Nutr Cycl Agroecosyst (2016) 105:51–59 DOI 10.1007/s10705-016-9773-2
ORIGINAL ARTICLE
Phosphorous stock changes in agricultural soils: a case study in Turkey ¨ zbek . Adrian Leip . Fethi S¸ aban O Marijn Van der Velde
Received: 21 January 2016 / Accepted: 18 March 2016 / Published online: 22 March 2016 Ó Springer Science+Business Media Dordrecht 2016
Abstract Given that phosphorus (P) is a non-renewable and finite resource, there is an increasing need to sustainably use P in agriculture. To this end, accurately assessing P budgets in agricultural soils is critical. On one hand P deficiency negatively affects plant and animal growth, while on the other hand P surplus can cause significant problems that affect water quality (e.g. eutrophication and low oxygen level). The method to estimate the Phosphorus Budget as proposed by Eurostat and the OECD was developed under the assumption of zero changes in soil P stock (SSC-P), due to the lack of available data. However, studies have shown that SSC-P cannot be neglected to assess P efficiency properly. In this study, an approach is proposed that allows estimating SSC-P, as well as related indicators. The largest uncertainty in the method derives from the lack of evidence on the maximum achievable Phosphorus Use Efficiency (PUEmax). The national and regional P
Electronic supplementary material The online version of this article (doi:10.1007/s10705-016-9773-2) contains supplementary material, which is available to authorized users. ¨ zbek (&) F. S¸ . O Agricultural Statistics Department, Turkish Statistical Institute, Devlet Mah. Necatibey Cad. No:114, C¸ankaya, 06420 Ankara, Turkey e-mail:
[email protected];
[email protected] A. Leip M. Van der Velde Institute for Environment and Sustainability, European Commission - Joint Research Centre, Ispra, VA, Italy
budgets of agriculture in Turkey were estimated using the improved methodology at the level of administrative regions for the period 2007–2011. Results give a range of regional SSC-P of -4.8 and 3.9 kg P ha-1 year-1 with a mean national SSC-P of -0.3 kg P ha-1 year-1 (assuming a PUEmax of 100 %). However, the average SSC-P for Turkey could be as large as -1.4 kg P ha-1 year-1 for a PUEmax of 80 %. Keywords Phosphorus budgets Phosphorus surplus Phosphorus use efficiency Soil depletion
Introduction Phosphorus (P) is a crucial nutrient for all living cells (Johnston and Dawson 2005; White et al. 2010). Given that P is a non-renewable and finite resource, there is an increasing awareness for the need to sustainably use P in agriculture (e.g. Pen˜uelas et al. 2013). To this end, accurately assessing phosphorus budgets in agricultural soils is critical. On one hand low soil P status can lead to P deficiency in crops negatively affecting their growth and production (e.g. MacDonald et al. 2011). On the other hand phosphorus surplus (PS) can cause problems that affect water quality (e.g. Johnston and Dawson 2005) with negative effects on biodiversity, eutrophication, and low oxygen level in waters (Sharpley et al. 1994; Smith 1998; Hansen et al. 2002). Indeed, the risk of P movement from agricultural fields to adjacent water bodies increases with P
123
52
accumulation in soil (e.g. Djodjic et al. 2005; Sims et al. 2000). Phosphorus budgets are used to estimate P deficiencies (PD) or PS in agricultural lands (e.g. OECD 2001; MacDonald et al. 2011; CAPRI 2013; Eurostat 2013). Eurostat (the statistical office of the European Union, http://ec.europa.eu/eurostat), has developed a methodology for quantifying P budgets (Eurostat 2013), hereafter referred to as PB-Eurostat. European Union Member States are required to apply the PBEurostat methodology to estimate their national P budgets (Eurostat 2013). PB-Eurostat yields valid results only when there are no substantial changes in soil phosphorus stock (SSC-P). Data on SSC-P, however, are scarce and a general guidance for its estimation at the regional or national level could therefore