the effect of in situ excreta deposition on plant, soil, and water responses has received limited attention. The research reported in this th esis has as its objectives to characterize the impact of frequency and type of excreta application on bahiagrass herbage dry matter harvested, chemical composition, and excr eta nutrient recovery, and to measure P and NO3-N concentration in shallow soil water under excreta applications. Treatments were two management intensities (Average and High), two types of excreta (dung and urine), and three frequencies of excreta application (one, two, or three times per year).

PAGE 116

116 Excreta was collected from animals grazing ba hiagrass pasture managed at Average and High intensities and applied to ungrazed bahiagrass plots fertilized with the same amount of N as the excreta source pasture from which the excret a was obtained. Average and High management intensities imposed on the excreta source pastures were defined based on N fertilizer amount and stocking rate. Average management intens ity source pastures received 60 kg N ha-1 yr-1 and were stocked with two yearling heifers ha-1, and High pastures received 120 kg N ha-1 yr-1 and were stocked with four yearling heifers ha-1. These two management intensities were selected to represent fertilization and stoc king regimes that are common in the Florida livestock industry (Average) or represent approximately the most intensive management applied to grazed bahiagrass (High) in Florida. Excreta from cattle grazing Average management intensity pastures was applied to bahiagrass pl ots that received 60 kg N ha-1 yr-1, and excreta from cattle grazing High management intensity pastures was a pplied to plots that received 120 kg N ha-1 yr-1. Data were reported in two chapters. Chapter 3 describes shallow soil water responses and those of herbage within a 45-cm radius of th e excreta application. In Chapter 4, the spatial patterns of herbage responses, be ginning at the center of excret a application and moving away from the application, were described. For herbage responses in the circle of 45cm radius surrounding the excreta application (Chapter 3), the primary treatments affecting th e responses were management intensity and the excreta application frequency X excreta type interaction. For plots r eceiving excreta, as management intensity increased from Average to High, forage DM harvested increased from 3230 to 3850 kg ha-1. High management intensity plots had greater N concentration than Average intensity (14.7 and 14.2 g kg-1, respectively), greater IVDOM than Average plots (577 vs. 567 g kg-1), greater N harvested than Average (57 vs. 47 kg ha-1), and greater P harvested

PAGE 117

117 than Average management intensity (10.9 vs. 8.0 kg ha-1 in dung-treated plots, respectively, and 13.7 vs. 12.4 kg ha-1 in urine-treated plots, respectively). There was excreta application frequency X ex creta type interaction for most herbage responses in Chapter 3. Intera ction occurred because dung app lication had no effect on most herbage responses, whereas respon ses to urine were significant consistently. In urine-treated plots, herbage DM harvested increased from 2760 at zero urine applications to 4670 kg ha-1 with three applications. Over the same urine appl ication frequencies, herbage N concentration increased from 13.3 to 16.2 g kg-1, P concentration from 3.45 to 3.87 g P kg-1, in vitro digestible organic matter concentration from 569 to 588 g kg-1, N harvested from 37 to 75 kg ha-1, and P harvested from 9 to 18 kg ha-1. Greater herbage response to urin ethan dung-treated plots was expected because of the high proportion of dung N th at is in an organic form and the greater availability to plants of nutrients in urine. The general absence of response to dung was not expected and could be attributed to a number of fact ors including physical interference of the dung pat, the high concentration of organic N as a proportion of total N, limited apparent activity of dung beetles, and a drier than normal year leading to rapid drying and crusting of dung. Excreta N recovery in harvested herbage was greater for urine than dung, averaging 22 and 2%, respectively. Recovery was also affected by excreta application frequency, decreasing from 28 to 18% as urine application frequency increased from one to three yr-1, and from 4 to less than 1% as dung application frequency in creased from one to three. Thes e recoveries are in the lower part of the range reported in th e literature, but it does not appear that leaching lo sses explain this response. Significant responses of shallow soil water NO3-N concentrations to treatment occurred at only one sampling da te (15 Sept. 2006), and these va lues were less than 0.15 mg L-1. Greater concentrations occurred in individual wells at two other dates in August, but these were

PAGE 118

118 restricted to a single replicat e within a treatment. Warm, dry weather may have increased volatilization losses of urine N (R usselle, 1996), and bahiagrass has been reported to store large quantities of N in rhizomes and roots (Blue, 1973). Low values for dung N recovery and absence of herbage response to dung suggest limited minera lization of nutrients in dung during the course of the growing season. Greater pe rcent recovery of N in excret a occurred with single excreta applications suggesting that grazing management practices which increase uniformity of excreta deposition will likely increase efficiency of N cycling in grazed grasslands. The objectives of the research reported in Chapter 4 were to determine the effects of excreta type and application fre quency to bahiagrass swards on sp atial patterns of herbage DM harvested, herbage N and P concentration, and N and P harvested responses. Spatial patterns of response were determined by sampling swards in concentric rings (termed ring numbers, with Ring 1 being the area closest to the center of the deposit) su rrounding urine and dung deposits. Responses to urine and dung were distinctly di fferent. Urine applica tion typically resulted in greater herbage DM harvested, N and P conc entration, and N and P ha rvested. In contrast, dung application had no measurable impact on he rbage N and P concentrations and decreased herbage DM harvested because of physical in terference in the area to which dung was applied. Spatial variation in herbage responses reflected th ese trends. For urine-treated plots, greatest DM harvested and nutrient concentrations occurred generally in Ring 1, enco mpassing the center of the area of urine application. He rbage responses typically decreas ed as distance from the center of urine application increased. For dung-treated plots, especially those receiving multiple dung applications, physical interference reduced herbag e DM harvested in Ring 1 by more than 500 kg ha-1.

