Biogenic Phosphate Minerals in Manure - ACS Publications

1525 Observatory Drive, Madison, Wisconsin 53706. Phosphorus (P) is present ... ruminant livestock (cattle), and hence, their manures should differ ch...
8 downloads 0 Views 323KB Size
Environ. Sci. Technol. 2002, 36, 5075-5082

Biogenic Phosphate Minerals in Manure: Implications for Phosphorus Loss to Surface Waters LESLIE R. COOPERBAND* AND LAURA WARD GOOD Department of Soil Science, University of Wisconsin, 1525 Observatory Drive, Madison, Wisconsin 53706

Phosphorus (P) is present in waterways throughout the United States at concentrations that impair water quality. Agriculture, particularly livestock production, has been identified as a major cause of this impairment. Excess manure P applied to croplands has increased P losses in runoff, leading to surface water eutrophication. We conducted a long-term (36-week) incubation with poultry and dairy manures applied to a silt loam soil to elucidate mechanisms controlling manure P loss to water. Manures were applied to supply the same total P rate to soils with different antecedent plant-available P concentrations (soil test P). There was a strong synergistic effect between dairy manure and soil test P on water extractable P, while soil test P did not affect P loss from poultry manureamended soils. Using scanning electron microscopy and energy dispersive X-ray spectroscopy, we found that poultry manure contained sparingly soluble calcium and magnesium phosphate minerals that controlled soil solution P concentrations, while dairy manure did not. These minerals resemble other biogenic phosphate minerals. Our findings refute current assumptions that all manure P behaves similarly in soils and that organic forms control manure-soil P loss to water.

Introduction Nonpoint source nutrient pollution and eutrophication of surface waters has been one of the most recalcitrant environmental issues in the U.S. While the Clean Water Act was promulgated in 1972, federal and state governments and policies have yet to address nutrient pollution from nonpoint sources, including agriculture. In a U.S. EPA report to Congress, 36% of all national river miles cited were impaired for designated uses; agriculture contributed to over 60% of these impairments (1). In the U.S., livestock producers traditionally spread manures on croplands as fertilizers, particularly as a source of nitrogen (N). Intensively managed livestock production systems have exacerbated conditions where manure use in crop production is more akin to waste disposal than beneficial fertilization. Manure application to meet crop N needs frequently leads to over-application of P because N and P are present in manure in near-equal concentrations, yet crops need approximately six times more N than P. In 1997, about 400 000 Mg of manure P were produced at the farm-level in excess of the amount needed for crop production; this excess is about 65% of the recoverable manure P in the U.S (2). The * Corresponding author phone: (608)265-4654; fax: (608)265-2595; e-mail: [email protected]. 10.1021/es025755f CCC: $22.00 Published on Web 11/02/2002

