Nutr Cycl Agroecosyst DOI 10.1007/s10705-015-9718-1

ORIGINAL ARTICLE

Phosphorus mobilization from sugarcane soils in the tropical environment of Mauritius under simulated rainfall Tesha Mardamootoo . Chris C. du Preez . Andrew N. Sharpley

Received: 2 February 2015 / Accepted: 31 July 2015 Ó Springer Science+Business Media Dordrecht 2015

Abstract The continuous application of phosphorus (P) to agricultural systems to ensure profitable crop productivity can lead to an accumulation of P in agricultural soils. While this long-term residual pool of soil P is desirable from an agronomic perspective, there is some concern about its possible impacts on surface water quality as it may lead to eutrophication. Since a better understanding of P transport processes provides useful information for the development of site-specific P management strategies, rainfall simulation studies (100 mm h-1 for 30 min) on runoff plots (2.1 m 9 0.75 m) were conducted at 20 field sites to study the mobilisation of soil P from sugarcane fields of

Electronic supplementary material The online version of this article (doi:10.1007/s10705-015-9718-1) contains supplementary material, which is available to authorized users. T. Mardamootoo (&) Agricultural Chemistry Department, Mauritius Sugarcane Industry Research Institute, Mauritius Cane Industry Authority, Re´duit, Mauritius e-mail: [email protected] C. C. du Preez Department of Soil, Crop and Climate Sciences, University of the Free State, Bloemfontein 9300, South Africa e-mail: [email protected] A. N. Sharpley Department of Crop, Soil and Environmental Sciences, University of Arkansas, Fayetteville, AR, USA e-mail: [email protected]

Mauritius. The research findings indicated that the edge-of-plot P losses were insignificant from an agronomic perspective but only small amounts of P can actually trigger eutrophication in freshwaters. The results also showed that total P concentrations in runoff are more strongly associated with runoff sediments (r2 = 0.92) than runoff volume (r2 = 0.49) indicating that a greater proportion of the P transported in runoff occurred as particulate P rather than dissolved P. Actually, about 89 % of total P loss in runoff waters was mobilised in particulate form, pointing to the importance of erosion as a mechanism for mobilising soil P. The research findings suggest that in addition to the current management practices aiming at reducing runoff volume, such as conservation tillage and trash cover, measures that reduce sediments in runoff, such as grassed waterways and riparian buffers, may further attenuate P losses. Keywords Eutrophication  Phosphorus loss  Surface runoff  Soil erosion  Soil phosphorus testing  Water quality

Introduction The continuous application of phosphorus (P) either in the form of mineral fertilisers or organic sources to agricultural systems to ensure profitable crop productivity can lead to an accumulation of P in agricultural

123

Nutr Cycl Agroecosyst

soils (Bender et al. 2013; Sharpley et al. 2013). From an agronomic point of view, this build-up of P in soils is desirable but on the other hand represents a threat to water quality (Jarvie et al. 2015). The environmental significance of P lies in its dominant role in accelerating the eutrophication of waters, impairing it for fisheries, recreation and drinking purposes (Foy 2005; Dodds et al. 2009). Although nitrogen (N) and carbon (C) are also necessary to the growth of aquatic biota, most of the attention has focused on P due to the difficulty in controlling exchange of N and C between the atmosphere and water, and fixation of atmospheric N by certain cyanobacteria (Kleinman et al. 2015; Sharpley and Wang 2014). Phosphorus is in fact the most limiting nutrient influencing eutrophication of surface waters, generally at a P concentration which is tenfold lower than that required for plant growth (Guidry et al. 2006). In general, to maintain the quality of waters, total P should not exceed 50 lg L-1 in streams entering lakes/reservoirs, or 25 lg L-1 within lakes/reservoirs as per directives of the U. S. Environmental Protection Agency (Daniel et al. 1998; EvansWhite et al. 2013; Kleinman et al. 2011). All forms of P in the soil whether they are soluble, adsorbed, precipitated, or organic are susceptible to transport to water bodies (Pierzynski et al. 2000). Transport of soil P occurs primarily via surface flow when the water flowing across the soil surface either dissolves and transports soluble P or detaches and transports particulate P (Sharpley and Withers 1994). Sharpley et al. (1996) found that dissolved (soluble) P in runoff originates from the release of P from a thin zone of surface soil (2–2.5 cm) and/or from vegetative material that interacts with rainfall. Leaching and subsurface lateral flow of P are a concern only in some specific circumstances, for example in soils with a high degree of P saturation or in fields where the artificial drainage systems provides a pathway for the water and dissolved solutes to move through the soil (Smith et al. 2015). However, it is reasonably well established that in most watersheds, P export occurs mainly in overland flow (Hart et al. 2004). Phosphorus loss via surface runoff and erosion may be reduced by conservation tillage and crop residue management, buffer strips, riparian zones, terracing, contour tillage, cover crops and settling basins (Sims and Kleinman 2005; Sharpley et al. 2015). Agricultural runoff has been identified as one of the potential threats to water quality in the tropical region

123

of Mauritius (Ministry of Environment and Sustainable Development 2011). As highlighted by Janeau et al. (2014), most work on nutrient transport from agricultural lands has focused on temperate agricultural production systems with tropical and sub-tropical ecosystems being ignored. Moreover, few studies have been undertaken in Mauritius to study P mobilization from the agricultural landscape, which is predominantly occupied by sugarcane fields (about 53,500 ha which represents 86 % of the agricultural land). In general, erosion from sugarcane fields is thought to be efficiently controlled due to the perennial characteristics of the crop and the related cultivation practices such as post-harvest management practices (Anon 2007). Actually in some countries (e.g., Guinope, Honduras) sugarcane itself has been employed as a ‘live barrier’ or in border planting to act as a soil conservation cover crop to reduce erosion (Cheesman 2004). The relative value of sugarcane as a cover crop is enhanced by the fact that the crop tends to remain in the ground for a number of years, due to ratooning. This multi-year growing cycle produces an extensive root system and the presence of a closed canopy during a major portion of the crop cycle, protects the soil from the erosive effects of rain (Cheesman 2004). In addition, sugarcane post-harvest management practices such as green cane trash blanketing following mechanised harvesting has also been found to reduce erosion since the mulch formed reduces the impact of raindrops on the soil surface thus decreasing surface runoff and also acts as a filter retaining soil particles in suspension in the surface water (Ng Cheong et al. 2003). Jeong et al. (2011) assessed P runoff from sugarcane fields in Louisiana, USA under different postharvest management practices and found that the adoption of practices that retain residue after harvest (e.g. green cane trash blanketing) improved water quality of both surface runoff and subsurface leachate. An assessment of land use impact on water quality in northern Queensland, Australia revealed that the concentrations of total N, P and suspended solids in streams increased as the proportion of land under sugarcane production increased (Bramley and Roth 2002). In order to address this issue, Bramley et al. (2003) developed a risk assessment tool which combined P sorption indices and runoff susceptibility of different soil types, to estimate the potential for P loss,