not be developed (Eurostat 2013). Like the PB-Eurostat methodology, large-scale P-budget assessments generally lack the accounting for SSC-P, even though it is a crucial component of the P budget. For instance, the estimation of residual soil P built up in P-intensive agricultural systems led Sattari et al. (2012) to calculate reduced estimates of future fertilizer P requirements for these systems. The budget approach also provides information to quantify P use efficiency (PUE) (Chien et al. 2012; Syers et al. 2008; Johnston and Syers 2009) as an alternative to the difference method, a traditional method where percent recovery of fertilizer P is calculated (Chien et al. 2012). Johnston et al. (2014) showed that there was a strong relationship between SSC-P and PUE. The budget approach leads to biased PUE estimates under non-zero SSC-Ps conditions (e.g. Mahisarakul et al. 2002; Chien et al. 2012). Under conditions of soil P depletion, apparent PUE values exceeding 100 % can be obtained with the budget approach (Syers et al. 2008; Johnston et al. 2014). Clearly, there is need to better quantify soil P stock changes. This can improve the PB-Eurostat methodology and lead to more accurate assessments of PUE in experimental studies. Characterizing residual soil P is a challenge. In contrast to nitrogen, which is generally mobile in soil, the concentration of P in the soil solution is generally low, while large amounts of P are bound to organic material or soil minerals (e.g. Cross and Schlesing 1995). The P mineralization rate depends on the nature of the organic matter, soil moisture, oxygen and importantly pH, which in turn affect plant available P through the reverse process of
123
Nutr Cycl Agroecosyst (2016) 105:51–59
immobilization (e.g. Zou et al. 1992). In addition, at low soil pH levels iron and aluminium compounds will fix P rendering it unavailable to plants. Not withstanding this complexity, developing a simplified methodology to estimate SSC-P is required to assess regional P budgets. Estimates of soil N depletion or accumulation in agricultural soils have been obtained using a hyperbolic regression model based on N-input and N-output data of a gross nitrogen budget both at the national (Ozbek and Leip 2015; Lassaletta et al. 2014) and regional level (Ozbek and Leip 2015). In the current work, the same approach is used for developing a model to estimate SSC-Ps of the sub-regions of Turkey by using data at NUTS2 (Nomenclature of Territorial Units for Statistics) administrative level. The purpose of this study is twofold: first, to develop a method that allows estimating SSC-P, as well as PUE under conditions of SSC-P; and second, to apply the method to estimate the SSC-P values for the 26 subregions of Turkey for the period from 2007 to 2011. Materials and methods The present study estimates SSC-Ps in Turkish NUTS2 regions by developing a model that uses a hyperbolic relationship between P input and P output in the PB, and improves PS and PUE values of Turkish regions by taking account of their SSC-P values. The PB-Eurostat method used in this study is based on the methodology recommended in Eurostat/OECD common guideline (Eurostat 2013). Input data required for the PB-Eurostat Phosphorus surplus is estimated by subtracting the total amount of P contained in the outputs from the total amount of P contained in the inputs and the result is divided by the reference area Aref . Aref is the total UAA (utilized agricultural area), comprising arable land, permanent crop land, and utilized permanent grassland. PUE is estimated as the ratio of the total amount of P contained in the outputs and the total amount of P contained in the inputs. The inputs and outputs used in PS and PUE estimations, and the methodology and the data sources used in estimations of these inputs and outputs are presented in Table 1. The analysis is done at NUTS2 level for the time period from 2007 to 2011 for which data was available .