PAGE 119

119 In spite of the magnitude of the response to urine, significant eff ects did not extend large distances beyond the edge of the urine application. For DM harvested and herbage P harvested, these effects extended only 15 cm beyond the edge of the area to which urine was applied, while this distance was 30 cm for herbage N concentr ation and N harvested. In contrast, herbage DM harvested and N and P concentrations in areas ou tside of Ring 1 were similar for plots treated with dung and the no dung control. Thus, unlike urine there was no measurable effect of dung beyond the immediate area to which it was applied. Lack of positive response to dung was due to physical interference, as already mentioned, but it was accentuated by limited apparent dung beetle activity, the drier than normal weather that may have re duced breakdown of dung, and the high proportion of nutrients in dung that are in an organic form and only slowly released for plant growth. Data from both chapters provide justifica tion for management practices that increase uniformity of excreta deposition on pastures. Respons es to single applicati ons of urine are often significant, especially in terms of DM harvested, and additional a pplications have lesser impact. Single applications of dung and urine result in a greater percentage N recovery than multiple applications. In addition, by avoi ding repeated dung applications in the same area, there is opportunity to avoid much of the negative impact of physical interference. Thus management practices that increase the uni formity of excreta distributi on should also increase herbage accumulation on pastures and increase the efficiency of recapture by plants of nutrients in excreta. Dubeux (2005) and Dubeux et al. (2006) provide evidence that rotational stocking, especially rotational stocking with short grazing periods, in creases uniformity of excreta distribution. Though rotational stoc king is not widely used at present by Florida livestock producers, rapidly increasing fertil izer costs and concerns about the impact of cattle excreta on

PAGE 120

120 water quality may result in greate r use of this practice in the future. Implementation of rotational stocking may make important contributions to the overall economi c and environmental sustainability of bahiagrass-li vestock systems in Florida.

PAGE 121

121APPENDIX STATISTICAL TABLES Table A-1. P values for interactions and main effects on response variables discussed in Chapter 3. Source of Variation Response variable Management intensity (MI) Application frequency (AF) Excreta type (ET) MI x AF MI x E AF x E MI x AF x E DM Harvested 0.0004 0.2224 <0.0001 0.7093 0.2288 0.0633 0.2996 Herbage N 0.0385 <0.0001 <0.0001 0.5707 0.7387 0.0001 0.9200 Herbage P 0.9235 0.0033 0.8355 0.7151 0.1743 0.0019 0.1776 Herbage IVDOM 0.0097 0.0045 0.0014 0.1892 0.9280 0.0987 0.1210 N Harvested <0.0001 0.0025 <0.0001 0.6270 0.4111 0.0008 0.3279 P Harvested 0.0006 0.0001 <0.0001 0.6256 0.0327 <0.0001 0.6859 N Recovery 0.5657 0.1249 <0.0001 0.5204 0.4631 0.5453 0.3771

PAGE 122

122 Table A-2. P values for main effects, interactions and polynomial contrasts for ring number for dung-treated plots as discussed in Chapter 4. Source of Variation Excreta Type Response variable Management Intensity (MI) Application Frequency (AF) Ring # (R)MI x AFMI x R AF x R MI x AF x R Dung DM Harvested <0.0001 0.1247 L-0.0918 Q-0.0009 0.3522 0.9914 0.0271 0.8681 Herbage N <0.0001 0.2144 L-0.5948 Q-0.1608 0.1771 0.7763 0.6869 0.7432 Herbage P 0.1498 0.9932 L-0.6190 Q-0.4989 0.4429 0.3765 0.8761 0.6519 Herbage IVDOM 0.0005 0.1888 L-0.0108 Q-0.4093 0.7276 0.7383 0.2491 0.9789 N Harvested <0.0001 0.2143 L-0.0328 Q-0.0009 0.5474 0.9210 0.0194 0.8024 P Harvested <0.0001 0.1845 L-0.1447 Q-0.0030 0.5862 0.5769 0.0584 0.4858 L = linear and Q = quadratic effect of ring number.

PAGE 123

123Table A-3. P values for main effects, interactions and polynomial contrasts for ring number for urine-treated plots as discussed in Chapter 4. Source of Variation Excreta Type Response variable Management Intensity (MI) Application Frequency (AF) Ring # (R) MI x AFMI x R AF x R MI x AF x R Urine DM Harvested <0.0001 <0.0001 L-<0.0001 Q-0.5930 C-0.0188 0.1053 0.5152 <0.0001 0.4439 Herbage N <0.0001 <0.0001 L-<0.0001 Q-0.7160 C-0.3752 0.0431 0.0009 <0.0001 0.2391 Herbage P 0.8317 <0.0001 L-0.3421 Q-0.5378 C-0.3637 0.9125 0.1458 0.1839 0.4481 Herbage IVDOM 0.0329 0.0028 L-0.0828 Q-0.9581 C-0.5357 0.0661 0.7085 0.0539 0.7354 N Harvested <0.0001 <0.0001 L-<0.0001 Q-0.7771 C-0.0357 0.0796 0.4811 <0.0001 0.5471 P Harvested 0.0024 <0.0001 L-<0.0001 Q-0.7856 C-0.0097 0.3842 0.3090 <0.0001 0.1828 L = linear, Q = quadratic, and C = cubic effect of ring number.

PAGE 124

124 REFERENCES Adeli, Ardeshir, and Jac J. Varco. 2001. Sw ine lagoon effluent as a source of nitrogen and phosphorus for summer forage grasses. Agron. J. 93:1174-1181. Arthington, John, Patrick Behlen, and Fritz R oka. 2003. Effect of stocking rate on measures of cow-calf pro ductivity and nutrient loads in surface water runoff. EDIS. May 2003. http://edis.ifas.ufl.edu/AN141 Allen, A.G., S.C. Jarvis, and D.M. Headon. 1996. Nitrous oxide emissions from soils due to inputs of nitrogen from excreta return by liv estock on grazed grassland in the UK. Soil Biology & Biochemistry 28:597-607. Anger, M., C. Hoffman, and W. Kuhbauch. 2003. Nitrous oxide emissions from artificial urine patches applied to different N-fert ilized swards and estimated annual N2O emissions for differently fertilized pastures in an upland location in Germany. Soil Use and Management 52:191-201. Ball, P.R., and D.R. Keeney. 1981. Nitrogen lo sses from urine affected areas of a New Zealand pasture. Proceedings of the XIV International Grassland Congress. Edited by J. Allan Smith and Virg il W. Hays. June 15-24. p. 42-344. Ball, Roger, D.R. Keeney, P.W. Theobald, a nd P. Nes. 1979. Nitrogen balance in urine affected areas of a New Zealand pasture. Agron. J. 71:309-314. Bardgett, R.D., and D.A. Wardle. 2003. Herbivore-mediated linkages between aboveground and belowground communities. Ecology 84:2258-2268. Barnes, R.F., and C.J. Nelson. 2003. Forages a nd grasslands in a changing world. p. 3-23. In R.F. Barnes et al. (ed.), Forages: An introduction to grassland ag riculture. Iowa State Press, Ames, IA. Barrow, N.J. 1967. Some aspects of the effects of grazing on the nutrition of pastures. J. Aust. Inst. Agric. Sci. 33:254-262. Barton, L., A. Schipper, G.F. Barkle, M. Mc Leod, T.W. Speir, M.D. Taylor, A.C. McGill, A.P. van Schaik, N.B. Fitzgerald, and S.P. Pandey. 2005. Waste Management: Land application of domestic effluent onto four so il types: plant uptake and nutrient leaching. J. Environ. Qual. 34: 635-643. Bathurst, N.O. 1952. The amino acids of sheep and cow urine. J. Agric. Sci. 42:476-478. Beaty, E.R., R.H. Brown and J.B. Morris. 1968. Response of Pensacola bahiagrass to intense clipping. p. 538-542. In Proc. 11th Int. Grassl. Congress. Surfers Paradise, Australia.