Nutr Cycl Agroecosyst

thereby promoting an improved management of P fertilizers in sugarcane fields. In Mauritius, a study carried out at a sub-catchment scale at Valetta on a Humic Ferruginous Latosol soil over three sugarcane growth seasons from 1997 to 2001, to study the movement of agrochemicals and nutrients including P (Ng Kee Kwong et al. 2002) showed that P concentrations of limnological significance (i.e. [50 lg P L-1) in streams flowing past sugarcane fields were measured, particularly after intense rainfall events. As climate, soil type, topography, and other management practices vary across the island, data from that site was valid only for that soil type and climatic conditions prevailing during the years when the measurements were made. Another study to evaluate erosion on different soil types of Mauritius and to estimate different parameters influencing erosion (e.g. soil erodibility and erosivity factors) was undertaken, but did not measure P losses (Seeruttun et al. 2007). Therefore, the objective of this study was to evaluate the extent of P losses occurring during erosion and surface runoff from sugarcane fields in the tropical environment of Mauritius during simulated rainfall events.

Materials and methods Study area The rainfall simulation experiments were carried out on the tropical island of Mauritius, situated in the South West Indian Ocean, approximately 800 km east of the Malagasy Republic. Although Mauritius occupies an area of 186,000 ha, there is owing to its topography a high variation in rainfall over very short distances. Mean annual rainfall changes abruptly from 800 mm on the west coast to over 4000 mm in the central tableland over a distance of only 20 km. Mauritius is actually divided into three agro-climatic zones, namely the sub-humid (\1600 mm annual rainfall), humid (1600–2600 mm annual rainfall), and super-humid zones ([2600 mm annual rainfall). On the basis of the differences in annual rainfall and age of volcanic parent materials, Parish and Feillafe´ (1965) used a genetic classification system previously adopted in Hawaii to map 13 different soil groups with the prominent ones shown in Fig. 1. However, nearly 90 % of the sugarcane lands in Mauritius are located

on five soil groups, the Low Humic Latosol, Humic Latosol, Humic Ferruginous Latosol, Latosolic Reddish Prairie, and Latosolic Brown Forest soil groups. The rain simulations were conducted at 20 sites (in duplicate) representing those five dominating soils under which sugarcane is cultivated. Soil samples (0–15 and 15–30 cm) were collected at each study site pled adjacent to the runoff plots, so as to avoid disrupting the hydrologic conditions of the soil inside the plots (National Phosphorus Research Project 2015). In general, the soils of Mauritius are generally acidic to near neutral with pH values ranging from 5.3 to 6.2, with soil texture varying from clay loam to clayey with an average of not less than 32.6 % clay (Table 1). Differences in the cation exchange capacity (CEC) and concentration of the exchangeable bases among the five main soil groups in Mauritius are related to rainfall, which affects the intensity of leaching in the soils (Cavalot et al. 1988). Thus, it can be noted that the Low Humic Latosol and Latosolic Reddish Prairie, which are located in the low rainfall regions (800–2600 mm) have the highest CEC among the Latosols and between the two Latosolic soils, respectively. The organic matter content of the soils varies between 3.7 and 6.5 % with the Low Humic Latosols, as the name suggests having the lowest soil organic matter and the Latosolic Reddish Prairie having the highest organic matter content (Table 1). Rainfall simulation tests and sample collection Rainfall simulation on runoff plots was a preferred choice over natural rainfall as it is not only relatively cheaper but also provides a more rapid, efficient, and reliable means of investigating the effects of site variability, such as soil type and P concentration on P runoff under controlled and constant rainfall (Dougherty 2006). In fact, rainfall simulation was originally developed to assess factors controlling erosion (Meyer 1965) and was later adapted and found suitable to evaluate P mobilisation under a wide range of conditions and management practices (Kleinman et al. 2006; Miller et al. 2006; Sharpley and Kleinman 2003). Simulated rainfall runoff studies as mentioned previously were carried out on 20 sites (Fig. 1) having slopes ranging between 8 to 20 % on sugarcane fields in Mauritius following the protocol of the National

123

Nutr Cycl Agroecosyst

Fig. 1 The different soil groups in Mauritius according to Parish and Feillafe´ (1965) and the location of study sites

123

Mean annual rainfall (mm)

1500

Humic Latosol (Humic Nitosol)c

c

b

0–15 15–30

5.5 ± 0.1 5.3 ± 0.1

6.3 ± 0.2 6.3 ± 0.1

0–15

15–30

5.9 ± 0.2 5.6 ± 0.2

0–15

15–30

5.5 ± 0.2 5.5 ± 0.2

0–15

6.0 ± 0.3

15–30

15–30

6.2 ± 0.3

pH (H2O)

0–15

Soil depth (cm)

6.5 ± 0.7 30.2 ± 6.8 6.3 ± 0.1 32.7 ± 1.2

3.9 ± 0.3 48.5 ± 2.0

4.0 ± 0.3 47.0 ± 3.4

3.7 ± 0.4 31.4 ± 3.4

4.4 ± 0.5 34.0 ± 1.9

4.3 ± 0.2 42.4 ± 4.3

4.5 ± 0.1 40.5 ± 3.5

3.8 ± 0.2 58.1 ± 5.0

34.1 ± 2.4 35.7 ± 5.0 30.5 ± 0.8 36.8 ± 2.0

24.4 ± 1.2 27.1 ± 2.5

22.5 ± 1.1 30.6 ± 2.9

32.5 ± 2.0 36.1 ± 2.3

30.2 ± 2.1 35.8 ± 2.4

29.0 ± 4.3 28.6 ± 1.6

31.0 ± 2.4 28.6 ± 2.4

21.8 ± 1.6 20.1 ± 3.4

23.4 ± 2.0 22.6 ± 3.7

Sand

Soil group as per FAO classification (Arlidge and Wong You Cheong 1975)