Nutr Cycl Agroecosyst (2016) 105:51–59
53
Table 1 The inputs and outputs used in PB-Eurostat, and the calculation methodology and the data sources used in estimations Inputs and outputs in P balance
Calculation method
Data sources and explanations
Inputs 1. Mineral fertilizer
Pmin
The amount of mineral fertilizer * P content
MFAL (yearly sales data)
2. Manure production
Pman
Number of animals * P excretion ratio
Number of animals: TurkStat (yearly administrative data), all animal types Excretion coefficients: Eurostat, OECD
3. Net manure imports/exports, withdrawals, stocks
PDmin
Manure imports - manure exports - withdrawals ? stock exchange
Negligiblea
4. Other organic fertilizer
Porg
The amount of organic fertiliser * P content
Negligiblea
5. Atmospheric deposition
Patm
Utilized agricultural area * Regional average deposition rate
Negligiblea
6. Seed and planting materials
Pseed
Cropped areas of wheat and potatoes * P seed input rate
Cropped area: TurkStat (yearly administrative data) Nutrient seed input rate: Eurostat, crop specific
7. Total inputs = sum (1, 2, 3, 4, 5, 6)
Pinput
¼ Pmin þ Pman þ PDman þ Porg þ Patm þ Pseed
Pcrop
Crop production * P content
Outputs 8. Crop production
Crop production: TurkStat (yearly administrative data), all crop types P content: Eurostat, OECD, crop specific
9. Fodder production
Pfod
(Fodder production * P content) ? (pasture and meadows area * yield * Consumption ratio * P content)
Fodder production: TurkStat (yearly administrative data) Pasture and meadows area: TurkStat (2001 General Agricultural Census) Yield, consumption: OECD; P content: Eurostat, OECD, crop specific
10. Crop residues removed
Pres
Crop residues removed * P content
Negligiblea
11. Stock changes of P in soil
PDsoil
12. Total outputs = sum (8,9,10,11)
Poutput
b
See Eq. 2b
See Eq. 2
Pinput Poutput Aref
P surplus = 7–12
PS
PS ¼
P use efficiency = 12/7
PUE
PUE ¼
Poutput Pinput
100
TurkStat Turkish Statistical Institute, MFAL Ministry of Food, Agriculture and Livestock a
See text (‘‘Input data required for the PB-Eurostat’’ section)
b
See text (‘‘The PB-soil stock change model (PB-model)’’ section)
Data show that the import and export of animal and plant manure for Turkey (TurkStat 2013) is negligible (less than 0.1 %) (Ozbek and Leip 2015), and manure stock exchange is excluded in accordance with the Eurostat/OECD common guidelines (Eurostat 2013). Therefore, import, export, and stock changes of farm manure have not been considered in the calculations. Atmospheric P deposition was neglected in accordance with the Eurostat/OECD common guideline (Eurostat 2013).
The PB-soil stock change model (PB-model) P output was regressed against P input data for the Turkish NUTS2 regions using a hyperbolic relationship according to Eq. 1 (e.g. Willcutts et al. 1998; Mathews and Hopkins 1999; Overman and Scholtz 2002; Lassaletta et al. 2014; Ozbek and Leip 2015). To obtain a regression model representing conditions with low SSC-Ps only data points for which the calculated PUE was below a given assumed maximum PUE (PUEmax) were used. As evidence for an
123
54
Nutr Cycl Agroecosyst (2016) 105:51–59
upper limit of PUEmax is lacking, an analysis of the sensitivity of the SSC-P estimation to PUEmax was performed with PUEmax ranging from 80 to 100 % (in 5 % steps) and included in ‘‘SSC-P estimations under different PUEmax’’ section. Nevertheless, the main quantitative results are presented using a PUEmax of 100 % (i.e. leading to smaller estimates of soil P depletion). A total of 106 data points (from a possible 130) at NUTS2 level were included for the model development when using a PUEmax of 100 %. Pout:model ¼
a Pin b þ Pin
ð1Þ
