Thije Teunis
30-05-2021, Almere
Supervisor: Dr. Boris Jansen
Second supervisor: Dr. John Parsons
Mentor: Donya Danesh
Spatial N and P distribution and its
causes in a potato field in Flevoland
Abstract
Manure in combination with mineral fertilizers are the dominant form of fertilizers in the Netherlands. While they have been very effective in supplying nitrogen and phosphorus, and have thereby helped with the high crop output in the Netherlands, they also have their downsides. These include both local and global environmental impact, as well as the limited nature of mineral phosphorus fertilizers. To decrease the dependency of these fertilizers in the EU, the LEX4BIO project was set up. This project aims to decrease the dependency by looking at fertilizers based on nutrient rich side streams, and to improve the efficiency of fertilizer use. To do this, the spatial distribution of nutrients needs to be taken into account. Fieldwork was done at a potato field in Flevoland near Almere to take a look at the spatial variability of nitrogen and phosphorus there. NH4, NO3, NO2, DON and PO4 concentrations
were determined on 20 different locations in the field to examine the spatial distribution of these nutrients in the field. A high spatial variability was found for all of these nutrients, in each case the lowest determined value was many times lower than the highest found value. Two possible explanations for spatial variability were examined. First of all, the distance to the highway was assessed as a possible predictor for NOx concentrations in the field. However, the simple linear
regression indicated a positive relation, without statistical significance (p = .823). Next to this, the relationship between PO4 and DON was investigated, as a constant relationship across the field could
indicate spatial variability caused by unequal application of manure. The Spearman’s coefficient between these nutrients was statistically significant (p = .012). These findings indicate that efficiency of fertilizer use can be improved by more adequate application of fertilizers.
Introduction
Mineral fertilizers, specifically those that supply nitrogen and phosphorus, have contributed to the rapid increase in crop yields since the 1950s, as fixated nitrogen and phosphorus are the limiting factor for plant growth on most agricultural lands (Robertson & Vitousek, 2009). With a continuously growing world population, fertilization appears to be a necessity to produce enough food to feed this increasing population. However, mineral fertilizers have several unwanted side effects. First of all, they have a severe impact on their environment as they can cause eutrophication in nearby freshwater bodies, increase nutrient levels in nearby nature areas, and affect soil microbial activity and soil pH (Ayoub, 1999; Robertsen & Vitousek, 2009; Geisseler & Scow, 2014). Next to this, they influence the global climate system by increasing soil nitrous oxide emissions (Robertson & Vitousek, 2009). Additionally, the production of mineral nitrogen fertilizer is largely dependent on the energy intensive Haber-Bosch process, which is responsible for around 1% of global energy consumption, with most of this energy being produced from fossil fuels (Smith, 2002). Finally, mineral fertilizers may increase crop yields, but actually have a negative effect on the fertility of the soil itself (Ayoub, 1999). Furthermore, mineral phosphorus fertilizers are limited by the fact that they need to be imported, and that they are produced from a finite resource. While estimates on how long the worldwide reserves could provide for agriculture based on mineral phosphate fertilizers vary widely, it is accepted that current consumption levels would lead to increased prices and almost complete dependency on Morocco for worldwide phosphate production (Cooper et al., 2011; Edixhoven, Gupta & Savenije, 2014). These problems will be amplified when the mineral fertilizer consumption levels of developing countries will start to resemble those of developed countries (Motesharezadeh & al., 2017).
Meanwhile, manure also has its disadvantages. In its traditional unprocessed form it is limited by the fact that it is difficult to transport, due to its high water content (Burton & Turner, 2003). This is problematic since manure producers (livestock based agriculture) are often located away from manure consumers (crop production). This results in surpluses in some areas, which has negative effects on the environment, and shortages in other areas (Jongbloed & Lenis, 1998; Burton & Turner, 2003). Also, the nutrient levels of unprocessed manure are variable and often unknown, making optimal application difficult (Eghball et al., 2002).