Rocky shallow immature shallow soils developed from lava flows dating back to less than 100,000 years

3.63 ± 1.10 2.51 ± 0.50

10.76 ± 0.74

10.68 ± 0.97

3.10 ± 1.73

4.12 ± 1.76

3.38 ± 0.86

3.62 ± 0.84

7.38 ± 1.71

7.40 ± 1.87

cmol? kg-1

1.91 ± 0.50 1.13 ± 0.08

5.27 ± 0.57

4.92 ± 0.32

0.59 ± 0.13

0.83 ± 0.12

0.78 ± 0.06

0.91 ± 0.13

2.24 ± 0.94

2.23 ± 0.83

Mg

Ca

Silt

Clay %

Exchangeable bases

Particle-size analysis

3.9 ± 0.2 54.0 ± 5.5

Organic matter (%)

Rock-free deep mature soils developed from lava flows dating back to approximately 7 million years

2500

Latosolic Brown Forest (Dystric Cambisol)c

a

1200

Latosolic Reddish Prairie (Eutric Cambisol)c

2500 Humic Ferruginous Latosol (Humic Acrisol)c Latosolic soil groupb

1200

Low Humic Latosol (Humic Nitosol)c

Latosol soil groupa

Soil group

0.52 ± 0.10 1.80 ± 1.56

1.97 ± 0.47

1.95 ± 0.46

0.18 ± 0.07

0.37 ± 0.09

0.23 ± 0.05

0.31 ± 0.05

1.30 ± 0.77

1.72 ± 1.15

K

Table 1 Pertinent characteristics (mean ± SE) of the five main soil groups under sugarcane in Mauritius (adapted from Mardamootoo et al. 2013)

0.33 ± 0.03 0.25 ± 0.04

0.63 ± 0.08

0.67 ± 0.04

0.20 ± 0.03

0.29 ± 0.02

0.23 ± 0.03

0.28 ± 0.03

0.48 ± 0.15

0.70 ± 0.34

Na

6.4 ± 1.1 5.7 ± 1.2

18.6 ± 1.1

18.2 ± 1.4

4.1 ± 1.9

4.8 ± 1.9

4.6 ± 0.9

5.1 ± 1.0

11.4 ± 3.4

12.0 ± 3.6

Cation exchange capacity

Nutr Cycl Agroecosyst

123

Nutr Cycl Agroecosyst

Phosphorus Research Project (2015). The experiments were carried out just after planting since it is at this time, when plant cover is absent, that the risk of soil detachment and P movement due to recent P fertiliser application is most probable. Moreover, the study plots were wetted and brought to field capacity a day prior to the rainfall simulation tests so as to reduce hydrologic variability between plots related to spatial differences in antecedent soil moisture. Rainfall was generated using a rainfall simulator similar to that described by Loch et al. (2001) and consisted of three Veejet 80100 nozzles mounted 1 m apart on a manifold, with the nozzles controlled to sweep across a plot width of 1.5 m (Loch et al. 2001). The operational sequence of the rainfall simulator relied on a continuous water flow through the nozzles and the water was supplied to the rainfall simulator from an electric water pump (0.55 kW) operating at a pressure of 70 kPa. A coefficient of uniformity of 90 % was obtained for the rainfall over a two meter square plot. Rainfall was simulated for a total duration of 30 min at each site with a rainfall intensity of 100 mm h-1 such that runoff and erosion could be generated within a reasonable time period. While rainfall return periods are not available for Mauritius, the rainfall rate used in this research represents the upper 75 % percentile of events across the island. At each study site, the runoff plots consisted of two abutting plots, each 2.1 by 0.75 m, with the longer side installed in the direction of the slope and along the cane rows. Metal frames were installed with their edges 5 cm above the ground to isolate runoff plots from the rest of the field. The surface runoff originating from the plots was diverted towards a down slope runoff gutter into a collection vessel. This surface runoff plot configuration is consistent with that established by the National Phosphorus Research Project (2015) and found to represent P mobilisation and transport from larger plots (i.e. 10 m2; Sharpley and Kleinman 2003). However, P runoff data collected from these 1.5 m2 plots is used to compare loss as a function of site variables and does not reflect losses that might occur on a field scale under natural rainfall. During the 30-min simulated rainfall, the soil eroded during each runoff event was collected using a 2 mm sieve and the dry weight was used to estimate soil loss per unit area. The remaining fraction (\2 mm) was collected in runoff samples. Whenever it occurred, the runoff was collected and its volume

123

measured at 5 min intervals during the 30 min simulated rainfall event. The individual runoff samples collected were then thoroughly mixed to ensure a uniform sediment suspension before a subsample (1 L) was collected in a Nalgene bottle and transported to the laboratory for analysis.

Laboratory determinations The amount of sediment present in the runoff sample was determined by evaporating to dryness a sufficient aliquot volume, usually ranging between 50 and 100 ml of homogenized runoff sample. This suspended sediment load represents the finer particles usually held in suspension in flowing water, which eventually settles out when water velocity decreases (Falken 1995). The method of Tandon et al. (1968) was used to determine total P in the eroded soil particles while the total P concentration of runoff samples (i.e. in the suspended load) were determined in perchloric acid digested samples (American Public Health Association 1992) for the standard methods for the examination of water and waste water. Mardamootoo et al. (2013) found that a good correlation (r2 = 0.93) existed between total dissolved P as determined by perchloric acid digestion and orthophosphate P concentrations in runoff samples, thus avoiding the need to analyse for total dissolved P concentration in the runoff samples by acid digestion. The dissolved P in this study was therefore determined directly in 0.45 lm filtered runoff samples as orthophosphate-P by the method of Murphy and Riley (1962). Particulate P was then estimated as the difference between total P in runoff and total dissolved P. The agronomic soil P levels were determined by the method of Cavalot et al. (1988), which involved shaking 1 g of air-dried soil with 50 ml of dilute acid (0.1 M H2SO4) and determining the amount of extracted P by the colorimetric by the method of Murphy and Riley (1962). This extraction method of Cavalot et al. (1988) has been validated by field experimental data which also provided the basis for formulating fertiliser recommendations at planting. The application of fertiliser P is not recommended to sugarcane soils with more than 80 mg P kg-1 as extracted by 0.1 M H2SO4. Moreover, field experiments have shown that above 80 mg P kg-1, the residual effect of past P fertilization would be

Nutr Cycl Agroecosyst

expected to maintain the maximum yield level during a sugarcane growth cycle of 6–7 years in Mauritius (Cavalot et al. 1988). Data processing The flow weighted concentration of P in the runoff waters during the 30 min simulated rainfall event was calculated as follows: Flow weighted concentration of P in runoff waters 6   P runoff Pj  Vj ¼ i¼1 6 P ðVj Þ i¼1

where runoff Pj (i = 1, 2, 3, 4, 5, 6) and Vj (i = 1, 2, 3, 4, 5, 6) represent respectively the concentration of P and the volume of runoff. Regression and correlation analysis was utilized to study the experimental data.