where Pout;model is the simulated total P output including also changes in SSC-P in any NUTS2 region, Pin is the total P input in the region obtained from the statistical data, and a and b are model parameters, fitted using the Excel Solver tool and minimizing RMSE (root mean square error—a measure of the misfit of the model to the data). On the basis of Eq. 1 it is possible to calculate SSCP under the assumption that SSC-P is the difference of apparent (thus statistical) and real ‘useful output’ of P. According to Leip et al. (2011) and Ozbek and Leip (2015) soil accumulation (positive SSC-P) preserves the nutrient for future use and increases thus the ‘useful output’ while soil depletion (negative SSC-P) removes nutrient from the plant available pool and thus decreases the ‘useful output’ (Eq. 2). SSCP ¼ Pout;model Pout
the time period 2007–2011 for which the PB-Eurostat model resulted in a PUE less than 100 %. The Nash– Sutcliffe efficiency (NSE) is 0.66; the ratio of the root mean square error to the standard deviation of measured data (RSR) is 0.59. The performance of the model is thus considered satisfactory (Moriasi et al. 2007). Applying the model to all NUTS2 regions gives an average SSC-P in Turkey, averaged over the 5 years 2007–2011, of -0.3 kg P ha-1 year-1. Maximum P soil accumulation over the 5 years was calculated in East Marmara (TR42) and maximum P soil depletion was calculated in the Aegean (TR31) with the values of 3.9 and -4.8 kg P ha-1 year-1 obtained respectively. The lowest annual SSC-P value was estimated at -10 kg P ha-1 year-1 for the year 2010, and the highest at 5 kg P ha-1 year-1 for the year 2011. The results of 5-year average SSC-P for the two regions [in Northeast Anatolia (TRA1) and in Mediterranean (TR63)] for which PB-Eurostat estimated PUE is greater than 100 % were estimated as -2.1 and -4.0 kg P ha-1 year-1, respectively (Fig. 2). A table with all data used and results obtained is given in the Supplementary Information. As seen in Fig. 3, the absolute ratio of SSC-P to Pout;model was low (3 %) in Turkey, when averaged over the 5-year period. At the regional level, it varied between -2 and -54 % in soil P depleting NUTS2 regions, and varied between 1 and 33 % in soil P accumulating NUTS2 regions.
ð2Þ
where Pout is the observed output of P obtained from statistical information. This information can now be used to calculate the indicators PS and PUE (Eqs. 3 and 4). PSmodel ¼ Pin Pout;model PUEmodel ¼
Pout;model Pin
ð3Þ ð4Þ
where PSmodel is the modelled P surplus and PUEmodel is the modelled P use efficiency. Results SSC-P estimations Figure 1 shows the scatter of P output versus P input in the statistical data for all NUTS2 regions in Turkey over
123
Fig. 1 The scatter plot diagram of Poutput and Pinput at NUTS2 regions with PUE-Eurostat \PUEmax = 100 % in Turkey for the period of 2007–2011. Nash–Sutcliffe efficiency (NSE): 0.66; ratio of the root mean square error to the standard deviation of measured data (RSR): 0.59
Nutr Cycl Agroecosyst (2016) 105:51–59
55
Fig. 2 Soil P stock changes (kg P ha-1 year-1) for Turkey and its NUTS2 regions for the years 2007–2011
PS and PUE estimations under SSC-P Over all regions, the 5 years average PSmodel values varied between 0.3 kg P ha-1 year-1 (Northeast Anatolia—TRA1) close to the northeast border of Turkey and 11.9 kg P ha-1 year-1 in East Marmara (TR42) close to the northwest border of Turkey. Over the 5-year period, the highest average PUEmodel value was calculated for TRA1 with a value of 93 %, and the minimum PUEmodel was calculated for TR42 at 56 % (Fig. 4). SSC-P estimations under different PUEmax A sensitivity analysis of the PUEmax on the SSCP estimation was performed with PUEmax ranging from 80 to 100 % (in 5 % steps, see Table 2). Applying the hyperbolic relationships for respectively a PUEmax of 100 % and 80 % results in 5-year average SSCP’s in Turkey ranging from -0.3 to -1.4 kg P ha-1 year-1. SSC-P values calculated for the other PUEmax values (95, 90, and 85 %) respectively equalled -0.4, -0.7, and -0.9 kg P ha-1 year-1.