Currently, the vast majority of fertilization used in the Netherlands is either unprocessed manure or mineral fertilizer (Velthof et al., 2017). An evaluation on the effect of fertilizer regulations by Velthof et al., (2017) provided the following results: in 2014, the share of nitrogen being provided by mineral fertilizers in the Netherlands was around 35% for all agricultural land, but up to 55% for cropland. This estimated share was the highest for cropland on clay soils, for which nitrogen provided by mineral fertilizers was estimated at 63%, due to the lower availability of manure there. The estimated share of phosphorus provided by mineral fertilizers was much lower, at 8%, sharply declining since 2005 due to stricter regulations and higher prices. However, this share was also higher for cropland on clay soil. Because of these limitations of the traditional fertilizing methods combined with the Netherlands’ current dependency on them, it is necessary to transition to more sustainable forms of fertilization. Promising alternatives are fertilizers based on nutrient rich resource streams (NRRS). Since these fertilizers are circular in nature, they reduce or eliminate some of the negative side effects related to the currently used mineral fertilizers. Besides diminishing the impact fertilizers have on their
environment by closing the nutrient cycle, there is no need for the Haber-Bosch process and phosphate imports. The LEX4BIO project, a cooperative European project with 21 partners from 14 countries, aims to acquire more knowledge on these fertilizers to be able to supply crops of sufficient nitrogen and phosphorus in more sustainable ways.
Next to this, the LEX4BIO project focuses on ways to optimize application of fertilizers, as this is necessary for both traditional fertilizers based on manure and mineral fertilizers, as well as for the fertilizers based on nutrient rich resource streams, to decrease their potential negative effects on the environment. To enable optimal application it is necessary look at the spatial distribution of nutrients. Spatial variability can result in insufficient amounts of nutrients in some areas of a field, while other areas will have excessive amounts (Nolin, Guertin & Wang, 1996). Spatial variability can be caused by several factors. First of all, it can be caused by the variability of nutrient levels of manure and the difficulty in spreading it consistently over an entire field (Burton & Turner, 2003). Mineral fertilizers are more homogenous in content, but spatial variability can still be caused by uneven distribution. Another reason for the spatial variability may be runoff of nutrients via water streams caused by rainfall, which nitrogen is especially susceptible to (Udawatta et al., 2006). Finally, outside factors may play a role as they can also supply nutrients to a field (Khalid et al., 2020; Dragosits et al., 2002). To take a deeper look at the distribution of nitrogen and phosphorus, a potato field in Flevoland near Almere was investigated. In this field, several nutrients of particular interest were measured. First of all, NO3 as the primary form of nitrogen that plants take up. Additionally, NH4 is also directly available
for plant uptake but usually reacts to NO3. Dissolved organic nitrogen was also measured, while not
being directly available for plant uptake, it can also mineralize and form NO3. Finally, phosphate was
measured as this is the only form of phosphorus that is directly available for plant uptake. These nutrients are supplied to the soil by the farmer via manure, which contains NH4, dissolved organic
nitrogen, and PO4,and mineral fertilizer, cadmium ammonium nitrate, which contains equal parts NH4
and NO3
This research is one of five bachelor projects that will be carried out in relation to the LEX4BIO project at the University of Amsterdam, which is one of the partners of LEX4BIO. One of the other projects will examine the spatial distribution of nitrogen and phosphorus on a nearby field of agricultural grassland, as well as the feasibility of changing fertilization regimes to biobased fertilizers there. The results of this study can be compared to find out if the spatial distribution patterns are similar or not. Another two projects will be dedicated to investigating the risk of heavy metal pollution associated with sewage sludge based fertilizers. Finally, one project will examine to what extent several simple, commercially available test kits can be used to assess N, P, K and pH in agricultural soils, and if these test kits can be used by farmers to monitor the nutrient levels in the soil of their fields.
