The objective of this dissertation research was to fully assess the impact of narasin on nitrogen dynamics in an agricultural soil. Agricultural soils are exposed to narasin, an anti-coccidiodal agent, when poultry litter is applied as a nitrogen fertilizer. Though it has a relatively short half-life in soil, narasin may persist at concentrations in the pg·kg-1 to ng·kg-1 range. Because sustainable fertilizer practices are based on known parameters of soil nitrogen cycle variation, microbial inhibition or delayed activity caused by antibiotics may undermine the ability of modeling tools to make strong fertilizer management recommendations, leading to reduced fertilizer use efficiency and increased losses of pollutant N species, including N2O and NO3–. Three soil incubation experiments conducted at 40%, 60%, and 80% WFPS were conducted to evaluate the impacts of 1-1000 ng kg-1 narasin on nitrogen transformation rates, changes to nutrient plant availability, and the accumulation of eco-toxic nitrogen compounds such as NO3– and N2O. The results indicate that even ultralow doses of narasin (1-1000 ng kg-1) can significantly alter one or more steps in the nitrogen cycle, which may lead to nutrient deficiencies in crops and/or an increase in non-point source nitrogen pollution.
Nitrogen fertilizer treatments in excess of 10 Tg N yr-1 are necessary to meet increasing demand for global food supplies, but are concomitant with a number of environmental concerns. For example, there is little debate that fertilizers are the primary source of nitrate nitrogen in groundwater, streams, and coastal aquatic ecosystems. In addition to being linked to a number of human health risks including methemoglobinemia (blue-baby syndrome), colon cancer, and reproductive disorders, excess nitrate (NO3–) in groundwater and streams also degrades aquatic ecosystems. Widespread eutrophication in rivers, lakes, estuaries, and coastal areas has been positively correlated to the increasing availability of reactive nitrogen from agricultural fertilizer use, leading to biodiversity loss, fishery collapse, and increased production of paralytic shellfish toxins. Increasing nitrate loads in estuarine environments have also been demonstrated to reduce coastal buffering that protects adjacent oceans from eutrophication. Nitrogen fertilizer use in agriculture is also the leading contributor to global nitrous oxide (N2O) emissions. N2O is a powerful greenhouse gas and the leading modern contributor to stratospheric ozone depletion. Since pre-industrial times, atmospheric N2O has increased from 270 parts per billion (ppb) to over 320 ppb, with the majority of this increase attributed to use of nitrogen fertilizers in agriculture.
Although microbial denitrification (NO3– →N2O→N2) can significantly mitigate nitrate losses, it is also the dominant source of N2O from soils. Alternate N2O production pathways (e.g. NH2OH→N2O or NO2–→NO→N2O) are also mediated by soil microbial activity. The many factors controlling microbial N-cycling are well-characterized and the resulting relationships have been used to develop a number of different modeling tools to better manage nitrogen fertilizer use. NLEAP-GIS and ADAPT-N are two such models, each used to estimate nitrogen fertilizer application rates that will maximize crop yield and minimize non-point source nitrogen pollution. Currently, however, these models do not take into account potential shifts in the nitrogen cycle that may arise when soil microbes are exposed to antibiotics.
The primary pathway by which antibiotics enter agricultural soils includes the use of animal manure as a fertilizer. Up to 90% of antibiotics administered to livestock are excreted as active, non-metabolized compounds. When manure is then spread as a nitrogen fertilizer supplement, residual antibiotic compounds are introduced to crop soils where they often persist and remain bioavailable. A number of recent studies have shown that commonly administered veterinary antibiotics can modify soil microbial community structure or function, even at environmentally relevant concentrations. However, a comprehensive assessment of the impact that antibiotics may have on soil nitrogen dynamics has not been put forth. Even a small shift in the microbial community structure, a delay to the onset of microbial activity, or a change in nitrogen transformation rates may considerably affect fertilizer use efficiency and have direct consequences to both crop yields and non-point source N2O or NO3– pollution. The overall objective of this research is to fully assess the impact of narasin on soil nitrogen transformations using incubation studies and isotopic tracers.