Discussion The application of the PB-Eurostat (Eurostat 2015) calculates values of PUE [ 100 % for several countries in Europe (Estonia, Bulgaria, and Hungary) which is a clear indication that P depletion occurs (MacDonald et al. 2011; Leip et al. 2011; Ozbek 2014;
Johnston et al. 2014) and the PB-Eurostat method is thus not valid for those countries. This method is based on the assumption that changes in soil nutrient stocks are small compared to input and output levels of nutrients and uncertainties in derived indicators are thus small (Eurostat 2013; OECD 2013). For Turkey, the values of P indicators of PBEurostat and the PB-model at the national level are very close (Fig. 4); however, the agreement between PB-Eurostat and the PB-model at the regional level is poor. Indeed, the data suggests that the contribution of SSC-P in P-output is significant (Fig. 3) and therefore cannot be neglected for many regions. Hence, efforts have to be made towards reliable SSC-P estimates when calculating P budgets. This indicates that the regionalisation of P balance is important to obtain more accurate results, and that there is a need to monitor the P indicators at regional level. Most negative SSC-Ps and thus soil P depletion were observed in the Aegean (TR31) and in the Mediterranean (TR62) regions. The main reason for soil depletion in these regions is intensive cropping cycles with insufficient fallow periods. Indeed, for all regions, a high negative correlation (r = -0.7) between P removed with crops and fodder and SSCP was observed for the period 2007–2011. The highest SSC-Ps and thus soil P accumulation were calculated in the West Black Sea (TR81) and in the East Marmara (TR42) regions. The main reason for soil P accumulation there is the high manure input per unit area in these regions. TR81 is the second highest region in
123
56
Nutr Cycl Agroecosyst (2016) 105:51–59
Fig. 3 Five years average (2007–2011) of soil P stock changes (SSC-P) and P-output-model (kg P ha-1 year-1) for Turkey and its NUTS2 regions
Turkey in terms of the number of cattle per unit area, and TR42 is the highest region in terms of the number of poultry per unit area. In almost half of the regions soil P accumulation occurs. Unfortunately, to our knowledge there are no studies reporting P indicators based on the budget approach for Turkey to compare our data with. The accumulation of soil P suggests that P is being used inefficiently in these regions (c.f. Johnston et al. 2014). This represent a potential risk to water quality impairment (e.g. eutrophication and low oxygen level in waters) as P accumulation in
123
soil increases the risk of P movement from agricultural fields to adjacent water bodies (e.g. Djodjic et al. 2005; Sims et al. 2000). On the other hand, if appropriately considered in farm management plans, P accumulated in soils gives the opportunity to reduce P application rates in following years (Sattari et al. 2012). GDWW (2015) have assessed the pollution level of water bodies in Turkey with regard to P (and other nutrients/compounds) and found high pollution levels, which were set by assessing P levels in water bodies with water quality standards set in national regulation
Nutr Cycl Agroecosyst (2016) 105:51–59
57
Fig. 4 PSEurostat and PUEEurostat (filled circle), PSmodel and PUEmodel (open triangle) for Turkey and its NUTS2 regions for the years 2007–2011
on Water Pollution Control, in the basins of Yesilırmak, Kızılırmak, Marmara, Susurluk, Western Black Sea, Eastern Black Sea, Sakarya. This is coherent with our findings of high P surpluses for NUTS2 regions located in these basins. There are a few studies for Turkey (Ozturk et al. 2005; Ibrikci et al. 2005) which have calculated P recovery percentage and the P efficiency ratio in terms of relative shoot growth. P recovery percentages were calculated by dividing fertilized minus non-fertilized P uptake by grain by fertilizer P added. Ibrikci et al. (2005) obtained values ranging between 10.8 and 46.4 % as a function of P application rates. Phosphorus efficiencies were estimated as the ratio of shoot dry matter production under low to that under adequate P supply. Results
from Ozturk et al. (2005) varied between 46.7 and 78.6 % as a function of genotype. Even though different methodologies were employed, this range is coherent with our results when compared with the 5-year average national PUE (73.6 %) and the regional PUE values that were obtained (see Fig. 4). In comparison with the results obtained for nitrogen (SSC-N) in Turkey (Ozbek and Leip 2015), the absolute percentage of SSC-N in N-output (0.3 %) was lower than the share of SSC-P in P-output (3 %). This result is consistent with the fact that P is less mobile in soil as compared to N (Chien et al. 2012) and thus more easily accumulates in soil compared to N. This confirms the necessity to estimate SSC-P values in an attempt to obtain reliable results for interpreting PS and PUE.