Aim & Objectives
The aim of this research is to answer the following question: What is the spatial distribution of different forms of nitrogen and phosphorus in a potato field in Flevoland, and what are the causes for possible variability?
To be able to answer this question, the following objectives are laid out:
a. Determine the NH4, NO3, NO2, DON and PO4 values on 20 different locations on the field to examine
the spatial distribution of these nutrients in the field.
b. Investigate possible patterns that caused the observed spatial variability. Two types of possible explanations will be examined in this research:
1. Variability caused by outside factors. In this case of this particular field the outside factor that will be investigated is the highway that lies to the west of the field as a source of NOx.
2. Variability caused by the application methods of fertilizers. This was examined by looking at the relation between PO4 and dissolved organic nitrogen at each sample point. These are
both added to the field via manure, which is not completely homogenous in nutrient distribution but relatively so (Kemppainen, 1989). However, the chemical reactions or leaching happen under different circumstances. Therefore if the relation is fairly constant, it could indicate spatial variability is caused by unequal application of manure across the field.
Methods
Figure 1. Map of the fieldwork site including the numbered sample points, the highway to the west and the farm directly north of the fieldwork area. The map in the top right corner provides the location in relation to Almere and Amsterdam.
Fieldwork site
The fieldwork was conducted at a farm near Almere which currently uses a combination of manure and mineral nitrogen fertilizers. The field, including the farm and the nearby highway, is shown in figure 1. During the time of the fieldwork, the field was used for potato production. However, in the past flower bulbs were cultivated in this field. Remnants of this were still found during the fieldwork. The field is 370 meters by 150 meters, with a cut-out in the north-west corner of the field in the resulting in a surface area of 5.1 hectares. The local climate is oceanic, and the soil type at the farm is marine clay. As mentioned in the introduction, crop agriculture on clay soils consumes the highest share of mineral fertilizers compared to total fertilizers in the Netherlands (Velthof et al, 2017). The fieldwork was done in April. Research by Van Merveinne and Hofman (1989) has shown that the spatial variability of nitrate in a polder field was the smallest during spring, specifically in April. The fact that the fieldwork was conduct in April has as a benefit that the mean amount of nitrate is easier to determine with a limited amount of samples. However, it also means that some of the spatial variability that may present in other months may not have have been observable during this research.
Sampling design & analysis
With ArcGIS, a map was created with 20 random locations. This map was then loaded into the ArcGIS Collector app to be able to trace these locations in the potato field. This randomized design was chosen to account for spatial variability within the field. Similarly, the quantity of sample locations was decided to be 20 as this is cited as a minimum to account for the spatial variability of nitrate concentrations within a field (Brandenberger & al., 2016; Zhang & Johnson, n.d.). From these locations soil samples of 250g were taken at a depth of around 30 cm. However, since the land is ploughed, the concentrations should be fairly homogenous throughout the ploughed layer.
From each soil sample, we extracted 30 grams twice to create analytical duplicates. To these samples, we added 75 mL of water. For 5 of the soil samples, we mixed 20 grams of sample with 50 mL of water. However, since we ran into shortages we increased this to 30 grams and 75 mL of water. This mixture was then shaken in the shaker for two hours. The next day, the samples were put in the centrifuge for 20 minutes. Then, the samples went through a filter with a pore size of 0.45 μm. Of the filtrate, 10 mL was used for analysing on NH4, NO3, NO2, PO4 and total N concentrations with an auto analyser. The
auto-analyser used is the Skalar San++ continuous flow analyzer. The dissolved organic nitrogen content was calculated by subtracting the NH4, NO3 and NO2 from the total N concentration.