Results and discussion Vulnerability of soils to erosion in Mauritius The amount of eroded soil particles transported from the five soils varied considerably as observed in Table 2, clearly reflecting a different erosion potential across the soil groups studied. Variation in erosion rates of the different soils has been attributed to differences in soil properties such as soil texture, aggregate stability, shear strength, infiltration capacity and organic matter content (Woodward and Foster

1997; Sallaway et al. 2001). Indeed the presence of organic matter in soils is known to act as a cementing agent which renders soils less erodible (Cheesman 2004). Based on that principle, it would have been expected that the Low Humic Latosol soils which has the lowest organic matter content (3.9 %) out of the five soils would be the most erodible. In fact, that was not the case and results shown that the Latosolic Reddish Prairie soils were the most erodible under the current field and experimental conditions. A possible explanation could be that the removal of rocks and land grading to facilitate cultivation of the naturally rocky Latosolic Reddish Prairie soils prior to planting, inevitably disturbed the stable soil structure causing more soil particles to be eroded during the simulated rainfall runoff events. Moreover, derocking and land grading during field preparation for sugarcane planting is known to reduce organic matter content of soil (Ng Cheong et al. 2009) and may, therefore, increase vulnerability of soils to erosion. As reviewed by Gburek et al. (2005), erosion is a size-selective process and preferential erosion of smaller sized particles often occurs. Consequently, Latosolic Reddish Prairie soils which has a relatively high clay content (47 %) represents a particular risk to P transport. Of particular importance in Table 2 is that the risks of soil particles being detached from the soil surface and transported to the edge of the plot during runoff events is of the order of 62 and 67 % on Humic Ferruginuous Latosol and Latosolic Brown Forest soils respectively, which are located in the high rainfall regions of Mauritius ([2600 mm). Even though the amount of soil eroded on these soils is not as high as Latosolic Reddish Prairie, the propensity of having runoff events following rainfall combined

Table 2 Amount of eroded soil particles and sediments (expressed as oven-dried weight per unit area) carried during a 30 min simulated rainfall of 100 mm h-1 for the five main soil groups of Mauritius Soil type

Occurrence of soil erosion during runoff events (%)

Amount of eroded soil particles (kg ha-1)

Sediment load (g ha-1)

Range

Range

Organic matter (%)

Clay (%)

Low Humic Latosol (LHL)

3.9

54.0

37

0.00–10.10

2.14 ± 1.31

0.02–0.17

0.07 ± 0.05

Humic Latosol (HL) Humic Ferruginuous Latosol (HFL)

4.5 4.4

40.5 34.0

100 62

4.16–22.38 0.00–25.84

9.34 ± 2.76 6.77 ± 3.79

0.04–0.62 0.02 –0.09

0.16 ± 0.09 0.04 ± 0.01

Latosolic Reddish Prairie (LRP)

4.0

47.0

67

0.00–107.08

32.67 ± 13.74

0.01–0.37

0.15 ± 0.04

Latosolic Brown Forest (LBF)

5.5

30.2

67

0.00–12.54

3.87 ± 1.93

0.01–0.21

0.08 ± 0.04

Mean ± SE

Mean ± SE

123

Nutr Cycl Agroecosyst

with small erosion events could account for a significant portion of P loss from fields. Quinton et al. (2001) showed that high frequency of small magnitude erosion events have disproportionately large potential to cause P pollution than infrequent large events. Therefore, soil conservation practices (such as contour planting along hills and mountain slopes, planting ve´tiver or muguet as a cover crop or buffer strip on field borders on sloping lands and adoption of minimum tillage) should not only be targeted to Latosolic Reddish Prairie soils, where the amount of soil loss is highest, as well as where there is a combination of high occurrence of soil particle transport during runoffs (e.g., Humic Latosol, Humic Ferruginuous Latosol, Latosolic Brown Forest) and relatively high amount of eroded soil particles transport. Total P transported by erosion In an attempt to evaluate the risk of P mobility associated with erosion, the correlation between the amount of eroded soil particles transported during simulated rainfall runoff event and P attached to the eroded material was established. A strong positive linear relationship (y = 1.433x; r2 = 0.97) was obtained showing that the amount of total P mobilised tended to increase with increasing rates of eroded material (Fig. 2a). The amount of P mobilised during erosion from the runoff plots ranged between 2 and 163 g ha-1 for corresponding soil losses of 1–107 kg ha-1 at simulated rainfall intensities of 100 mm h-1 for 30 min. As the loss of eroded particles enriched in P is a function of erosion, conservation control measures that mitigate soil losses would concomitantly decrease associated P losses. Although these P losses appear insignificant from an agronomic point of view, only small quantities of P can impair water quality and cause accelerated eutrophication (Smith et al. 2015). Therefore, soil conservation practices, as proposed by the Mauritius Sugar Industry Research Institute (2007), which include contour planting along hills and mountain slopes, planting of ve´tiver or muguet on field borders of sloping lands and adoption of minimum tillage would also be effective in controlling P losses. Minimum tillage practices have indeed been adopted on sloping lands in Mauritius since 1980s for more effective erosion control (Ismael et al. 2008).