PAGE 125

125 Beaty, E.R. and John D. Powell. 1978. Gr owth and management of Pensacola bahiagrass. Journal of Soil a nd Water Conservation 33-34:191-193. Beaty, E.R., K.H. Tan, R.A. McCreery, and J. B. Jones. 1975. Root-herbage production and nutrient uptake and retention by berm udagrass and bahiagrass. Journal of Range Management. 28:385-389. Blue, W.G. 1970. Fertilizer nitrogen uptake by Pensacola bahiagrass from Leon fine sand, a Spodosol. Proc XI Int. Grassla nd Congress, Surfers Paradise, Queensland, Australia. p. 389-392. Blue, W.G. 1973. Role of Pensacola bahiagrass st olon-root systems in fertilizer nitrogen utilization on Leon fine sand. Agronomy Journal. 65: 88-91. Boddey, R.M., R. Macedo, R.M. Tarr, E. Ferreira, O.C. de Oliveira, C. de P. Rezende, R.B. Cantarrutti, J.M. Pereira, B.J.R. Alves, and S. Urquiaga. 2004. Nitrogen cycling in Brachiaira pastures: The key to understadi ng the process of pasture decline. Agriculture, Ecosyste ms and Environment. 103:389-403. Burson, B.L., and V.H. Watson. 1995. Bahiag rass, dallisgrass, and other Paspalum species. p. 431-440. In R.F. Barnes et al. (ed.), Forages. Vol. 1. 5th ed. Iowa State University Press, Ames, IA. Burton, G.W. 2004. Warm Season (C4) Grasses. In L.E. Moser, B.L. Burson, L.E. Sollenberger (eds.), Warm-season (C4) Grasses. ASA, CSSA, SSSA Agronomy Monograph no. 45. 217-255. Burton, G.W., R.N. Gates, and G.J. Gascho. 1997. Re sponse of Pensacola bahiagrass to rates of nitrogen, phosphorus and potassium fertilizer s. Soil Crop Sci. Soc. Fla. Proc. 56:31-35. Burton, G.W.. 1967. A search for the origin of Pensacola bahiagrass. Econ. Bot. 21:319-382. Bussink, D.W., and O. Oenema. 1998. Ammoni a volatilization from dairy farming systems in temperate areas: A review. Nutrient Cycling in Agroecosystems. 51:19-33. Carran, R.A., and P.W. Theobald. 1999. Effect s of excreta return on properties of a grazed pasture soil. Nutrient Cycling in Agroecosystems. 56:79-85. Carran, R.A., P.W. Theobald, and J.P. Evans. 1995. Emission of nitrous oxide from some grazed pasture soils in New Zealand. Australian Journal of Soil Research. 33:341-352. Chambliss, C.G., and M.B. Adjei. 2006. Bahi agrass. Publication SS-AGR-36, Agronomy Department, University of Florida, Gainesville.

PAGE 126

126 Cherney, Debbie J. R., Jerome H. Cherne y, and Elena A. Mikhailova. 2002. Orchardgrass and tall fescue utilizati on of nitrogen from dairy manure and commercial fertilizer. Agron. J. 94:405-412. Clough, Tim J., Francis M. Kelliher, Robert R. Sh erlock and Colleen D. Ford. 2004. Lime and soil moisture effects on nitrous oxide emissions from a urine patch. Soil Science Society of America Journal. 68:1600-1609. Coleman, S.W., J.E. Moore, and J.R. W ilson. 2004. Quality and utilization. p. 267-308. In L.E. Moser, B.L. Burson, L.E. Sollenberger (eds.), Warm-season (C4) Grasses. ASA-CSSA-SSSA, Madison, WI. Correll, David L. 1996. Environmental impact of pasture systems on surface water quality. p. 13-28. In R.E. Joost and C.A. Roberts (ed.) Nutrie nt cycling in forage systems. Potash and Phosphate Institute and The Foundation for Agronomic Research, Manhattan, Kansas. Crowder, L.V., and H.R. Chheda. 1982. p. 177. In Tropical grassland husbandry. Longman Publishers, New York. Cuttle, S.P., and P.C. Bourne. 1993. Uptake and l eaching of nitrogen from artificial urine applied to grassland on different dates dur ing the growing season. Plant and Soil. 150:77-86. Dai, Xiaobing. 2000. Impact of cattle dung de position on the distribution pattern of plant species in an alvar limestone gra ssland. Journal of Vegetation Science. 11:715-724. Dalrymple, R.L., R. Stevens, T. Carroll, and B. Flatt. 1994. Forage production benefits from nutrient recycling via beef cattle and how to manage for nutrient recycling in a grazing cell. p. 269-273. In American Forage and Grassland Council Proc. Lancaster, PA. 6-10 March 1994. Day, T.A., and J.K. Detling. 1990. Grassla nd patch dynamics and herbivore grazing preference following urine de position. Ecology. 71:180-188. de Klein, C.A.M. 2001. An analysis of envir onmental and economic im plications of nil and restricted grazing systems designed to reduce nitrate leaching from New Zealand dairy farms. II. Pasture prod uction and cost/benefit analysis. New Zealand Journal of Agricu ltural Research. 44:217-235. Decau, M.L., J.C. Simon, and A. Jacquet. 2003. Fate of urine nitr ogen in three soils throughout a grazing season. J. Environ. Qual. 32:1405-1413.