Statistical analysis
After we received the data the outliers were identified for the NO3, NO2, dissolved organic nitrogen
and PO4 values. To do this, the values were omitted where one of the duplicates was 1.5 times the
interquartile range below the first quartile or 1.5 times the interquartile range above the third quartile, and the other duplicate was not. In two instances, that of the PO4 value of sample point 14 and the
NO2 value of sample point 18, both of the duplicates were 1.5 times the interquartile range above the
third quartile. Here, the values were not classified as outliers. Since both these. Then, these values were recalculated to μmol per kilogram of dry soil. Finally, the averages were calculated of each value per sample point. If one of the duplicates was classified as an outlier, the value of the other duplicate was used for that sample point, and if one or more duplicates was below the limit of detection without being an outlier it was classified as 0. Then, the mean and the standard deviation for each nutrient was determined using Excel.
To test the relation between the distance to the highway and the NOx concentrations in the soil, a
simple linear regression was constructed with the distance to the highway as the predictor, and the combined NO2 and NO3 concentrations to represent NOx as the dependent variable. Even though the
not all the assumptions that come with a simple linear regressions were met, it is still robust enough to present reasonably reliable results. Additionally, a simple linear regression was set up for just the NO3 concentration as the outcome variable against the distance to the highway as the predictor
variable, as that is the form of NOx that is directly available for plant uptake. Then, the p-value of these
coefficients was determined using R to examine the statistical significance.
To test the relation between the PO4 concentrations and the dissolved organic nitrogen concentrations
per sample point, the Spearman's rank correlation coefficient was set up. This correlation test was chosen instead of the Pearson correlation coefficient as the data was not normally distributed, which is one of the assumptions for the Pearson correlation coefficient. The p-value was again determined using R.
Results
To give an initial visualization tool for the spatial distribution of the nutrients, figure 2 and table 1 were created. Table 1 shows the mean values and the standard deviation of each of the nutrients across the field. The appendix includes a table that gives the average value of each sample point for the researched nutrients. As can be seen in table 1, NO3 is the primary form of nitrogen providing around
90% of the total nitrogen in the field. Dissolved organic nitrogen is the second most represented form of nitrogen, while NO2 and NH4 form very minimal components. NO2 and NH4 also have the highest
standard deviations compared to their mean, which is the result of many sample points having very low values. The standard deviations of NO3, dissolved organic nitrogen and PO4 are all relatively similar
in relation to their respective means.
Table 1. Mean and standard deviation of the researched nutrients. NO3 (μmol/kg) NO2 (μmol/kg) DON (μmol/kg) NH4 (μmol/kg) PO4 (μmol/kg) μ 2706.39 4.66 268.06 2.29 19.62 σ 1051.10 111.56 112.89 4.09 11.23
Figure 2 shows the distribution of the most important form of nitrogen (NO3) and phosphate (PO4) by
using a map with a colour ramp. The colour ramp shows that for both NO3 and PO4 the highest
concentration found at a sample point is several times higher than the lowest concentration found at a sample point. Additionally, the figure 2 shows that the distribution of NO3 and PO4 sometimes shows
similarities, but is not the same. For example, sample point 4 had a low NO3 concentration and a low
PO4 concentration, while sample point 5 had the highest NO3 concentration but a below average PO4
concentration.
Figure 2. Visualization of the spatial distribution of NO3 (left) and PO4 (right) across the potato field.
Figure 3 plots the distance to the highway of each sample point against the NOx and NO3 values of that
sample point, to investigate the influence of the highway on NOx and NO3 values in the field. When
looking at the NOx values, no direct trend becomes visible, as the points seem to be randomly
scattered across the graph. The simple linear regression indicates a slightly positive relation (Δ𝑦
Δ𝑥 = 0.053). However, the simple linear regression that was setup lacks statistical significance (p = . 823). These findings also hold for the NO3 values. Again, there is a slightly positive relationship
(Δ𝑦Δ𝑥 = 0.034), but this simple linear regression also lacks statistical significance (p = . 887).
Figure 3. NOx concentrations (left) and NO3 concentrations (right) plotted against the distance to the highway, including trendlines.