123

Sugarcane cultural practices, which reduce erosion (e.g. trash blanketing), will help mitigate P losses while soil compaction due to mechanization of cultural operations may exacerbate P transport via erosion. Despite the fact that erosion control has been shown to decrease particulate P losses, considerable concentrations of dissolved P can occur in runoff from soils where erosion is kept at a minimum, especially from soils with a high P status or having recently received P applications (Penn et al. 2005). Furthermore, the longterm adoption of conservation tillage practices as highlighted by Sims and Kleinman (2005) may actually exacerbate dissolved P losses in surface runoff due to stratification of P at the soil surface with an accumulation of P in the upper few centimetres, due to repeated surface applications of fertilisers or organic amendments without the mixing induced by plowing. The regression slope of the linear relationship in Fig. 2 actually represents the amount of P transported for every kilogram of soil that would be transported. It was observed that the regression slopes varied across the different soil types due to differences in the intrinsic soil properties. For example, the regression slopes were higher for the Low Humic Latosol of the Latosols and Latosolic Reddish Prairie of the Latosolic soils (Fig. 2b, c). This coincides with a comparatively higher clay content of 54 % and 47 % for the Low Humic Latosol and Latosolic Reddish Prairie, respectively, compared to the other soils of their respective groups (Table 2). Indeed, Quinton et al. (2001) showed that there was a strong relationship between clay and P content of soils. Penn et al. (2005) further explained that the high surface area and presence of various P sorbing minerals of high clay soils often absorb more P compared to other soils (e.g. coarsetextured soils). Therefore it would be expected that for similar erosion rates, more P will be mobilised from the high clay soils such as Low Humic Latosol and Latosolic Reddish Prairie. Erosion is actually a P transport mechanism that preferentially removes finersized soil particles and as a result the P content and reactivity of eroded material is usually greater than the soil from which it originated (Sharpley 1985). The term P-enrichment ratio is often utilised to represent the integrated effective P loss from soils by accounting for the disproportionate effect of finer particles (Gburek et al. 2005).

Nutr Cycl Agroecosyst Fig. 2 Relationship between the amount of eroded soil particles and the associated P losses for a all soils combined, b latosols and c latosolic soils following simulated rainfall events of 100 mm h-1 for a total duration of 30 min

(a)

(b)

0 180

5

10

15

20

25

30

80

100

120

(c)

160 140 120 100 80 60 40 20 0 0

20

40

Phosphorus losses by surface runoff In addition to erosion, surface runoff during rainfall events also represents a considerable pathway for P

60

losses. In that context, the ability of the agronomic soil P test to also estimate total P losses was tested by establishing the relationship between the soil P test levels (i.e., 0.1 M H2SO4 extractable P) and the total P

123

Nutr Cycl Agroecosyst

transported in runoff waters. From Fig. 3, it is clear that the agronomic soil P test, which was initially designed to measure the amounts of P that would be released during the crop growing season, did not show any correlation with total P losses in surface runoff during the simulated rainfall events. Agronomic soil P tests such as the 0.1 M H2SO4 extraction is related to the solubilisation of P and was found to better predict dissolved P losses as indicated in Mardamootoo et al. (2013). Withers et al. (2007) suggested that soil P methodologies with some predictive value for assessing the degree to which surface runoff becomes entrained with eroding particles and enriched in P are potentially useful. Besides, the poor correlation between soil test P levels and total P losses also indicates that there are other factors apart from soil P which are contributing to total P transfer in runoff. Actually, none of the factors controlling surface runoff such as rainfall characteristics, soil detachment and soil permeability is directly related to soil phosphorus, and it was therefore expected that soil P tests will not be suitable to predict off-site movement of total P. Total runoff P losses observed during the 30 min simulated rainfall of 100 mm h-1 ranged between 0.2 and 1346 g ha-1. In addition, a linear correlation (y = 7.29x - 3.71; r2 = 0.49) was obtained between runoff total P concentration and runoff volume (Fig. 4a). This relationship was considerably

1600

Total P in runoff (g ha-1)

1400 1200 1000 800 600 400 200 0 0

50

100

150

200

250

300

Agronomic soil P test level (mg kg-1) No runoff event

Total P losses during runoff events

Fig. 3 Relationship between the agronomic soil test P levels (using the 0.1 M H2SO4 extraction) and the total P loads in runoff waters following 30 min simulated rainfall events of 100 mm h-1

123

improved when total runoff P concentration was correlated to suspended load in runoff instead of runoff volume. In fact, the total P concentration was found to be linearly correlated to the suspended load present in runoff waters with a correlation coefficient of 0.92 (Fig. 4b). This close correlation indicates that a greater proportion of the P transported in runoff occurred as particulate P, rather than dissolved P. It also demonstrates that P compounds in soils are sparingly soluble and that P is strongly adsorbed to soil particles (Wang et al. 2010). Similar research findings have been obtained by Cox and Hendricks (2000) showing that total P losses in runoff were highly influenced by the amount of suspended sediment load in runoff waters. These findings suggest that management practices which reduce runoff volume (e.g. conservation tillage, cover crops) will to a certain extent reduce total P transport from fields. Moreover, reducing suspended sediments in runoff waters will also attenuate P losses and this can be achieved by having grassed waterways, riparian buffers and buffer strips. Forms of P mobilized Soil P may be transported by surface runoff when the water flowing across the soil surface either dissolves and transports soluble P or erodes and transports particulate P. From Fig. 5, it can be observed that irrespective of soil type, particulate P was the dominant form and accounted for 67–99 % of the total P mobilised in runoff waters. These results actually concur with previous studies done at a watershed scale in Mauritius by Ng Kee Kwong et al. (2002) which showed that P transport in surface runoff was intimately linked to sediment. Although dissolved P constitutes only a minor portion (1–33 %) of the total P mobilised in runoff waters, it is in a form which is immediately available for algal uptake causing accelerated eutrophication of freshwaters (Sharpley 1993). In contrast, particulate P constitutes a long-term source of P once deposited in water bodies and its bioavailability depends on the nature of the sediment as well as the surrounding environment (Truman et al. 1993). Results shown in Fig. 5, further clarify why the soil P test levels (i.e. 0.1 M H2SO4-P) did not correlate with total P losses (refer to Fig. 3) in runoff because most of the P was lost in particulate forms which

Nutr Cycl Agroecosyst

(a)

(b)

Total P in runoff (g ha-1)

Fig. 4 Relationship between total P in runoff waters and a runoff volume and b sediment load following a 30 min simulated rainfall event of 100 mm h-1

0.00

Runoff volume (L)

0.20

0.40

0.60

0.80

Sediment load (g ha-1)

99%

100

67%

80 60 40

33% 20 0

Low Humic Latosol (LHL)