The first objective, quantifying the impact of Narasin on soil nitrogen transformation rates, was to be achieved by conducting a comprehensive series of soil incubation tests designed to compare the rate of mineralization, nitrification, denitrification, N2O emissions, and accumulation of NO3– in soils that were not exposed to antibiotic and to those receiving 1, 10, 100, or 1000 µg kg-1 narasin, an anticoccidiodal agent administered to poultry. Incubations were performed at 40%, 60%, and 80% water-filled pore space (WFPS). The resulting mineral N content (NO3– and NH4+) was quantified colorimetrically, the rates of mineralization, nitrification, and denitrification were quantified using the paired isotope methoed, and N2O flux was determined via isotope ration mass spectrometry.
The second objective is to develop and test a dose-response model for the rate of mineralization, nitrification, denitrification, N2O emission, and potential NO3– leaching as a function of soil moisture and narasin concentration. This objective was scheduled for completion by May 31, 2015 but was delayed due to unexpectedly long turnaround times for acqisition of isotope data. Published turnaround times are 6-8 weeks and though all samples were submitted by January 2015, the final data reports were not received until late April 2015. Since this completion of this second objective is dependent upon the successful completion of Objective 1 (completed May 31, 2015) the scheduled deadline could not be met. Progress is currently underway an a related field experiment is tentatively scheduled for October 2015.
The experimental portion of this proposed research was conducted in two phases. In the first phase, soil was randomly sampled from the top 10 cm of the field site, sieved to <2mm, frozen, and returned to the laboratory for a comprehensive set of soil incubation tests. During the second phase a dose-response model will be developed from regression analysis and tested at the field site.
The field site, Bulls Eye Farm, was selected for three reasons. First, manure fertilizers have not been applied to this farm in the last 15 years and the soils (Ultisol) are representative of the region where intense poultry production encourages the application of litter as a fertilizer amendment. Second, the owner of the land, Pete Okie, has granted essentially unlimited access to the property and has agreed to allow collection of soil samples and to conduct field testing. Finally, the independent farmer working this field has been extremely forthcoming with providing fertilization and crop history for the site.
Narasin has been selected for this study for three reasons: (1) narasin is commonly administered to poultry, (2) poultry litter is commonly applied as a fertilizer amendment, which can introduce narasin to the soil in concentrations as high as 615 µg kg-1, (3) narasin has limited mobility in soils and will tend to remain in the root zone where changes to nitrogen transformation rates have the most significant impact on plant nutrient uptake and N2O that is produced as a by-product is most readily lost to the atmosphere.
Objective 1: Quantifying the rate of mineralization, nitrification, denitrification, N2O emissions, and accumulated NO3–
The first objective, quantifying the impact of narasin on soil nitrogen transformation rates, was achieved by conducting a comprehensive series of soil incubation tests designed to simultaneously quantify the rate of mineralization, nitrification, denitrification, N2O emissions, and accumulation of NH4+ and NO3–. At the start of each incubation experiment, eighteen 50 cm3 plastic cups were filled with 75 g air-dried, sieved soil and treated with 1 mL of inorganic fertilizer substrate containing 7 mg NH4+-N and 3.5 mg NO3—N ( 72 kg N ha-1). Half the samples received substrate enriched with 10% atom excess 15N-(NH4)2SO4 and the remaining half were enriched with 10% atom excess 15N-KNO3. The soils were then treated with a 1 mL narasin solution resulting in dosages of 0 (control), 1, 10, 100, 1000 ng kg-1. The total soil moisture content was raised to 40%, 60%, or 80% WFPS using Milli-Q and then the cups containing each soil sample were placed in the bottom of a 500 mL glass jar and capped with a stainless steel lid outfitted with two gas-tight sampling ports and three-way stopcocks. Once sealed, one the jars were incubated in the dark at 23°C. After incubating for 24 hours, headspace samples were collected from three jars enriched with 15N-KNO3 and three jars enriched with 15N-(NH4)2SO4 using a 25-mL sample lock syringe attached to one of the sampling ports. Immediately after collection, 25 mL of the sample will be flushed through a pre-evacuated Exetainer vial, with the final 12 mL retained for 15N-N2 and N2O analysis. The gravimetric water content of each soil was measured, and an approximately 5 g sample was extracted with 40 mL of 2M KCl. The concentration of NH4+ and NO3– in the soil extracts was determined colorimetrically, and then the extracts were diffused onto acidified glass fiber filters and the filters submitted to determine the 15N enrichment of the mineral nitrogen pools. The remaining samples were sampled and extracted at 48 and 72 hours, respectively. All isotopic enrichments (N2, NH4+, and NO3–) were quantified by Isotope Ratio Mass Spectrometry (IRMS) at the University of California, Davis. The rates of mineralization, nitrification, and denitrification corresponding to each treatment were determined using the paired isotope technique.