Figure 4 plots the PO4 values of a sample point against the dissolved organic nitrogen values at the
same point, to investigate the possible relationship between the two nutrients. Figure 4a shows the absolute amounts, while figure 4b ranks each sample point and includes a trendline, visualizing the Spearman's rank correlation coefficient. This relationship is statistically significant (p = .012).
Figure 4. DON and PO4 relations per sample point. The absolute values are presented on the left graph, while on the right each sample point is ranked on both DON and PO4 to visualize Spearman's rank correlation coefficient
Discussion
Highway influence
While a negative relationship was expected between the distance to the highway and the NOx values,
the simple linear regression obtained in the results was actually slightly positive (Figure 3). The same trend is true for the NO3 concentrations in relation to the distance to the highway, which is not
surprising since in practice the vast majority of the NOx measured was NO3. However, both these
regressions had high corresponding p-values. These findings point against an influence from the highway in providing NOx to the field. This is most likely due to a combination of two reasons. First of
all, in other studies the emissions of highways the increased nitrogen values were mostly found in the first couple of meters next to the highway (Kirchner et al., 2005; Khalid et al, 2020). Next to this, it is likely that the NO3 values are so high in the field due to the application of fertilizers, that this
overshadows the possible trend of having more NO3 deposition closer to the highway, if there is one.
Farm influence
The original intention was to also look at the influence of the dairy farm directly north of the field. This dairy farm is home to 225 cows whose feces are stored beneath the barn for later use as manure. From this location, NH3 escapes as a gas. This mostly reacts to form NH4 in the moist and pH-neutral
circumstances of the soil. NH4 is one of the forms of nitrogen that is immediately available for plant
uptake. Therefore, the NH3 emissions could result in spatial variability of plant available nitrogen
within the field. Research by Verhagen & van Diggelen (2006) found increased nitrogen deposition as far as 1500 from the source dairy farm. However, as with the highway influence, it is questionable if this increase would be noticeable on fertilized soil, as it could be overshadowed by nitrogen applied with fertilizers.
In this research, the NH4 values were very low and in most cases below the detection limit, making the
setup of a simple linear regression impossible. These low NH4 values could be caused by most of the
NH4 had already formed NO3. However, that process usually takes at least several weeks which would
mean that this field was not recently fertilized. PO4 and DON relation
The relation between PO4 and DON was similar across the field. These are both added to the field by
manure which is relatively homogenous in its distribution of nutrients, while the processes that could remove PO4 and dissolved organic nitrogen happen in different circumstances (Kemppainen, 1989).
This indicates that the nutrient variability across the field is at least partially explained by the unequal distribution of manure. This is however not a full-proof conclusion but just an indication, since there is the possibility that these nutrients were actually removed from the field at a similar rate at each sample location.
For the PO4 values of sample point 13, I decided not to classify them as outliers since both the
duplicates gave PO4 values that were more than 1.5 times the interquartile range above the third
quartile. Therefore I interpreted that it was likely that the values were in fact very high at this location, as the chance that it was caused by a misread of the auto-analyzer was low. However, when visualized in the graph projected against the dissolved organic nitrogen values, these values did somewhat resemble outliers. In case these values would have been classified as outliers, it would have increased
the p-value from the Spearman coefficient to .033. This is an increase, but still statistically significance since it is below p = .05.
The lack of results on NH4 values also made it impossible to look at the relation of the two nutrients
present in the mineral fertilizer that is used on this field. Since without it, only one of the two nutrients that form components of the mineral fertilizer were properly measured so no comparison can be set up. If the theory that all NH4 had already reacted to form NO3 is true, the spatial variability in NO3
could possibly be in part explained by the unequal distribution of mineral fertilizer when it was applied.
Other explanations
Another likely explanation for the spatial variability of NO3 across the field is nitrate leaching or runoff.