Humic Latosol (HL)

7% Humic Ferruginous Latosol (HFL)

Particulate P

Latosolic Reddish Prairie (LRP)

Latosolic Brown Forest (LBF)

Dissolved P

Fig. 5 Average particulate and dissolved P losses (expressed as a percentage of total P) in runoff following 30 min simulated rainfall events of 100 mm h-1 on the different soil groups of Mauritius

involves the detachment of soils and sediments during rainfall and runoff instead of a desorption-dissolution reaction controlling P release from surface soils in dissolved forms. The agronomic soil P test (0.1 M H2SO4 extraction) utilised in Mauritius, extracts plant-available P fraction from the soil (i.e. soil P concentration) and gives no indication of the potential for particulate P losses in erosion or runoff. Quinton et al. (2003) emphasized that soil P tests in combination with sediment concentrations would provide a better means of assessing total P mobilisation in surface runoffs. In addition, it was found that the suspended sediment load present in the runoffs

accounted for about 93 % of the variation in the amount of particulate P measured in runoff samples collected during the simulated rainfall events (Fig. 6). Furthermore, the high correlation observed in Fig. 4b between the total P and suspended load was due to the apparent relationship between suspended sediment load and particulate P losses in runoffs (Fig. 6). The suspended load obviously plays an important role as a carrier of particulate P and controlled runoff P losses as explained by Miller et al. (2006). This therefore confirms that the adoption of management practices such as cover crops, grassed waterways and conservation buffers which intercept

123

Nutr Cycl Agroecosyst

y = 1899x R² = 0.93

0.00

0.20

0.40

0.60

0.80

Sediment load (g ha-1)

Fig. 6 Relationship between sediment load and particulate P measured in runoff samples collected during 30 min simulated rainfall of 100 mm h-1 at 20 sites in Mauritius

runoff flow and trap sediments from agricultural land will be effective in reducing P losses from cane fields in Mauritius.

Conclusions This study conducted under simulated rainfall, showed that surface runoff and erosion represents potential pathways for P mobilization from sugarcane fields in the tropical environment of Mauritius. The amount of P load mobilised during erosion ranged between 2 and 163 g P ha-1 while total P in runoff waters including sediments ranged between 0.2 and 1346 g ha-1 at 30 min simulated rainfall intensities of 100 mm h-1. While this amount of P loss appear insignificant from an economic and agronomic perspective, only small amounts of P can cause accelerated eutrophication of freshwaters. The research findings also showed that total P mobilised during runoff events was closely associated with suspended sediments present in runoff waters, indicating that soil detachment during erosion represents a significant mechanism for P losses during rainfall events. It was also observed that irrespective of the soil type, the major portion of P losses (89 % on average) occurred in particulate forms which were in turn strongly linked to suspended sediments present in runoff waters, confirming that suspended soil particles controlled runoff particulate P.

123

Additionally, the 0.1 M H2SO4 soil P test used to formulate P fertiliser recommendations for sugarcane in Mauritius was inappropriate to predict total P losses during surface runoff and erosion. Indeed, such soil P tests provide only an estimate of soil desorbable P without assessing the degree to which runoff waters becomes entrained with eroding particles and enriched in P. Knowledge about soil P transport via erosion and surface runoff in conjunction with soil P status (as determined by soil P tests) are key components required not only to predict off-site P movement but to also choose the best P management practices. Therefore, in order to adequately predict P losses from agricultural landscapes, a more holistic approach which combines the source of P (e.g. soil P status, fertiliser application) along with field hydrologic characteristics affecting P transport should be considered (Lemunyon and Gilbert 1993). Such a holistic approach includes the P indexing system, which integrates major source and transport factors controlling P movement so as to identify areas in a field which are most susceptible to P losses and thereby implementing improved P management to those areas at risk. This study using simulated rainfall on small runoff plots has provided information on factors governing P transport from sugarcane soils in Mauritius. Additional results gathered from this study will be incorporated into a nutrient management tool, notably the P index framework to identify areas in a field which are most vulnerable to P losses and therefore require sitespecific P management to reduce risks of P transport. This will be a step towards sustaining a profitable sugarcane industry while at the time reducing risks of water quality impairment in Mauritius. Acknowledgments The results reported in this paper are part of a research study for the award of a doctoral degree at the University of the Free State, South Africa. Funding of this project was provided by the European Union under the African Caribbean and Pacific Sugar Research Program (ACP-SRP). The contents of this document are the sole responsibilities of the authors and can under no circumstances be regarded as reflecting the position of the European Union.

References American Public Health Association (APHA), American Water Works Association (AWWA) and Water Environment

Nutr Cycl Agroecosyst Federation (WEF) (1992) Standard methods for the examination of water and wastewater. Franson MAH (eds), 18th edn Anon (2007) Soil erosion under sugar cane. In: Mauritius Sugar Industry Research Institute Annual Report 2006, Mauritius, pp 59–60 Arlidge EZ, Cheong YWU (1975) Notes on the land resources and agricultural suitability map of Mauritius 1: 50 000. Mauritius Sugar Industry Research Institute and FAO Occasional Paper No. 29, Mauritius, p 137 Bender R, Haegele JW, Ruffo ML, Below FE (2013) Nutrient uptake, partitioning, and remobilization in modern, transgenic insect-protected maize hybrids. Agron J 105:161–170. doi:10.2134/agronj2012.0352 Bramley RGV, Roth CH (2002) Land-use effects on water quality in an intensively managed catchment in the Australian humid tropics. Mar Freshwater Res 53:931–940. doi:10.1071/MF01242 Bramley RGV, Roth CH, Wood AW (2003) Risk assessment of phosphorus loss from sugarcane soils—a tool to promote improved management of P fertiliser. Soil Res 41:627–644. doi:10.1071/SR02099 Cavalot PC, Deville J, Ng Kee Kwong KF (1988) Refinement of method for prediction of soil P available to sugarcane in Mauritius. Revue Agricole et Sucrie`re Ile Maurice 67:55–63 Cheesman OA (2004) Use and impacts of by-products. Environmental impacts of sugar production: The cultivation and processing of sugarcane and sugar beet. CABI Publishing, CAB International, Wallingford, pp 151–172 Cox FR, Hendricks SE (2000) Soil test phosphorus and clay content effects on runoff water quality. J Environ Qual 29:1582–1586. doi:10.2134/jeq2000.0047242500290005 0027x Daniel TC, Sharpley AN, Lemunyon JL (1998) Agricultural phosphorus and eutrophication: a symposium overview. J Environ Qual 27:251–257. doi:10.2134/jeq1998. 00472425002700020002x Dodds WK, Bouska WW, Eitzmann JL, Pilger TJ, Pitts KL, Riley AJ, Schloesser JT, Thornbrugh DJ (2009) Eutrophication of US freshwaters: analysis of potential economic damages. Environ Sci Technol 43:12–19. doi:10.1021/ es801217q Dougherty WJ (2006) The mobilisation of soil phosphorus in surface runoff from intensively managed pastures in southeast Australia. Doctoral Thesis, University of Adelaide Evans-White MA, Haggard BE, Scott JT (2013) A review of stream nutrient criteria development in the United States. J Environ Qual 42(4):1002–1014. doi:10.2134/jeq2012. 0491 Falken M (1995) Runoff, erosion and sedimentation prediction and measurement. In: Land and water integration and river basin management. FAO Land and water Bulletin No. 1. FAO, Rome. http://www.fao.org/docrep/V5400E/V5400 E00.htm Accessed 27 Jan 2015 Foy RH (2005) The return of the phosphorus paradigm: Agricultural phosphorus and eutrophication. In: Sims JT, Sharpley AN (eds) Phosphorus: agriculture and the environment. Agronomy Monograph No. 46. American Society of Agronomy, Madison, Wisconsin, USA, pp 911–939