PAGE 127

127 Deenen, P.J., and N. Middelkoop. 1992. Eff ects of cattle dung and urine on nitrogen uptake and yield of perennial ryegrass. Netherlands Journal of Agricultural Science. 40:469-482. Dickinson, C.H., and G. Craig. 1990. Effect s of water on the decomposition and release of nutrients from cow pa ts. New Phytologist 115:139-147. Doak, B.W. 1952. Some chemical changes in th e nitrogenous constituents of urine when voided on pasture. Journal of Agricultural Science 42:162-171. Dubeux, J.C.B., Jr. 2005. Management strategies to improve nutrient cycling in grazed Pensacola bahiagrass pastures. Ph.D. dissert ation, Univ. of Florida, Gainesville. Dubeux, J.C.B, Jr., L.E. Sollenberger, B.W. Math ews, J.M. Scholberg, and H.Q. Santos. 2007. Nutrient cycling in warm-climate grasslands. Crop Sci. 47:915-928. Dubeux, J.C.B., Jr., R.L. Stewart, Jr., L.E. Sollenb erger, J.M.B. Vendramini, and S.M. Interrante. 2006. Spatial heterogeneity of he rbage response to management intensity in continuously stocked Pensacola bahiagrass pastures. Agron. J. 98:1453-1459. During, C., and W.C. Weeda. 1973. Some ef fects of cattle dung on soil properties, pasture production, and nutrient uptake. N.Z. J. Agric. Res. 16:423-430. Edwards, D.R. 1996. Recycling livestock manure on pastures. p. 45-63. In R.E. Joost and C.A. Roberts (ed.) Nutrient cycling in forage syst ems. Potash and Phosphate Institute and The Foundation for Agronomic Research, Manhattan, Kansas. Eghball, Bahman, Brian J. Wienhold, Brya n L. Woodbury, and Roger A. Eigenberg. 2005. Phosphorus management. Agron. J. 97:542-548. Elliot, H.A., G.A. OConnor, P. Lu, and S. Brinton. 2002. Influence of water treatment residuals on phosphorus solubility and leaching. J. Environ. Qual. 31:1362-1369. Finlayson, E.H. 1941. PensacolaA new fine leaved bahia. South. Seedsman. 12: 9. Firestone, M.K. 1982. Biologi cal denitrification. p. 289-326. In F.J. Stevenson (ed.) Nitrogen in agricultural soils. Agron. Monogr. No.22. ASA and SSSA, Madison, WI. Fuentes, Barbara, Nanthi Bolan, Ravi Na idu, and Maria de la Luz Mora. 2006. Phosphorus in organic waste-soil systems. Revista de la ciencia del suelo y nutricion vegetal. 6:64-83. Gallaher, R. N., C.O. Weldon, and J.G. Futral. 1975. An aluminum block digester for plant and soil analysis. Soil Sci. So c. Amer. Proc. 39:803-806.

PAGE 128

128 Gates, Roger N., Camilo L. Quarin, and Carl os G.S. Pedreira. 2004. Bahiagrass. p. 651-675. In L.E. Moser et al. (ed.) Warm-season (C4) Grasses. Agronomy Monograph no. 45. ASA, CSSA, SSSA. Madison, WI. Graetz, D.A., V.D. Nair, K.M. Portier, and R.L Voss. 1999. Phosphorus accumulation in manure-impacted Spodosols of Flor ida. Agriculture, Ecosystem and Environment. 75:31-40. Greenhalgh, J.F.D., and G.W. Reid. 1968. The effects of grazing intensity on herbage consumption and animal production. III. Dairy cows grazed at two intensities on clean or contaminated pasture. J ournal of Agric. Sci., Camb. 71:111-222. Griffin, T.S., C.T. Honeycutt, and Z. He. 2003. Changes in soil phosphorus from manure application. Soil Sci. Soc. Am. J. 67:645-653. Griffith, W.K., and L.S. Murphy. 1996. Macr onutrients in forage production. p. 13-28. In R.E. Joost and C.A. Roberts (ed.) Nutrient cycli ng in forage systems. Potash and Phosphate Institute and The Foundation for Agr onomic Research, Manhattan, Kansas. Hack-ten Broeke, M.J.D., W.J.M. De Groot a nd J.P. Dijkstra. 1996. Impact of excreted nitrogen by grazing cattle on nitrate l eaching. Soil Use and Management 12:190-198. Hambleton, L.G. 1977. Semiautomated method for simultaneous determination of phosphorus, calcium and crude protein in animal feeds. J.A.O.A.C. 60:845-852. Haynes, R.J., and P.H. Williams. 1992. Changes in soil solution composition and pH in urine-affected areas of past ure. J. Soil Sci. 43:323-334. Haynes, R.J., and P.H. Williams. 1993. Nutrient cycling and soil fertil ity in the grazed pasture ecosystem. Adv. Agron. 49:119-199. Henzell, E.F., and Ross, P.J. 1973. The nitr ogen cycle of pasture ecosystems. p. 227-246. In O.W. Butler and R.W. Bailey (e d.), Chemistry and biochemistry of herbage. Academic Press, London, UK. Hirata, Masahiko. 1994. Response of bahiagrass ( Paspalum notatum Flugge) sward to nitrogen fertilization rate and cutting inte rval. J. Japan Gr assl. Sci. 40:313-324. Hirata, M., Y. Ogawa, N. Koyama, K. Shindo, Y. Sugimoto, M. Higashiyama, S. Ogura, and K. 2006. Fukuyama. Productivity of bahi agrass pastures in South-western Japan: Synthesis of data from grazing trials. J. Agronomy & Crop Science. 192:79-91. Hirata, Masahiko, Yasuhiro Sugimoto, and Masahiko Ueno. 1990. Quantification of cattle dung-related processes in bahi agrass pasture. JARQ. 23:189-195.