123
58
Nutr Cycl Agroecosyst (2016) 105:51–59
Table 2 Average SSC-P for NUTS2 regions in Turkey over the years 2007–2011 for different PUEmax values (kg P ha-1 year-1) PUEmax regions
100 %
95 %
90 %
85 %
80 %
TR10
-0.9
-0.9
-1.2
-1.5
-1.9
TR21
-4.3
-4.4
-4.7
-4.9
-5.4
TR22
0.2
0.1
-0.2
-0.4
-0.9
TR31
-4.8
-4.8
-5.2
-5.4
-5.8
TR32
-3.6
-3.7
-4.0
-4.2
-4.7
TR33
1.2
1.0
0.8
0.5
0.1
TR41
0.4
0.3
0.0
-0.2
-0.7
TR42
3.9
3.9
3.6
3.4
3.0
TR51
1.1
0.9
0.7
0.5
0.1
TR52
0.6
0.5
0.2
0.0
-0.4
TR61 TR62
-1.6 -4.6
-1.8 -4.7
-2.1 -5.0
-2.3 -5.2
-2.8 -5.7
TR63
-4.0
-4.2
-4.5
-4.7
-5.2
TR71
0.1
-0.1
-0.3
-0.5
-0.9
TR72
-0.6
-0.7
-0.9
-1.1
-1.4
TR81
3.5
3.4
3.1
2.9
2.4
TR82
2.2
2.1
1.8
1.6
1.1
TR83
-0.5
-0.7
-0.9
-1.1
-1.6
TR90
0.8
0.7
0.5
0.3
0.0
TRA1
-2.1
-2.2
-2.4
-2.5
-2.7
TRA2
0.6
0.4
0.2
0.0
-0.4
TRB1
-0.1
-0.2
-0.4
-0.5
-0.9
TRB2
-1.0
-1.2
-1.4
-1.6
-2.0
TRC1
-1.5
-1.6
-1.9
-2.0
-2.5
TRC2
-2.2
-2.3
-2.6
-2.8
-3.3
TRC3
-0.4
-0.5
-0.8
-1.0
-1.5
Region codes TR10: Istanbul; TR21, TR22: West Marmara; TR31, TR32, TR33: Aegean; TR41, TR42: East Marmara; TR51, TR52: West Anatolia; TR61, TR62, TR63: Mediterranean; TR71, TR72: Central Anatolia; TR81, TR82, TR83: West Black Sea; TR90: Eastern Black Sea; TRA1, TRA2: Northeast Anatolia; TRB1, TRB2: Central-east Anatolia; TRC1, TRC2, TRC3: Southeast Anatolia
Conclusion Data on SSC-P in agricultural soils in Turkey are not available. It is therefore not possible to validate the results obtained. The assumption of negligible P soil stock changes is not valid in many circumstances. Indeed, our data strongly suggest that long-term changes in soil P stocks are important for many regions in Turkey. Estimating P indicators accurately by taking account of the SSC-P and monitoring the
123
indicators regularly among the years is an important element to ensure sustainable farm management. One of the most uncertain elements for the estimation of SSC-P is the theoretical maximum PUE; the data suggest that on the average soil P losses range between -0.3 and -1.4 kg P ha-1 year-1 for Turkish agricultural soils, respectively for a PUEmax of 100 % and 80 %. Depletion of soil P stocks threatens soil fertility. P surplus is a risk for water quality and is a waste of a scarce resource. Accumulation of P in soils is also a potential risk for water quality unless this knowledge is transferred into farm management plans that adjust P application rates. The regional coherence of high water pollution levels and high soil P accumulation values suggests that in these regions opportunities to adjust P application rates exist.