Gburek WJ, Barberis, E, Haygarth, PM, Kronvang B, Stamm C (2005) Phosphorus mobility in the landscape. In: Sims JT, Sharpley AN (eds) Phosphorus: agriculture and the environment. Agronomy Monograph No. 46. American Society of Agronomy, Madison, Wisconsin, USA, pp 941–979 Guidry AR, Schindler FV, German DR, Gelderman RH, Gerwing JR (2006) Using simulated rainfall to evaluate field and indoor surface runoff phosphorus relationships. J Environ Qual 35:2236–2243. doi:10.2134/jeq2006.0156 Hart MR, Quin BF, Nguyen M (2004) Phosphorus runoff from agricultural land and direct fertilizer effects. J Environ Qual 33:1954–1972. doi:10.2134/jeq2004.1954 Ismael FM, Seeruttun S, Barbe C, Gaungoo A (2008) A review of changing cultural practices to improve productivity of sugar cane in Mauritius. Proc S Afr Sugar Technol Assoc 81:539–544 Janeau JL, Gillard LC, Grellier S, Jouquet P, Le TPQ, Luu TNM, Ngo QA, Orange D, Pham DR, Tran DT, Tran SH, Trinh AD, Valentin C, Rochelle-Newall E (2014) Soil erosion, dissolved organic carbon and nutrient losses under different land use systems in a small catchment in northern Vietnam. Agric Water Manage 146:314–323. doi:10.1016/ j.agwat.2014.09.006 Jarvie HP, Sharpley AN, Flaten D, Kleinman PJA, Jenkins A, Simmons T. (2015) The pivotal role of phosphorus in a resilient water-energy-food security nexus. J Environ Qual (in press). doi:10.2134/jeq2015.01.0030 Jeong CY, Weindorf DC, DeRamus A, Goodeaux LL (2011) Surface and subsurface phosphorus losses from sugarcane fields with different management practices. Water Air Soil Pollut 217:649–661. doi:10.1007/s11270-010-0617-4 Kleinman PJA, Srinivasan MS, Dell CJ, Schmidt JP, Sharpley AN, Bryant RB (2006) Role of rainfall intensity and hydrology in nutrient transport via surface runoff. J Environ Qual 35:1248–1259. doi:10.2134/jeq2006.0015 Kleinman PJA, Sharpley AN, McDowell RW, Flaten DN, Buda AR, Tao L, Bergstrom L, Zhu Q (2011) Managing agricultural phosphorus for water quality protection: principles for progress. Plant Soil 349:169–182. doi:10.1007/s11104011-0832-9 Kleinman PJA, Sharpley AN, Withers PJ, Bergstro¨m L, Johnson LT, Doody DG (2015) Implementing agricultural phosphorus science and management to combat eutrophication. Ambio 44(Suppl. 2):297–310. doi:10.1007/s13280-0150631-2 Lemunyon JL, Gilbert RG (1993) The concept and need for a phosphorus assessment tool. J Prod Agric 6:483–486. doi:10.2134/jpa1993.0483 Loch RJ, Robotham BG, Zeller L, Masterman N, Orange DN, Bridge BJ, Sheridan G, Bourke JJ (2001) A multipurpose rainfall simulator for field infiltration and erosion studies. Aust J Soil Res 39:599–610 Mardamootoo T, Ng Kee Kwong KF, Du Preez CC (2013) Assessing environmental phosphorus status of soils in Mauritius following long-term phosphorus fertilization of sugarcane. Agric Water Manage 117:26–32. doi:10.1016/j. agwat.2012.10.022 Mauritius Sugar Industry Research Institute (2007) Good management practices to improve the productivity, competiveness and sustainability of the Mauritian sugar cane