PAGE 129

129 Hirata, Masahiko, Yasuhiro Sugimoto, and Masahiko Ueno. 1991. Use of a mathematical model to evaluate the effects of dung from grazing animals on pasture production and util ization and animal production. Journal of Japanese Society of Grassland Science. 37:303-323. Holter, Peter. 1979. Effect of dung-beetles ( Aphodius spp.) and earthworms on the disappearance of cattl e dung. Oikos. 32:393-402. Hughes, R.D., P. Ferrar, A. Macqueen, P. Du rie, G.T. McKinney, and F.H.W. Morely. 1975. Introduced dung beetles and Austra lian pasture ecosystems: Papers presented at a symposium during the mee ting of the Australia and New Zealand Association for the advancement of sc ience at Canberra in January 1975. The Journal of Applied Ecology. 12:819-837. Hutchings, N.J., J.E. Olesen, B.M. Petersen, and J. Berntsen. 2007. Modelling spatial heterogeneity in grazed grassland an d its effects on nitrogen cycling and greenhouse gas emissions. Agriculture, Ecosystems and Environment. 121:153163. Ibrikci, H., E.A. Hanlon, and J.E. Rechci gl. 1999. Inorganic phosphorous and manure effects on bahiagrass production on a spodosol. Nutrient Cycling in Agroecosystems. 54:259-266. Interrante, S.M., L.E. Sollenberger, and A.R. Blount. 2007. Bahiagrass ge notype yield and tiller responses to defoliation frequency and intensity. In 2007 Annual Meeting Abstracts [CDROM]. ASA, CSSA, SSSA, Madison, WI. Iyamuremye, F., R.P. Dick, and J. Baham. 1996. Organic amendments and phosphorus dynamics: I. Phosphorus chemistry and sorption. Soil Sci. 161:426-435. Jabro, J.D., W.L. Stout, S.L. Fales, and R.H. Fox. 1997. Nitrate leaching from soil core lysimeters treated with urine or fe ces under orchardgrass: Measurement and simulation. Journal Envir onmental Quality. 26:89-94. Jaramillo, Victor J., and James K. Detling. 1992. Small-scale hete rogeneity in a semi-arid North American grassland. II. Cattle grazing of simulated urine pa tches. Journal of Applied Ecology. 29:9-13. Jarvis, Steve C., David Scholefield, and Brian Pa in. 1995. Nitrogen cycling in grazing systems. In P.E. Bacon (ed.), Nitrogen fertilization in the environment. Marcel Dekker, New York. Jarvis, S.C., D.J. Hatch, and Daphne H. Roberts. 1989. The effects of grassland management on nitrogen losses from grazed swards through ammonia volatilization; the relationship to excretal N returns from cattle. J. Agric. Sci. Camb. 112:205-216.

PAGE 130

130 Joblin, K.N. 1981. Effect of urine on the elemen tal composition of spring regrowth herbage in a ryegrass pasture. New Zealand Journa l of Agricultural Research. 24:293-297. Joblin, K.N., and R.G. Keogh. 1979. The element composition of herbage at urine patch sites in a ryegrass pasture. J. Agric. Sci. 92:571-574. Johnson, A.F., D.M. Vietor, F.M. Rouquette Jr., and V.A. Haby. 2004. Fate of phosphorus in dairy wastewater and poultr y litter applied on gr assland. Journal Environmental Quality. 33:735-739. Joost, Richard E. 1996. Nutrient cycling in fo rage systems. Nutrient cycling in forage systems. p. 1-11. In R.E. Joost and C.A. Roberts (ed.) Nu trient cycling in forage systems. Potash and Phosphate Institute and The F oundation for Agronomic Research, Manhattan, Kansas. Jorgensen, Finn Vanman and Erik Steen Jense n. 1997. Short-term effects of a dung pat on N2 fixation and total N uptake in a perenn ial ryegrass/white clover mixture. Plant and Soil. 196:133-141. Kanneganti, V.R., and S.D. Klausner. 1994. Nitrogen recovery by orchardgrass from dairy manure applied with or without fert ilizer nitrogen. Commun. Soil Sci. Plant Anal. 25:2771-2783. Kemp, D.R., and D.L. Michalk. 2005. Grassla nds for production and the environment. p. 193-208. In D.A. McGilloway (ed.) Grassland: A global resource. Wageningen Academic Publishers, Wageningen, The Netherlands. Lantinga, E.A., J.A. Keuning, J. Groenwold, and P.J.A.G Deenen. 1987. Distribution of excreted nitrogen by grazing cattle and its effects on sward quality, herbage production and utilization. In H.G. Van der Meer, R.J. Unwin, T.A. van Dijk, and G.C. Ennik (eds.), Animal manure on grassland and fodder crops. Fertilizer or waste? Developments in Plant and Soil Science 30:103-117. Ledgard, S.F., K.W. Steele, and W.H.M Saunde rs. 1982. Effects of cow urine and its major constituents on pasture properties. New Zealand Journal of Agricultural Research. 25:61-68. Lee, C.M., and R. Wall. 2006. Cow-dung colonization and decomposition following insect exclusion. Bulletin of Entomological Research. 96:315-322. Leterme, Philippe, Camille Barre, and Francoise Vertes. 2003. The fate of N15 from dairy cow urine under pasture receiving different rates of N fertilizer. Agronomie. 23: 609-616. Lotero, J., W.W. Woodhouse, Jr ., and R.G. Petersen. 1966. Lo cal effect on fertility of urine voided by grazing cat tle. Agron. J. 58:262-265.

PAGE 131

131 Luo, J., R.W. Tillman and P.R. Ball. 1999. F actors regulating denitrification in a soil under pasture. Soil Biology and Biochemistry. 31:913-927. Ma, Xiuzhi, Shiping Wang, Gaoming Jiang, Siliva Haneklaus, Ewald Schnug, and Paul Nyren. 2007. Short-term effect of targeted placements of sheep excrement on grassland in Inner Mongolia on soil and plant parameters. Communications in Soil Science and Plant Analysis. 38:1589-1604. Marschner, H. 1995. Mineral nut rition of higher plants. 2nd ed. Academic Press, New York. Mathews, B.W., S.C. Miyasaka, and J.P. Tritschler. 2004. Mine ral nutrition of C4 forage grasses. p. 217-266. In L. E. Moser et al. (ed.) Warm-season (C4) Grasses. ASA/CSSA/SSSA, Madison, WI. Mathews, B.W., L.E. Sollenberger, P. Nkedi-Kizza, L.A. Gaston, and H.D. Hornsby. 1994. Soil sampling procedures for monitoring potassium distribution in grazed pastures. Agron. J. 86:121-126. Mathews, B.W., L.E. Sollenberger, and J.P. Tritschler II. 1996. Grazi ng systems and spatial distribution of nutrients in past ures: Soil considerations. p. 213-229. In R.E. Joost and C.A. Roberts (ed.), Nutrient cycling in fora ge systems. Potash and Phosphate Institute and the Foundation for Agronomi c Research, Manhattan, Kansas. MacDiarmid, B.N., and B.R. Watkin. 1972. The cattle dung patch: III. Distribution and rate of decay of dung patche s and their influence on graz ing behaviour. J. Brit. Grassland Soc. 27:48-54. McCaleb, J.E., C.L. Dantzman, and E.M. Hodges. 1966. Response of panolagrass and Pensacola bahiagrass to different am ounts of phosphorus and potassium. Soil Crop Sci. Soc. Fla. Proc. 26:248-256. Mislevy, P., G.W. Burton, A.R. Bl ount, and F.G. Martin. 2005. Dry biomass yield and nutritive value of bahiagrass cultivars in Central Flor ida. Soil Crop Sci Soc. Florida Proc. 64:7579. Monaghan, R.M. and D. Barraclough. 1993. Nitrous oxide and dinitrogen emissions from urineaffected soil under controlled condi tions. Plant and Soil. 151:127-138. Moore, J. E., and G. O. Mott. 1974. Recovery of residual organic matter from in vitro digestion of forages. J. Dairy Sci. 57:1258-1259. Mylavarapu, R., D. Wright, G. Kidder, and C.G. Chambliss. 2007. UF/IFAS standardized fertilization recommendations for agronomic crops. Fla. Agric. Exp. Stn. Fact Sheet SL129. Univ. of Florida, Gainesville.