The initial plan is to also take a deeper look into this on a larger scale in this project. However, the processes associated with this are quite complex and would require much additional data. Ideally, another experiment should be set up monitoring the runoff of water at several points in the field. Both the volume and the NO3 concentrations of the water at these points would need to be measured to
give an indication of the amount of NO3 being lost at these locations. Runoff of NO3 is known to cause
lower NO3 concentrations near the edges of fields, especially where ditches are present (Harmel et al.,
2009). The potato field directly borders a ditch on the southern edge of the field. The NO3 values are
indeed among the lowest on the two points closest to the ditch, which are sample points 15 and 16 (figure 2). This could indicate that water containing high amounts of NO3 was drained from the field
into the ditch. However, sample point 20 was only slightly further away from the ditch, and actually had a slightly higher than average NO3 concentration.
A final factor that might have caused spatial variability is the bulb field that lies directly west of the potato field. The highest NO3 value was found directly next to this other field at sample point 5, while
other sample locations that lie nearby the bulb field (points 3, 19 and 20) also showed above average NO3 concentrations (Figure 2). Via runoff, NO3 could have been transferred from the bulb field to the
western sample locations of the potato field. This could also explain why the NO3 concentration of
sample point 20 was not lower than average, while the two other sample points that were close to the ditch on the south side did have much lower values.
A limiting factor during this research was the lack of precise information on fertilizer use by the farmer. Several details, including the exact date and amount of application, as well as the method used to apply the fertilizer, were not able to be retrieved. This resulted in several problems. First of all, it made it harder to asses how much of the spatial variability was caused by the method of application, and to what extent improvements could be made using different application methods. Additionally, it would have been interesting to compare the nutrient levels in the soil to the amount of nutrient originally added, and investigate how much of it could still be found in the soil.
It also needs to be mentioned that since this research only uses data that was collected during one fieldwork session, it is in fact just a snapshot of the nutrient levels at that time. Since these nutrients are subjected to many different processes, such as chemical reactions, plant-uptake, leaching and runoff, their concentrations vary over time. To get the full picture, more measurements would need to be taken over time, ideally over a whole year.
Solutions
Possible solutions that could reduce the spatial variability across the field, and in that way ensuring more adequate nutrient levels across the field, mostly include a more equal distribution of fertilizers. Due to the lack of knowledge on current fertilization methods, it is hard to what extent additional improvements can be achieved. However, not all spatial variability can be sustainably reduced by a more adequate application of fertilizers. For example, solving the lower NO3 levels found at the
southern end of the field bordering the ditch would require extra fertilizer application there. While this may work in theory, it would only increase the environmental impact of the fertilization regime, as more NO3 would runoff to the surface water. A solution to reduce edge of field nitrogen runoff
could be introducing integrated buffer zones between the ditch and the field. Research by Zak et al. (2018) has shown that this could decrease total nitrogenrunoff by 10-67%, while total phosphorus runoff could be reduced by 31-69%. Another method that could reduce NO3 leaching is the addition
of organic amendments to the soil. Radersma & Smit (2011) found that the addition of paper pulp could reduce nitrogen leaching by as much as 63-70%.
Conclusion
In conclusion, the nitrogen and phosphorus concentrations determined during this research showed substantial spatial variability. The lowest values of both NO3 and PO4 where several times lower than
the highest measured values for each respective nutrient. These large differences in nutrient levels indicate that further optimization in supplying nutrients is still possible, which would result in more efficient fertilizer use. Options to improve optimisation include more careful fertilizer application, the introduction of integrated buffer zones, and the additions of organic amendments to the soil. The simple linear regression analysis that was setup to look at the influence of the distance to the highway on the NOx and NO3 values did not have a significant result (p = . 823, and (p = . 887, respectively). This
indicates that no significant influence from the highway on nutrient levels was found in our results. The relatively constant relation between PO4 and DON (p = .012) indicates that manure applications
resulted in spatial variability across the field. Also, NO3 runoff near the ditch at the southern end of
the field may be the cause for the low NO3 values found there, while the high NO3 values found near
the border with the bulb field to the west may in turn be caused by runoff from the bulb field. However, due to the few amount of sample points at these location no hard conclusions can be drawn. To get a more complete picture of the causes of spatial variability in this field, additional research would be necessary. This research would ideally monitor nutrient runoff and changes in nutrient levels over time. Additionally, exact information on the amount of fertilizer used and the date and method of application could give additional information on the amount of nutrients lost and the reasons for spatial variability. Finally it could help giving more concrete advise to be able to reduce spatial variability in this field.