Figure 1 illustrates the size of the NH4+ nitrogen pool (mg NH4+-N kg-1) under the three moisture regimes as a function of dose and time. Based on the one-way ANOVA, a significant dose-response is observed at 40% WFPS at each sampling period (see Table 1). For the first two days following narasin exposure, all but the 10 ng kg-1 dose correspond to an increase in the median availability of NH4+-N. The magnitude of this effect starts to diminish by Day 2 and is essentially absent three days after treatment. NH4+-N availability in soils treated with 10 ng kg-1 narasin appears to follow an inverse trend in which the initial pool sizes is comparable to the control but is nearly 25% smaller by the end of the experiment. No significant affects where observed for soils at 60% WFPS or 80% WFPS.
The NO3–-N pool was significantly affected by narasin exposure at all three sampling points for each of the moisture levels tested (Table 2). At 40% WFPS (Figure 2, left panel), a bi-modal response is observed on Day 1 in which the NO3–-N pool is larger relative to the control at the lowest antibiotic dose (1 ng kg-1) and smaller relative to the control at the remaining doses. As time proceeds, little change in the net pool size is observed at the higher doses and by Day 3, the median NO3–-N pool in the control soil exceeds that of all four sets of narasin-treated soils. At 60% WFPS (Figure 2, center panel) and 80% WFPS (Figure 2, right panel), the dose-responses are generally linear but opposite in effect. At 60% WFPS, the NO3–-N pool size is inversely related to narasin dose, an effect that increases in magnitude over time. The opposite effect is observed at 80% WFPS where the trend is also magnified over the course of time, but the availability of NO3–-N increases as a function of dose.
Exposure to narasin appears to have little impact on the rate of nitrogen mineralization. Under more aerobic conditions (40% WFPS), a statistically significant dose response (p = 0.03) was observed on Day 2 of the experiment where mineralization appears to be inhibited at the 1000 ng kg-1 dose but otherwise remains relatively constant over time. At 60% WFPS, it appears as if higher doses may correlate to a reduced mineralization rate (Figure 3, right panel), but one-way ANOVA shows no significant dose response (Table 3).
The rate of nitrification was significantly affected by narasin exposure at both 40% and 60% WFPS (Table 5). Under low-moisture conditions (Figure 5, left panel), the effects are apparent on Day 1, where nitrification is inhibited relative to the control at all four doses, most strongly at the 10 ng kg-1 dose (p = 0.0003). As time progresses, the dose-response remains inhibited at all doses on Day 2 (p = 0.00004). By Day 3, the dose-response curve shows an inversion where nitrification at the 10 ng kg-1 which initially corresponded to the greatest inhibition exhibits a rebound and exceeds that of the control at this sampling period while inhibition is still observed at the remaining doses (p = 0.00002). At 60% WFPS a statistically significant dose-response relationship emerges on Day 2 and persists through Day 3 (Table 5). On Day 2 (Figure 5, right panel, p=0.03), nitrification is accelerated relative to the control at the lowest dose (1 ng kg-1), no significant difference is observed at the two middle doses (10 and 100 ng kg-1), and inhibition is observed at the highest dose (1000 ng kg-1). The increased rate of nitrification at the low dose on Day 2 appears temporal, however, as the rate at this and the 10 ng kg-1 dose are comparable to the control by Day 3. Apparent inhibition at the highest dose is maintained from Day 2 to Day 3, at which time it is also apparent at the 100 ng kg-1 dose.
The rate of denitrification was nearly 2 orders of magnitude greater at 60% WFPS than 40% WFPS, which is consistent with the fact that denitrification is favored by anaerobic conditions. While denitrification was not significantly affected by narasin treatment on Day 1 at either moisture regime (Table 5), denitrification in the low-moisture soil (Figure 5, left panel) at the 1 ng kg-1 dose diminished considerably less between Days 1 and 2 than did the control. Denitrification remains high at this dose through Day 3, whereas the rate at higher doses (10 and 100 ng kg-1) is reduced, suggesting overall inhibition of denitrification at these doses. The dose-response at 60% WFPS (Figure 5, right panel) undergoes at temporal inversion in which denitrification rate is accelerated relative to the control at all four doses on Day 1 but appears to slow rapidly, resulting in an overall inhibition of denitrification rate at all four doses on Day 2 (p =0.00005) and Day 3 (p = 0.00008).