References CAPRI (2013) CAPRI modelling system common agricultural policy regionalised impact modelling system. http://www. capri-model.org/dokuwiki/doku.php?id=capri:concept:nBal Chien SH, Sikora FJ, Gilkes RJ, McLaughlin MJ (2012) Comparing of the difference and balance methods to calculate percent recovery of fertilizer phosphorus applied to soils: a critical discussion. Nutr Cycl Agroecosyst 92:1–8 Cross AF, Schlesing WH (1995) A literature review and evaluation of the Hedley fractionation: applications to the biogeochemical cycle of soil phosphorus in natural ecosystems. Geoderma 64(3–4):197–214 Djodjic F, Bergstro¨m L, Grant C (2005) Phosphorus management in balanced agricultural systems. Soil Use Manag 21:94–101 Eurostat (2013) Nutrient budgets, EU-27, NO, CH. Methodology and handbook. Version 1.02. Luxembourg, Eurostat Eurostat (2015) Gross nutrient balance. http://appsso.eurostat. ec.europa.eu/nui/show.do?dataset=aei_pr_gnb&lang=en. Eurostat GDWW (2015) Basin protection action final reports (in Turkish). http://suyonetimi2.ormansu.gov.tr/su/AnaSayfa/Hav za_Yonetimi_Planlamasi_Daire-Baskanligi/HavzaKoru ma.aspx?sflang=tr Hansen NC, Daniel TC, Sharpley AN, Lemunyon JL (2002) The fate and transport of phosphorus in agricultural systems. J Soil Water Conserv 57:408–417 Ibrikci H, Ryan J, Ulger AC, Buyuk G, Cakir B, Korkmaz K, Karnez E, Ozgenturk G, Konuskan O (2005) Maintenance of phosphorus fertilizer and residual phosphorus effect on corn production. Nutr Cycl Agroecosyst 72:279–286 Johnston AE, Dawson CJ (2005) Phosphorus in agriculture and in relation to water quality. Peterborough, UK, Agricultural Industries Confederation Johnston AE, Syers JK (2009) A new approach to assessing phosphorus use efficiency in agriculture. Better Crops 93(3):14–16
Nutr Cycl Agroecosyst (2016) 105:51–59 Johnston AE, Poulton PR, Fixen PE, Curtin D (2014) Phosphorus: its efficient use in agriculture, advances in agronomy, vol 123. Academic Press, Burlington, pp 177–228 Lassaletta L, Billen G, Grizzetti B, Anglade J, Garnier J (2014) 50 year trends in nitrogen use efficiency of world cropping systems: the relationship between yield and nitrogen input to cropland. Environ Res Lett 9:105011 Leip A, Britz W, Weiss F, de Vries W (2011) Farm, land, and soil nitrogen budgets for agriculture in Europe calculated with CAPRI. Environ Pollut 159:3243–3253. doi:10.1016/ j.envpol.2011.01.040 MacDonald GK, Bennett EM, Potter PA, Ramankutty N (2011) Agronomic phosphorus imbalances across the world’s croplands. PNAS 108(7):3086–3091 Mahisarakul J, Siripaibool C, Claimon J, Pakkong P (2002) Field assessment of the relative agronomic effectiveness of phosphate rock materials in a soybean-maize crop rotation using 32P isotopic techniques. In: Assessment of soil phosphorus status and management of phosphatic fertilizers to optimize crop production. IAEA-TECDOC-1272, International Atomic Energy Agency (IAEA), Vienna, Austria, pp 265–274 Mathews BW, Hopkins KD (1999) Superiority of S-shaped (sigmoidal) yield curves for explaining low-level nitrogen and phosphorus