 2002 American Chemical Society

repeated application of animal manure P to cropland in excess of crop needs has led to elevated concentrations of P in soils and an increase in the amount of P that may be lost as runoff, through subsurface drainage, or leaching to groundwater (3, 4). The term “animal manure” encompasses a diverse range of materials. Monogastric livestock like poultry and swine have different diets and digest feed P differently than ruminant livestock (cattle), and hence, their manures should differ chemically. Monogastric animals often receive mineral P supplements (calcium phosphates) because they cannot digest organic forms of P common in grains, especially phytic acid. Ruminants, particularly dairy cattle, can digest these organic P compounds, but often receive supplemental mineral P because of perceived benefits for reproductive health. Additional sources of variability in manure include bedding materials (straw, sawdust, sand, oat and rice hulls), presence of undigested feed and urine, and storage duration prior to land spreading. Manures from different animal types are known to have different concentrations of total, inorganic and organic P compounds (5). Despite known differences in amounts and types of P compounds in different livestock manures, the current paradigm driving manure management practices and manure application to croplands assumes that P in all manures behaves similarly. The conventional wisdom is that as organic forms of P undergo microbial decomposition, inorganic P (in the form of H2PO4- or HPO42-) becomes available for plant uptake. Consequently, most assumptions about P release from manure to soils are based on organic matter decomposition kinetics (6). Moreover, current guidelines for soil P management assume that manure P and antecedent plant-available P (or “soil test P”) will interact in an additive manner (7). To date, most of the studies of manure P and soil interactions have focused on inferring environmental (runoff) P loss from empirical plant-available P measures with little regard for mechanisms controlling P loss from manureamended soils (8). The prevailing model for P loss from manure-amended soils is that of a threshold phenomenon: once a soil’s P retention capacity has been reached, P will be lost to the environment (3, 9). We found few studies comparing manure P chemistries from livestock with different digestive systems (10 -12) and no studies comparing different livestock species’ manures to explore mechanisms of manure P interactions with soils and runoff water. We began our research with the hypothesis that P forms in ruminant (dairy cow) and monogastric (chicken) manures will differ chemically because ruminant and monogastric livestock receive different diets and process feed differently. In turn, the difference in manure P chemistries should affect the solubility of P when manures are added to soils. Our objective was to evaluate the interactive effects of manure type and antecedent soil test P (STP) on P release to water from manure-amended soils. In addition, we sought to characterize chemical forms of P in manure directly using scanning electron microscopy (SEM) to identify mechanisms controlling manure P loss to water from manure-amended soils.

Experimental Methods Manure Soil Incubation. The basic experimental design for this research involved laboratory incubations of fresh soils (e.g., not dried or ground) mixed with fresh poultry or dairy manure applied at typical rates (i.e., based on N supply for cash grain) for Wisconsin crop production (7 Mg ha-1 for VOL. 36, NO. 23, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5075

TABLE 1. Chemical Characteristics of Animal Manures Used for Incubationa

manure

pH

dry matter (g kg-1)

poultry dairy

6.6 7.7

670 230

g kg-1 dry matter C

N

P

K

Ca

Mg

Fe

Al

WEPb

356 394

33 20

17 6

24 14

23 25

6 11

2 0.2

2 0.8

2.0 1.7

a N ) 2 or 3 for all analyses. Total C and N determined via dry combustion using a CHN Analyzer. Total minerals determined via nitric-perchloric digestion and ICP. b Water extractable P determined using 1:100 dilution shaken for 1 h, filtered through 0.45 µm filter and measured as dissolved reactive P.

TABLE 2. Characteristics of Soils Used for Incubationa soil

pH

CEC (cmolc kg-1)

low P soil high P soil

6.1 6.3

14.6 15.2

mg kg-1 soil

organic matter (g kg-1)

Bray-1 P

total P

total K

total Ca

total Mg

total Fe

total Al

34 33

12 30

317 376

1608 1685

2200 2700

2900 3100

17500 17800

14900 15400

a CEC or Cation Exchange Capacity determined by ammonium acetate extraction buffered at pH ) 7 (n ) 2); organic matter content determined via loss on ignition (n ) 1); total P and cations determined via nitric-perchloric digestion and ICP (n ) 2).