PAGE 132

132 NRC. 1996. Nutrient requirem ents of beef cattle; 6th ed. National Academy Press; Washington, DC. Nair, V.D., P.K.R. Nair, R.S. Kalmbacher, I.V. Ezenwa. 2007. Reducing nutrient loss from farms through silvopastoral practices in coarse-textured soils of Florida, USA. Ecological Engineering. 29:192-199. National Agricultural St atistics Service. 2006. http://www.nass.usda.gov:8080/ QuickStats/PullDataUS April 2006. Newman, Y.C., M.B. Adjei, J.M. Scholberg, C.G. Chambliss, and J. Rechcigl. 2005. Nitrogen rate effects on productivity and nutritive value of forages used on South Florida dairies. In K. Cassida et al. (ed.), Proc. Am Forage and Grassl. Counc. Vol. 14. Springfield, IL. Newman, Yoana C., Lynn E. Sollenberger, Kenn eth J. Boote, L. Hartwell Allen, Jr., Jean M. Thomas, and Ramon C. Little. 2006. N itrogen fertilization affects bahiagrass responses to elevated atmospheric carbon dioxide. Agron. J. 98:382-387. Norman, M.J.T., and J.O. Green. 1958. The lo cal influence of cattle dung and urine upon the yield and botanical composition of permanent pasture. Journal of the British Grassland Society. 13:39-45. Oenema, O., G.L. Velthof, S. Yamulki, S.C. Jarvis. 1997. Nitrous oxide emissions from grazed grassland. Soil Use and Management. 13:288-295. Overman, A.R. and R.L. Stanley. 1998. Bahiag rass response to appl ied nitrogen and harvest interval. Commun. So il Sci. Plant Anal. 29:237-244. Owens, L.B., W.M. Edwards, and R.W. Van Ke uren. 1992. Nitrate levels in shallow groundwater under pastures receiving ammoni um nitrate or slow release nitrogen fertilizer. J. Environ. Qual. 21:607-613. Pakrou, N. and Peter Dillon. 1995. Preferenti al flow, nitrogen transformations and N15 balance under urine-affected areas of irrigated and non-irrigated clover based pastures. Journal of Contaminant Hydrology. 20:329-347. Pedreira, C.G.S. and R.H. Brown. 1996a. Physiology, morphology, and growth of individual plants of selected and unselected bahiagrass populations. Crop Sci. 36:138-142. Pedreira, C.G.S. and R.H. Br own. 1996b. Yield of selected and unselected Bahiagrass populations at two cutting he ights. Crop Sci. 36:134-137.

PAGE 133

133 Perez, T., S.E. Trumbore, S.C. Tyler, P.A. Matson, I. Ortiz-Monasterio, T. Rahn, and D.W.T. Griffith. 2003. Identifying the agricultural imprint on the global N2O budget using stable isotopes. J. Geophys. Res. Atmos. 106:9869-9878. Petersen, R. G., H.L. Lucas, and W.W. Woodhous e, Jr. 1956. The dist ribution of excreta of freely grazing cattle and its eff ect on pasture fertility. I. Excretal distribution. Agron. Journal. 48:440-444. Peterson, P.R. and J.R. Gerrish. 1996. Grazi ng systems and spatial distribution of nutrients in pastures: Livestock ma nagement considerations. p. 203-212. Nutrient cycling in Forage Systems. 7-8 March 1996. Pineiro, Gervasio, Jos Maria Paruelo and Ma rtin Oesterheld. 2006. Potential long-term impacts of livestock introduction on carbon a nd nitrogen cycling in grasslands of Southern South America. Global Change Biology. 12:1267-1284. Powell, J.M., F.N. Ikpe, and Z.C. Somda. 1999. Crop yield and the fate of nitrogen and phosphorus following application of plant ma terial and feces to soil. Nutrient Cycling in Agroecosystems. 54:215-226. Powell, J.M., F.N. Ikpe, Z.C. Somda, and S. Fernandez-Rivera. 1998. Urine effects on soil chemical properties and the impact of urine and dung on pearl millet yield. Expl. Agric. 34:259-276. Powell, J.M., M.A. Wattiaux, G.A. Broderick, V.R. Moreira, and M.D. Casler. 2006. Dairy diet impacts on fecal chemical prope rties and nitrogen cycling in soils. Soil Sci. Soc. Am. J. 70:786-794. Rhoads, F.M., R.L. Stanley, Jr., and E.A. Hanl on. 1997. Response to bahiagrass to N, P, K on an Ultisol in North Florida. Soil Crop Sci. Florida Proc. 56:79-83. Robinson, Donald L. 1996. Fertilization and nutrient utilization in harvested forage systems Southern forage crops. p. 65-92. In R.E. Joost and C.A. Roberts (ed.) Nutrient cycling in forage systems. Potash and Phosphate Institute and The Foundation for Agronomic Research, Manhattan, Kansas. Rockmann, T., J. Kaiser and C.A.M Brennikme ijer. 2003. The isotopic fingerprint of the pre-industrial and anthropogenic N2O source. Atmos. Chem. Phys. 3:315-323. Rotz, C.A., F. Taube, M.P. Russelle, J. Oe nema, M.A. Sanderson, and M. Wachendorf. 2005. Whole-farm perspectives of nutri ent flows in grassl and agriculture. Crop Sci. 45:2139-2159.