N2O production was greatest at 60% WFPS, where daily flux ranged from 5-10 μg N2O-N kg-1 d-1 (Figure 6, middle panel). Flux was slightly lower at 80% WFPS (Figure 6, right panel), where itdid not exceed 3.5 5-10 μg N2O-N kg-1 d-1, and at 40% WFPS, where conditions are more favorable to anaerobic activity, average N2O flux was less than 200 ng N2O-N kg-1 d-1 (Figure 6, left panel). These results are consistent with present knowledge of the influence of moisture on N2O production rates. Furthermore, significant dose-responses to narasin treatment were observed under all three moisture regimes (Table 6). At lower moisture (40% WFPS), the impact is most clearly observed on Day 3 (p = 0.0003) where a direct relationship between dose and N2O is observed and the median flux rate at 1000 ng kg-1 is more than 3x that of the control. At 60% WFPS, where maximum N2O production occurs, the same linear trend is observable on Day 1 (p = 0.005) and maintained through Day 2 (p = 0.001) and Day 3 (p = 0.007). The magnitude of this effect peaks on Day 1 where the median N2O flux is nearly 4x that of the control (Day 1) at the 1000 ng kg-1 dose but remains significant (2x the control) by the final sampling period on Day 3. The dose response at 80% WFPS follows a different trend (Figure 6, right panel) but remains statistically significant for all three days (Table 6). Here an inversion similar to the one observed for denitrification rate (60% WFPS) occurs. Maximum N2O flux corresponds to the 1 ng kg-1 dose on Day 1 and, at this dose, N2O flux remains elevated relative to the control for the duration of the experiment. At the three higher doses, N2O flux is also elevated on Day 1, but slows considerably by Day 2 and by Day 3 N2O production at these doses is less than half that of the control.
- Figure 1. Box-whisker plots illustrating the NH4+-N pool over a three-day incubation period in moist soils (40%, 60%, and 80% WFPS) treated with narasin.
- Figure 4. Time-series plots illustrating changes in nitrification rates (mg N kg-1 d-1) as a function of narasin dose.
- Figure 2. Box-whisker plots illustrating the NO3–N pool over a three-day incubation period in moist soils (40%, 60%, and 80% WFPS) treated with narasin.
- Figure 3. Time-series plots illustrating changes in mineralization rates (mg N kg-1 d-1) as a function of narasin dose.
- Figure 6. Box-whisker plots illustrating the N2O-N flux over a three-day incubation period in moist soils (40%, 60%, and 80% WFPS) treated with narasin.
- Figure 5. Time-series plots illustrating changes in denitrificaton rates (mg N kg-1 d-1) as a function of narasin dose.
- Tables 1-6. Results of one-way ANOVA.
The evidence presented here shows that soils exposed to trace levels of the anticoccidiodal drug narasin exhibit dose and time-dependent shifts in nitrogen cycling rates that may affect plant nutrition and the production/loss of ecotoxic forms of nitrogen (NO3– and N2O) to the environment. Effects on mineralization, nitrification, and thus NH4+-N concentration are most considerable for low-moisture soils (40% WFPS). Although mineralization, which adds to the NH4+-N pool is slightly inhibited by narasin at most of the doses tested, the rate of mineralization is an order of magnitude less than that of nitrification, a process that depletes the NH4+-N pool. Since nitrification also exhibits significant dose-dependent inhibition, the net result is an increase in the size of the NH4+-N pool relative to the control, as seen in Figure 1. For crops that utilize NH4+ as their preferred N-source, the effects of narasin may therefore be beneficial. On the other hand, if the size of the resulting NH4+-N pool is larger than can be utilized, potential runoff losses of NH4+ may rise, thus reducing plant-available NH4+ and increasing non-point source nitrogen pollution. The potential effect of narasin on nutrient availability for crops that favor NO3– as an N-source is equally complex. Denitrification affects the size of the NO3–-N pool at several narasin doses and times, particularly in the first 24 hours after exposure. At 40% WFPS where leaching losses of NO3– are relatively small, nitrification is inhibited at most doses so less NO3– is added to the soil. Concurrently, denitrification is accelerated at several doses for 2 days. At the lowest dose (1 ng kg-1), the acceleration continues through Day 3. Enhanced removal NO3– combined with slower nitrification results in a smaller NO3–-N pool at most of the doses tested (Figure 2). Similar effects reduce the NO3–-N pool at 60% WFPS. At this moisture level, additional losses due to leaching are more probable, particularly following a rain event. Where soils are close to saturation (80% WFPS), the combined effects of narasin result in a slight increase in NO3–. Though possibly beneficial from the perspective of plant nutrition, leaching is a significant risk for soils near saturation and a rain event would obviate that benefit and increase NO3– input to groundwater.