poultry manure and 56 Mg ha-1 for dairy manure, wet weight). We incubated two soils mixed with poultry and dairy manure at a rate of 180 g total P kg-1 soil under controlled conditions for 36 weeks. We used water-extractable P (WEP) from incubated manure-soil mixtures as a measure of relative risk of P loss in runoff over time (13, 14). We collected dairy manure (DM; mixture of feces, urine and straw bedding) from the UW-Madison Dairy Barns where it had been piled for less than 48 h. The poultry manure or litter (PM; mixture of excreta, spilled feed and oat hull bedding) was collected from the floor of two broiler houses near Arcadia, Wisconsin immediately before the houses were cleaned. The litter had been accumulating for ∼42 days prior to clean out. We determined chemical characteristics of both manures including pH, dry matter content, total C and N (using a Carlo Erba CHN Analyzer) and total concentrations of P, K, Ca, Mg, Fe and Al (using nitric-perchloric acid digestion followed by inductively coupled plasma-optical emission spectroscopy, ICP) (Table 1). Poultry manure had three times greater dry matter and total P contents than dairy manure, while TC and total Ca and Mg contents were similar. Poultry manure Fe and Al contents were an order of magnitude greater than those in dairy manure. However, total Fe and Al concentrations in both manures were low compared to those in the study soils (Table 2). Despite large differences in total dry matter P concentrations, there was little difference in water-extractable P between the two manures. Soil samples were collected from the surface 5-cm of two plots in a long-term fertilizer trial at the University of Wisconsin’s Arlington Research Station. The field soil was classified as Plano silt loam (fine, silty, mixed, mesic Typic Argiudoll). With the exception of P and K fertilizer application rates, the plots had been under the same management for more than 40 years. The soil test used to assess plant available P was Bray-1, a dilute strong acid extractant of 0.025 M HCl and 0.03 M NH4F (15). Bray-1 P (STP) concentrations for these soils were 12 mg kg-1 (“low STP”) and 30 mg kg-1 (“high STP”). Despite a more than two-fold difference in Bray-1 P, total soil P contents were relatively similar between the two soils, as were pH, organic matter, total cations, and cation exchange capacity (Table 2). The manure soil incubation was a 2 × 3 factorial experiment (two soil test P levels × (two manures + no amendment control)) with six replications of each treatment. We mixed fresh manure and field moist soil and placed mixtures into 2.65 L plastic containers. The oven-dry equivalent (ODE) weight of soil per container was ap5076

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 23, 2002

proximately 1.7 kg. The incubations were kept at 25 °C, and soil moisture contents were adjusted gravimetrically during the incubation between 270 and 340 g kg-1 dry weight soil. Soil samples were removed from containers after 2, 4, 8, 14, 24 and 36 weeks of incubation to determine water extractable P and other parameters. To avoid changes in P solubility resulting from drying and grinding, all waterextractions used fresh soils. Soil samples were stored at 4 °C for no longer than 14 days prior to extraction. We conducted a 1:10 deionized water (DIW) extraction of soil samples (50 mL DIW: 5 g ODE soil) by shaking for 1 h on a platform shaker, centrifuging at 8500 rpm for 20 min and then filtering through 0.45 µm nylon filters. Filtrates were analyzed colorimetrically using the reduced phosphomolybdate method for dissolved reactive phosphorus (DRP, a measure of H2PO4and HPO42- in solution) (15). Soil scientists commonly consider filtrates passing through 0.45 µm filters to represent the total dissolved concentrations of both inorganic and organic constituents (16). In addition, we measured waterextractable Ca and Mg in the 1:10 DIW extracts at 2 and 36 weeks using ICP. We used the total dissolved cation and DRP concentrations to generate cation ratios and cation:DRP ratios. At the end of the 36-week incubation, we conducted a series of water extractions with different extraction times and shaking energies to simulate the effect of a range of runoff kinetic conditions on P concentrations in solutions from the incubated soils. Dilution ratios and contact times were as follows: 1:0.6 soil:water, 1 h, no shaking (saturated soil solution P), 1:10, 1 h shaking (readily desorbable P), 1:100, 18 h shaking and 1:1000, 18 h shaking (both represent sediment concentration ranges that could be found in runoff) (17). We used the same shaking, centrifugation, filtration and DRP analysis procedures as those described above. We also conducted 1:10 soil:solution extractions with 1 mM CaCl2 and 1 mM MgCl2 to observe Ca and Mg effects on WEP concentrations. We used the same amount of soil and solution (5 g ODE with 50 mL solution) and extracted under the same set of conditions as described previously for the 1:10 DIW extractions. Both cations and P were analyzed in filtrates using ICP. We measured total carbon (TC) added with the manures and total C of the soil-manure mixtures at the beginning and end of the incubation to evaluate organic matter decomposition effects on WEP. We determined TC from dried and ground samples (