PAGE 134

134 Russelle, Michael P. 1996. Nitrogen cycling in pasture systems. In Nutrient cycling in forage systems. p. 125-166. In R.E. Joost and C.A. Roberts (e d.) Nutrient cycling in forage systems. Potash and Phosphate Institute and The Foundation for Agronomic Research, Manhattan, Kansas. Russelle, M.P., B.A. Browne, N.B. Turyk, a nd B. Pearson. 2005. Den itrification under pastures on permeable soils helps protect ground water quality. p. 98. In S.C. Jarvis et al. (ed.), Optimisation of nutrient cycling and soil quality for sustainable grasslands. Wageningen Academic Publishers Wageningen, The Netherlands. Saarijarvi, K., P. Virkajarvi H. Heinonen-Tanski, and I. Taipalinen. 2004. N and P leaching and microbial contamination from intensively managed pasture and cut sward on sandy soil in Finland. Agriculture, Ecosystems, and Environment. 104: 621-630. Sakar, D., and G.A. OConnor. 2004. Plant and soil responses to biosolidsphosphorus in two Florida soils with high phosphorus content. Communications in Soil Science and Plant Analysis. 35:1569-1589. SAS Institute. 2007. SAS/STAT users guid e. Release version 9.1. SAS Inst., Cary, NC. Saunders, W.M.H. 1984. Mineral composition of soil and pasture from areas of grazed paddocks, affected and unaffected by dung and urine. New Zealand Journal of Agricultural Re search. 27:405-412. Schachtman, Daniel P., Robert J. Reid a nd S.M. Ayling. 1998. Phosphorus uptake by plants: From soil to cell. Plant Physiol. 116:447-453. Scott, J.M. 1920. Bahiagrass. University of Florida Agricultural E xperiment Stations, Gainesville, FL. Shigaki, Francirose, Andrew Sharpley and Lu is Ignacio Prochnow. 2006. Animal based agriculture, phosphorus management and wate r quality in Brazil: Options for the future. Sci. Agric. (P iracicaba, Braz.) 63:194-209. Sigua, G.C., M.J. Williams, and S.W. Cole man. 2004. Levels and changes of soil phosphorus in subtropical beef cattle pa stures. Comm. Soil Sci. Plant Anal. 35:975-990. Silva, R.G., K.C. Cameron, H.J. Di, and E.E. Jorgensen. 2005. A lysimeters study to investigate the effect of dairy effluent and urea on cattle urine N losses, plant uptake and soil retention. Water, Air and Soil Pollution. 164:55-78. Skerman, P.J., and F. Riveros. 1989. Tropical grasses. FAO, Rome.

PAGE 135

135 Smith, K.A., and J.P. Frost. 2000. Nitrogen excr etion by farm livestock with respect to land spreading requirements and contro lling nitrogen lo sses to ground and surface waters. I. Cattle a nd sheep. Biosource Tech. 71:173-181. Smith, K.A., P.E. Thompson, H. Clayton, I.P. McTaggart, and F. Connen. 1998. Effects of temperature, water content and nitr ogen fertilisation on emissions of nitrous oxide by soils. Atmospheric Environment 32:3301-3309. Sollenberger, L.E., J.C.B. Dubeux, H.Q. Santos, and B.W. Mathews. 2002. Nutrient cycling in tropical pasture ecosystems. p. 151-179. In A.M.V. Batista et al. (ed.), Proc. Brazilian Soc. Animal Sci., Recife, Brazil. 29 July-1 Aug. Sociedade Brasileira de Zootecnica, Brasilia, Brazil. Steenvoorden, J., H. Fonck, and H.P. Oo sterom. 1986. Losses of nitrogen from intensive grassland systems by l eaching and surface runoff. p. 85-97. In H.G. van der Meer, J.C. Ryden, and G.C. Ennik. (ed.) Nitroge n fluxes in intensive grassland systems. Steinauer, E.M., and S.L. Collins. 1995. Effects of urine deposition on small-scale patch structure in prairie ve getation. Ecology. 76:1195-1205. Stevenson, F.J., and M.A. Cole. 1999. Cy cles of soilcarbon, n itrogen, phosphorus, sulfur, and micronutrients. Wiley Publishers, New York. p. 427. Stewart, R.L., Jr. 2003. Management intensit y effects on animal performance and herbage response in bahiagrass pastures. M.S. thesis University of Florid a. Gainesville, FL. Stewart, R.L., Jr., J.C.B. Dubeux, Jr., L.E. So llenberger, and J.M.B. Vendramini, and S.M. Interrante. 2005. Stocking me thod affects plant responses of Pensacola bahiagrass pastures. Online. Forage and Grazi nglands doi:10.1094/FG-2005-1028-01-RS. Stewart, R.L. Jr., L.E. Sollenberger, J.C.B. Dube ux Jr., J.M.B. Vendramini, S.M. Interrante, and Y.C. Newman. 2007. Herbage and animal responses to management intensity of continuously stocked bahiagrass pastures. Agron. J. 99:107-112. Stout, W.L., S.A. Fales, L.D. Muller, R.R. Schnabel, W.E. Priddy, and G.F. Elwinger. 1997. Nitrate leaching from cattle urine and feces in Northeast USA. Soil Sci. Soc. Am. Journal. 61:1787-1794. Stout, W.L., S.L. Fales, L.D. Muller, R.R Schnabel, and S.R. Weaver. 2000. Water quality implications of nitr ate leaching from intensively grazed pasture swards in the Northeast US. Agric. Ecosyst. Environ. 77:203-210. Sugimoto, Y., M. Hirata, and M. Ueno. 1987. Energy and matter flows in bahiagrass pasture. V. Excreting behavior of Holstein heifer s. J. Japan. Soc. Grassl. Sci. 33:8-14.