fertilization responses in the humid tropics. J Hawaii Pac Agric 10:33–46 Moriasi DN, Arnold JG, Van Liew MW, Bingner RL, Harmel RD, Veith TL (2007) Model evaluation guidelines for systematic quantification of accuracy in watershed simulations. Trans ASABE 50(3):885–900 OECD (2001) Environmental indicators for agriculture: methods and results. OECD, Paris OECD (2013) OECD Compendium of agri-environmental indicators. OECD Publishing, Paris. doi:10.1787/97892 64186217-en Overman AR, Scholtz RV (2002) Mathematical models of crop growth and yields. Marcel Dekker, New York Ozbek FS (2014) Estimation of national and regional phosphorus budgets for agriculture in Turkey. Span J Agric Res 12(1):52–60. doi:10.5424/sjar/2014121-4327 Ozbek FS, Leip A (2015) Estimating the gross nitrogen budget under soil nitrogen stock changes: a case study for Turkey. Agric Ecosyst Environ 2(1):109
59 Ozturk L, Eker S, Torun B, Cakmak I (2005) Variation in P efficiency among 73 bread and durum wheat genotypes grown in a P-deficient calcareous soil. Plant Soil 269:69–80 Pen˜uelas P, Poulter B, Sardans J, Ciais P, Van der Velde M, Bopp L, Boucher O, Godderis Y, Hinsinger P, Llusia J, Nardin E, Vicca S, Obersteiner M, Janssens IA (2013) Human-induced nitrogen–phosphorus imbalances alter natural and managed ecosystems across the globe. Nat Commun 4(December):2934. doi:10.1038/ncomms3934 Sattari SZ, Bouwman AF, Giller KE, Van Ittersum MK (2012) Residual soil phosphorus as the missing piece in the global phosphorus crisis puzzle. Proc Natl Acad Sci USA 109(16):6348–6353. doi:10.1073/pnas.1113675109 Sharpley AN, Chapra SC, Wedepohl R, Sims JT, Daniel TC, Reddy KR (1994) Managing agricultural phosphorus for protection of surface waters: issues and options. J Environ Qual 23(3):437–451 Sims JT, Edwards AC, Schoumans OF, Simard RR (2000) Integrating soil phosphorus testing into environmentally based agricultural management practices. J Environ Qual 29:60–71 Smith VH (1998) Cultural eutrophication of inland, estuarine, and coastal waters. In: Pace ML, Groffman PM (eds) Successes, limitations, and frontiers in ecosystem science. Springer, New York Syers K, Johnston E, Curtin D (2008) Efficiency of soil and fertilizer phosphorus use. FAO Fertilizer and Plant Nutrition Bulletin, Rome TurkStat (2013) Foreign trade statistics. https://biruni.tuik.gov. tr/disticaretapp/menu_ing.zul. Accessed 15 Mar 13 White S, Cordell D, Moore D (2010) Securing a sustainable phosphorus future for Australia: implications of global phosphorus scarcity and possible solutions. Institute for Sustainable Futures, for CSIRO Sustainable Agriculture Flagship SENSE Earth Systems Governance, Amsterdam, 24th–31st August (2008) Willcutts JF, Overman AR, Hochmuth GJ, Cantliffe DJ, Soundy P (1998) A comparison of three mathematical models of response to applied nitrogen: a case study using lettuce. Hort Sci 33:833–836 (Williams, G. P. 1997) Zou X, Binkley D, Doxtader KG (1992) A new method for estimating gross phosphorus mineralization and immobilization rates in soils. Plant Soil 147(2):243–250
123