123

Nutr Cycl Agroecosyst Industry. Mauritius Sugar Industry Research Institute, Mauritius, Re´duit Meyer LD (1965) Simulation of rainfall for soil erosion research. Trans ASAE 8:63–65 Miller JJ, Olson ECS, Chanasyk DS, Beasley BW, Larney FJ, Olson BM (2006) Phosphorus and nitrogen in rainfall simulation runoff after fresh and composted beef cattle manure application. J Environ Qual 35:1279–1290. doi:10. 2134/jeq2005.0389 Ministry of Environment and Sustainable Development (2011) Chapter 3 freshwaters. In: Mauritius Environment Outlook Report 2011, Mauritius: pp 43–58 Murphy J, Riley JP (1962) A modified single solution method for the determination of phosphate in natural waters. Anal Chim Acta 27:31–36. doi:10.1016/S0003-2670(00)884 44-5 National Phosphorus Research Project (2015) National research project for simulated rainfall-surface runoff studies: Protocol. North Carolina State University, Raleigh, USA https://sera17dotorg.files.wordpress.com/2015/02/nationalp-protocol.pdf Accessed 29 May 2015 Ng Cheong LR, Ah Koon PD, Ng Kee Kwong KF (2003) Effects de la pluviome´trie et des pratiques culturales sur le ruissellement et l’e´rosion dans la canne a` sucre a` Maurice. Revue Agricole et Sucrie`re de l’ıˆle Maurice 82:72–77 Ng Cheong LR, Ng Kee Kwong KF, Du Preez CC (2009) Soil compaction under sugar cane (Saccharum hybrid sp.) cropping and mechanization in Mauritius. S Afr J Plant Soil 26:199–205. doi:10.1080/02571862.2009.10639955 Ng Kee Kwong KF, Bholah A, Volcy L, Pynee K (2002) Nitrogen and phosphorus transport by surface runoff from a silty clay loam soil under sugarcane in the humid tropical environment of Mauritius. Agric Ecosyst Environ 91:147–157. doi:10.1016/S0167-8809(01)00237-7 Parish DH, Feillafe´ SM (1965) Notes on the 1: 100 000 soil map of Mauritius. Mauritius Sugar Industry Research Institute Occasional paper no. 22, Mauritius Penn CJ, Mullins GL, Zelazny LW (2005) Mineralogy in relation to phosphorus sorption and dissolved phosphorus losses in runoff. Soil Sci Soc Am J 69:1532–1540. doi:10. 2136/sssaj2004.0224 Pierzynski GM, Sim JT, Vance FG (2000) Soil phosphorus and environmental quality. In: Soils and environmental quality, 2nd edn. New York: pp 55–202 Quinton JN, Catt JA, Hess TM (2001) The selective removal of phosphorus from soil: is event size important? J Environ Qual 30:538–545. doi:10.2134/jeq2001.302538x Quinton JN, Strauss P, Miller N, Azazoglu E, Yli-Halla M, Uusitalo R (2003) The potential for soil phosphorus tests to predict phosphorus losses in overland flow. J Plant Soil Sci 166:432–437. doi:10.1002/jpln.200321122 Sallaway M, Ah Koon D, Ng Cheong R (2001) Processes driving agrochemical movement. In: Offsite movement of agrochemicals in tropical sugarcane production. Extension Workshop, Bundaberg, Australia, 8–9 May 2001. Re´duit: Mauritius Sugar Industry Research Institute, Bundaberg: Queensland Department of Natural Resources and Mines; Canberra: Australian Centre for International Agricultural Research, pp 49–60

123

Seeruttun S, Ah Koon D, Ismael FM, Ng Cheong R (2007) Sugarcane reduces soil erosion by more than 80 percent in five major soil groups of Mauritius. Proc Int Soc Sugar Cane Technol 26:384–387 Sharpley AN (1985) The selective erosion of plant nutrients in runoff. Soil Sci Soc Am J 49:1527–1534. doi:10.2136/ sssaj1985.03615995004900060039x Sharpley AN (1993) Assessing phosphorus bioavailability in agricultural soils and runoff. Fert Res 36:259–272. doi:10. 1007/BF00748704 Sharpley AN, Kleinman PJA (2003) Effect of rainfall simulator and plot scale on overland flow and phosphorus transport. J Environ Qual 32:2172–2179. doi:10.2134/jeq2003.2172 Sharpley A, Wang X (2014) Managing agricultural phosphorus for water quality: lessons from the USA and China. J Environ Sci 26:1770–1782. doi:10.1016/j.jes.2014.06.024 Sharpley AN, Withers PJA (1994) The environmentally-sound management of agricultural phosphorus. Fert Res 39:133–146. doi:10.1007/BF00750912 Sharpley AN, Daniel TC, Sims JT, Pote DH (1996) Determining environmentally sound soil phosphorus levels. J Soil Water Conserv 51:160–166 Sharpley AN, Jarvie HP, Buda A, May L, Spears B, Kleinman PJA (2013) Phosphorus legacy: overcoming the effects of past management practices to mitigate future water quality impairment. J Environ Qual 42:1308–1326. doi:10.2134/ jeq2013.03.0098 Sharpley AN, Bergstro¨m L, Aronsson H, Bechmann M, Bolster C, Bo¨rling K, Djodjic F, Jarvie HP, Schoumans O, Stamm C, Tonderski K, Ulen B, Uusitalo R, Withers P (2015) Future agriculture with minimized phosphorus losses to waters: research needs and direction. Ambio 44:163–179 Sims JT, Kleinman PJA (2005) Managing agricultural phosphorus for environmental protection. In: Sims JT and Sharpley AN (eds) Phosphorus: agriculture and the environment. Agronomy Monograph No. 46. American Society of Agronomy, Madison, Wisconsin, USA, pp 1021–1068 Smith DR, King KW, Johnson L, Francesconi W, Richards P, Baker D, Sharpley AN (2015) Surface runoff and tile drainage transport of phosphorus in the Midwestern United States. J Environ Qual 44:495–502. doi:10.2134/jeq2014. 04.0176 Tandon HLS, Cescas MP, Tyner EH (1968) An acid-free vanadate-molybdate reagent for the determination of total phosphorus in soils. Soil Sci Soc Am J 32:48–51. doi:10. 2136/sssaj1968.03615995003200010012x Truman CC, Gascho GJ, Davis JG, Wauchope RD (1993) Seasonal phosphorus losses in runoff from a coastal plain soil. J Prod Agric 6:507–513. doi:10.2134/jpa1993.0507 Wang YT, Zhang TQ, Hu QC, Tan CS, O’Halloran IP, Drury CF, Reif DK, Ma BL, Ball-Coelho B, Lauzon JD, Reynolds WD, Welacky T (2010) Estimating dissolved reactive phosphorus concentration in surface runoff water from major Ontario soils. J Environ Qual 39:1771–1781. doi:10. 2134/jeq2009.0504 Withers PJA, Hodgkinson RA, Barberis E, Presta M, Hartikainen H, Quinton J, Miller N, Sisak I, Strauss P, Mentler A (2007) An environmental soil test to determine the intrinsic risk of sediment and phosphorus mobilisation

Nutr Cycl Agroecosyst from European soils. Soil Use Manage 23:57–70. doi:10. 1111/j.1475-2743.2007.00117.x Woodward J, Foster IAN (1997) Erosion and suspended sediment transfer in river catchments: environmental controls,

processes and problems. Geography 82:353–376. http:// www.jstor.org/stable/40572952

123