PAGE 136

136 Sumner, Sid, Wayne Wade, Jim Selph, Jerry Sout hwell, Vicky Hoge, Pat Hogue, Ed Jennings, and Travis Seawright. 1992. Fertilization of established bahiagrass pasture in Florida. Florida Cooperative Extens ion Service Circular 916. Taliaferro, C.M., F.M. Rouquette, Jr., and P. Misl evy. 2004. Bermudagrass and stargrass. p. 417-438. In L.E. Moser, B.L. Burson and L. E. Sollenberger (ed.) Warm-season (C4) grasses. ASA Monogr. 45. ASA, CSSA, SSSA, Madison, WI. Tayler, J.C. and J.E. Rudman. 1966. The distributi on of herbage at different heights in grazed and dung patch areas of a sward under two me thods of grazing management. J. Agric. Sci. 66:29-39. Thomas, R.J. 1992. The role of the legume in th e nitrogen cycle of productive and sustainable pastures. Grass For. Sci. 47:133-142. Thomas, R.J., and N.M. Asakawa. 1993. Decompos ition of leaf litter from tropical forage grasses and legumes. Soil Biology and Biochemistry 25:1351-1361. Thompson, R.B., and I.R.P Fillery. 1998. Fate of ur ea nitrogen in sheep urine applied to soil at different times of the year in the pasture-whea t rotation in south Western Australia. Aust. J. Agric. Res. 49:495-510. Tiffany, M.E., L.R. McDowell, G.A. OConnor, F.G. Martin, and N.S. Wilkinson. 1999. Variation of forage and extractable so il minerals over two grazing seasons in North Florida. Commun. Soil Sci. Plant Anal. 30:2743-2754. Twidwell, E., W.D. Pitman, and G.J. Cuomo. 1998. Bahiagrass production and management. Louisiana State Univ. Public ation # 2697, Baton Rouge, LA. Valk, H., and M.E.J. Hobbelink. 1992. Supplem entation of grazing dairy cows to reduce environmental pollution. In 14th General Meeting of the European Grassland Federation. Lahti, Finland. Eur opean Grassland Fe deration. p. 400-405. Vallis, I., L.A. Harper, V.R. Catchpoole, and K.L. Weier. 1982. Volatilization of ammonia from urine patches in a subtropical pasture. Aust. J. Agric. Res. 33:97107. van der Meer, H.G., and M.G. Van Uum-Va n Lohuyzen. 1986. The relationship between inputs and outputs of nitrogen in in tensive grassland systems. p. 1-15. In H.G. van der Meer, J.C. Ryden, and G.C. Ennik. (ed.), Nitr ogen fluxes in intensive grassland systems. Martinus Nijhoff, Dordrecht. Wageningen, The Netherlands. van Groenigen, Jan Willem, Peter J. Kuikman, W illy J.M. de Groot, and Gerard L. Velthof. 2005a. Nitrous oxide emission fr om urine-treated soil as influenced by urine composition and soil physical cond itions. Soil Biology & Biochemistry 37:463-473.

PAGE 137

137 van Groenigen, Jan Willem, Gerard L. Velthof Frank J.E. van der Bolt, An Vos and Peter J. Kuikman. 2005b. Season variation in N2O emissions from urine patches: Effects of urine concentration, soil co mpaction and dung. Plant and Soil. 273:1527. Wachendorf, Christine, Friedhelm Taube, and Michael Wachendorf. 2005. Nitrogen leaching from 15N labeled cow urine and dung app lied to grassland on a sandy soil. Nutrient Cycling in Agroecosystems. 73:89-100. West, C.P., A.P. Mallarino, W.F. Wedin, and D.B. Marx. 1989. Spatial variability of soil chemical properties in grazed pastur es. Soil Sci. Soc. Am. J. 53:784-789. White, R.E. 1988. Leaching. p. 193-211. In J.R. Wilson (ed.) Advances in nitrogen cycling in agricultural ecosystems. CAB Inte rnational, Wallingford, CT, USA. Whitehead, D.C. 1970. The role of nitrogen in grassland productivity. Commonwealth Agricultural Bureau, Ab erystwyth, Wales, UK. Whitehead, D.C. 1986. Sources and transformati ons of organic nitrogen in intensively managed grassland soils. p. 47-58. In H.G. van der Meer, J.C. Ryden, and G.C. Ennik. (ed.) Nitrogen fluxes in intensive grassla nd systems. Martinus Nijhoff, Dordrecht. Wageningen, The Netherlands Whitehead, D.C., D.R. Lockyer, N. Raistrick. 1989. Volatilization of ammonia from urea applied to soil: influence of hippuric acid and ot her constituents of livestock urine. Soil Biology & Biochemistry 21:803-808. Wilkinson, S. R., and R.W. Lowrey. 1973. Cy cling of mineral nutrients in pasture ecosystems. p. 247-315. In G.W. Butler and R.W. Bailey (ed.) Chemistry and biochemistry of herbage. Vol. 2. Academic Press, New York, NY. Williams, C.H. 1950. Studies on soil phosphorus. II. The nature of native and residual phosphorus in some South Australian so ils. J. Agric. Sci., Camb. 40:243-256. Williams, M.J., and A.C. Hammond. 1999. Rotational vs. continuous intensive stocking management of bahiagrass pasture for cows and calves. Agron. J. 91:11-16. Williams, P.H., and R.J. Haynes. 1995. Effect of sheep, deer, and cattle dung on herbage production and soil nutrient content. Grass and Forage Science. 50:263-271. Williams, P.H., and R.J. Haynes. 2000. Transfor mations and plant uptake of urine N and S in long and short-term pastures. Nutr ient Cycling in Agroecosystems. 56:109116.

PAGE 138

138 Woodard, K.R., E.C. French, L.A. Sweat, D. A. Graetz, L.E. Sollenberger, B. Macoon, K.M. Portier, B.L. Wade, S.J. Rymph, G.M. Prine, and H.H. Van Horn. 2002. Nitrogen removal and nitrate-N leachi ng for forage systems receiving dairy effluent. J. Environ. Qual. 31:1980-1992. Woodard, Kenneth R., Edwin C. French, Lewi n A. Sweat, Donald A. Graetz, Lynn E. Sollenberger, Bisoondat Macoon, Kenneth M. Portier, Stuart J. Rymph, Brett L. Wade, Gordon M. Prine, and Harold H. Van Horn. 2003. Nitrogen removal and nitrate leaching for two perennial, sod-based forage systems receiving dairy effluent. J. Environ. Qual. 32:996-1007. Wrage, N., G.L. Velthof, M.L. van Beusichem, and O. Oenema. 2004. Role of nitrifier denitrification in the production of nitr ous oxide. Soil Biol. Biochem. 33:1723-1732. Yokoyama, Kazuhira, Hideaki Kai, and Hirofu mi Tsuchiyama. 1991. Paracoprid dung Beetles and gaseous loss of nitrogen fr om cow dung. Soil Biol. Biochem. 23: 643-647.

PAGE 139

139 BIOGRAPHICAL SKETCH Una Renee White was born in Low Moor, Virginia, on 21 August 1982 and grew up in Millboro, Virginia. She was raised on a beef fa rm, where she tagged along with her dad feeding cattle. She moved away in 2003 to Raleigh, No rth Carolina, to atte nd North Carolina State University, where she received her B.S. in Ag ronomy. Upon completing her M.S. she would like to work for the government and then star t her own consulting company. One day she hopes to move back to Virginia and have a small farm.