References
Ayoub, A. (1999) Fertilizers and the environment. Nutrient Cycling in Agroecosystems 55, 117–121. Brandenberger, L. P., Bowser, T. J., Zhang, H., Carrier, L. K., & Payton, M. E. (2016). Evaluation of Testing Kits for Routine Soil Analyses. Discourse Journal of Agriculture and Food Sciences, 4(1), 1-10. Burton, C. H., & Turner, C. (2003). Manure management: Treatment strategies for sustainable
agriculture. Editions Quae.
Cooper, J., Lombardi, R., Boardman, D., Carliell-Marquet, C. (2011) The future distribution and production of global phosphate rock reserves. Resources, Conservation and Recycling 57, 78-86 Dragosits, U., Theobald, M. R., Place, C. J., Lord, E., Webb, J., Hill, J., ... & Sutton, M. A. (2002). Ammonia emission, deposition and impact assessment at the field scale: a case study of sub-grid spatial variability. Environmental Pollution, 117(1), 147-158.
Edixhoven, J.D, Gupta, J., and Savenije, H.H.G. (2014). Recent revisions of phosphate rock reserves and resources: a critique. Earth System Dynamics, 5, 491–507.
Eghball, B., Wienhold, B. J., Gilley, J. E., & Eigenberg, R. A. (2002). Mineralization of manure nutrients. Journal of Soil and Water Conservation, 57(6), 470-473.
Geisseler, D., Scow, K.M. (2014). Long-term effects of mineral fertilizers on soil microorganisms – A review. Soil Biology and Biochemistry, 75, 54-63
Harmel, R. D., Smith, D. R., Haney, R. L., & Dozier, M. (2009). Nitrogen and phosphorus runoff from cropland and pasture fields fertilized with poultry litter. Journal of soil and water conservation, 64(6), 400-412.
Jongbloed, A. W., Lenis, N. P. (1998). Environmental concerns about animal manure. Journal of
Animal Science, 76(10), 2641–2648
Khalid, N., Noman, A., Masood, A., Tufail, A., Hadayat, N., Alnusairi, G. S., ... & Aqeel, M. (2020). Air pollution on highways and motorways perturbs carbon and nitrogen levels in roadside
ecosystems. Chemistry and Ecology, 36(9), 868-880.
Kemppainen, E. (1989). Nutrient content and fertilizer value of livestock manure with special reference to cow manure. Journal of the Agricultural Research Centre 28(3).
Kirchner, M., Jakobi, G., Feicht, E., Bernhardt, M., & Fischer, A. (2005). Elevated NH3 and NO2 air concentrations and nitrogen deposition rates in the vicinity of a highway in Southern
Bavaria. Atmospheric Environment, 39(25), 4531-4542.
Motesharezadeh, B., Etesami, H., Bagheri-Novair, S. (2017). Fertilizer consumption trend in
developing countries vs. developed countries. Environmental Monitoring and Assessment, 189, 103. (2017).
Nolin, M. C., Guertin, S. P., & Wang, C. (1996). Within‐field spatial variability of soil nutrients and corn yield in a Montreal lowlands clay soil. In Proceedings of the Third International Conference on
Precision Agriculture (pp. 257-270). Madison, WI, USA: American Society of Agronomy, Crop Science
Society of America, Soil Science Society of America.