In addition to potential NH4+ and NO3– losses that may result from narasin exposure, production and losses of N2O, a strong greenhouse gas and ozone-reducing molecule was observed to increase at all four doses in low (40% WFPS) and moderate (60% WFPS) conditions. At the field scale, a 2-fold increase in N2O flux is both significant and of great concern, particularly because agriculture is already the primary contributor to increasing atmospheric N2O. Since NO3– was shown to decrease as a result of narasin exposure under several moisture conditions and additional losses to leaching are possible, increased N-additions may be required to compensate for N-deficiencies resulting from this shift. Increased N-additions may subsequently contribute to even higher N2O flux from affected soils, an outcome that is decidedly detrimental at both local and global scales.
Education & Outreach Activities and Participation Summary
DeVries, S.L.; Loving, M.; Li, X.; Zhang, P., The effect of ultralow-dose antibiotics exposure on soil nitrate and N2O flux. Submitted manuscript on 7/28/2015.
DeVries, S.L.; Zhang, P.; Loving, M.; Logozzo, L., The Effects of Ultra Low Narasin Exposure on Soil Nitrogen Processing. Presented at the Jeffrey Steiner Memorial Symposium, May 2015, New York, NY.
DeVries, S.L.; Loving, M.; Zhang, P., Stimulated Nitrate Reduction in Anaerobic Soil Exposed to Trace Veterinary Antibiotic. Presented at the American Geophysical Union Fall 2013 Meeting, San Francisco, CA.
DeVries, S.L.; Loving, M.; Li, X.; Zhang, P., The effect of ultralow-dose antibiotics exposure on soil nitrate and N2O flux. Presented at the Northeastern Ecosystem Research Cooperative Spring 2015 Meeting,Saratoga Springs, NY.
DeVries, S.L. Soil on (a little bit of) Drugs: Evidence against linear dose-response models and the implications for sustainability in agroecysostems. Presented at Rutgers University November 2013, Brusnwick, NJ.
Areas needing additional study
The present study was intended as a preliminary assessment of the effects of the anti-coccidiodal drug narasin on biogeochemical nitrogen cycling. The results indicate that even low-dose exposure to this antimicrobial may affect agricultural sustainability both from the perspective of plant nutrition and environmental health. In addition to replicating this work to test the effects of narasin on soil nitrogen in different types of soil, additional research is needed to: (1) examine the effects of additional antibiotic compounds that may be introduced to agriculture soils. Examples include oxytetracycline, chlorotetracycline, virginiamycin, gentamicin, bacitracin, sulfadiazine, and sulfamethoxazole, (2) determine whether the affects reported here persist or change following repeat applications of a given antibiotic compound. Though antibiotic resistance may develop, it is not clear how that may affect the functional composition of the soil microbial community, (3) estimate the field-scale impacts of antibiotics on soil nitrogen and non-point source nitrogen pollution using a a combination of field and modeling studies. In particular, a precision nutrient management tool such as ADAPT-N should be utilized alongside a field study to determine whether biogeochemical shifts resulting from antibiotic exposure significantly reduces the efficacy of this or similar tools. Outside of the scope of sustainability, additional studies are recommended to investigate: (1) structural and functional changes to the soil microbial community resulting from chronic, low-dose antibiotic exposure and (2) shifts in the N2O pathway resulting from chronic, low-dose antibiotic exposure.