Radersma, S., & Smit, A. L. (2011). Assessing denitrification and N leaching in a field with organic amendments. NJAS-Wageningen Journal of Life Sciences, 58(1-2), 21-29.
Robertson, G. P., Vitousek, P. M. (2009). Nitrogen in Agriculture: Balancing the Cost of an Essential Resource. Annual Review of Environment and Resources, 34(1), 97-125.
Smith, B. E. (2002). STRUCTURE: Nitrogenase Reveals Its Inner Secrets. Science, 297(5587), 1654-1655.
Udawatta, R. P., Motavalli, P. P., Garrett, H. E., & Krstansky, J. J. (2006). Nitrogen losses in runoff from three adjacent agricultural watersheds with claypan soils. Agriculture, Ecosystems &
Environment, 117(1), 39-48.
Van Meirvenne, M., Hofman, G. (1989) Spatial variability of soil nitrate nitrogen after potatoes and its change during winter. Plant Soil, 120, 103–110.
Velthof, G.L., Koeijer, T., Schröder, J.J., Timmerman, M., Hooijboer, A., Rozemeijer, J., van Bruggen., C., Groenendijk, P., (2017). Effecten van het mestbeleid op landbouw en milieu: Beantwoording van
de expostvragen in het kader van de evaluatie van de Meststoffenwet (Report No. 2782).
Wageningen Environmental Research. https://www.wur.nl/nl/download/Effecten-van-het-mestbeleid-op-landbouw-en-milieu.htm
Verhagen, R., & Van Diggelen, R. (2006). Spatial variation in atmospheric nitrogen deposition on low canopy vegetation. Environmental pollution, 144(3), 826-832.
Zak, D., Kronvang, B., Carstensen, M. V., Hoffmann, C. C., Kjeldgaard, A., Larsen, S. E., ... & Jensen, H. S. (2018). Nitrogen and phosphorus removal from agricultural runoff in integrated buffer
zones. Environmental science & technology, 52(11), 6508-6517.
Zhang, H., Johnson, G. How to get a good soil sample. Oklahoma Cooperative Extension Service. Available online at http://www.poultrywaste.okstate.edu/files/f2207web.pdf
Appendix Sample point NO3 (μmol/kg) NO2 (μmol/kg) NH4 (μmol/kg) DON (μmol/kg) PO4 (μmol/kg) Distance to highway (m) 1 3500,68 5,76 0,00 343,56 19,40 847,0 2 1694,21 19,10 8,75 299,46 19,11 684,0 3 3484,44 4,49 0,00 276,93 12,35 683,0 4 882,73 8,82 0,00 189,94 9,40 768,5 5 4728,56 6,07 0,00 159,50 16,14 727,0 6 1851,83 5,98 0,00 121,44 11,98 815,5 7 3279,80 4,61 0,00 238,14 13,26 859,0 8 2057,03 6,46 0,00 182,06 32,29 909,5 9 3567,79 15,67 8,20 451,30 28,50 894,5 10 1702,27 16,88 0,00 183,94 5,57 711,0 11 3103,23 6,33 0,00 138,26 20,29 757,0 12 3398,30 101,46 0,00 166,21 13,86 1035,0 13 4005,20 86,71 9,70 445,34 19,71 1052,0 14 3135,91 8,52 0,00 505,95 54,86 1045,5 15 1427,46 4,42 0,00 152,32 11,12 1067,5 16 1121,03 0,00 9,05 386,15 25,78 1008,0 17 3519,69 4,11 0,00 301,15 20,21 978,0 18 1887,26 501,66 10,15 253,06 18,59 925,0 19 2771,45 4,71 0,00 319,73 32,36 885,0 20 3008,85 61,53 0,00 246,69 7,72 926,0
The raw data is available via the following link:
https://docs.google.com/spreadsheets/d/170jd8FakP_KN83qPJLUpOizeW0YRsyu-a1DcC8Slt9w/edit?usp=sharing
