Evaluation of the phosphorus status of sugarcane soils in Mauritius using agronomic and environmental criteria

[摘要] Phosphorus input is vital to the maintenance of profitable sugarcane crop production in Mauritius. The intensive use of some 5,000 tonnes of P annually during the past 50 years is believed to have built up the P status of the sugarcane soils, perhaps even to excessive levels. While this accumulation of P is desirable from an agronomic perspective, there is growing concern in Mauritius about its possible effect on surface water quality. In response to that concern, a study was initiated with the following specific objectives: i. To review the usage of P fertilisers in sugarcane production in Mauritius and assess their resulting impact on the P status of the main soil groups under sugarcane. ii. To enlarge the scope of the current method used (0.1M H2SO4 extraction) for agronomic P testing so that it also indicates environmental status of sugarcane soils in Mauritius. iii. To determine the environmental threshold P in soils above which the P will represent a hazard to surface waters. The five yearly averages of fertiliser P usage by the Mauritian sugarcane industry showed that from the 790 tonnes of P2O5 (mainly as rock/guano phosphates) consumed at the beginning of the 20th century, P usage attained a peak of 5,675 tonnes in the 1970s before declining thereafter as a result of a decreasing land area under sugarcane. During the period 2005 to 2008, an average of 3,350 tonnes of P2O5 mainly as ammonium phosphates were applied annually to sugarcane which is cultivated in Mauritius mainly on five soil groups, namely the Low Humic Latosol (Humic Nitosol)*, the Humic Latosol (Humic Nitosol)*, the Humic Ferruginuous Latosol (Humic Acrisol)*, the Latosolic Reddish Prairie (Eutric Cambisol)* and the Latosolic Brown Forest (Dystric Cambisol)*. A method based on 0.1M H2SO4 as extractant is currently used as a routine soil test to assess P available to sugarcane in the soils of Mauritius. On the basis of soil P test values, four soil P fertility classes could be discerned, namely: Examination of the soil test P data obtained in 1997/1998 showed that 48% of the land still required P fertilisation while approximately 40% had an excess of P (P �?100 mg kg-1). Less than 10% of the soils had an optimum soil P (80 �?P < 100 mg kg-1). Moreover, soils with a highly excessive soil P status (P �?150 mg kg-1) rose from 23% in 1997/1998 to 34% in 2005/2006 indicating that with the current P management practice in sugarcane, the P status of soils in Mauritius will shift more and more towards an excess of P. In spite of the extensive information available on the soil P status, its significance from the freshwater protection angle was, prior to this study unknown due mainly to a lack of a suitable environmental soil P test method. From this perspective, as a laboratory extraction of soil with 0.01M CaCl2 gives a very reliable representation of the P in runoff, the P extractable in a 0.01M CaCl2 (0.01M CaCl2-P) solution was determined in 112 soil samples representing the five main soil groups under sugarcane. The soil samples whose characteristics of pH, organic matter content, exchangeable bases and cation exchange capacity were also determined, were selected to cover a range of 10 to 250 mg kg-1 P extractable by the 0.1M H2SO4 used for agronomic soil P testing in Mauritius. As the environmental soil test P must be independent of soil properties and the concept of degree of P saturation (DPS) meets that criteria, the ammonium oxalate DPS (DPSox) was determined in the 112 soil samples to provide a reliable pointer of P susceptibility to loss from soils. Since it is very unlikely that ammonium oxalate extraction would be used as a routine soil test, the relationship between DPSox and 0.1M H2SO4-P was established by conventional statistical regression techniques. The results obtained indicate that no single soil characteristic could be said to have a distinct influence on the amount of P extracted by either the 0.1M H2SO4 or the 0.01M CaCl2 or by the DPSox. Indeed the correlation (r2) between the 0.1M H2SO4-P, 0.01M CaCl2- P, DPSox with the individual measured soil characteristics was low and never exceeded 0.28 in the case of 0.1M H2SO4-P and 0.52 with 0.01M CaCl2-P. The DPSox exhibited the poorest relationship with the soil properties with none of the r2 values being above 0.16. Instead the low r2 values observed indicated as confirmed by multiple regression analysis that the amount of P extracted by each reagent would be the result of the combined effects of certain soil characteristics. The results moreover showed that for soil P not to constitute a hazard to the freshwaters in Mauritius, the DPSox should not exceed 3.10±0.10% and the 0.01M CaCl2-P must lie below 18±1µg L-1. Moreover the linear fit regression equation 0.1M H2SO4-P = 17.3 + 23.2 DPSox with r2 = 0.54 was found to most appropriately describe the relationship between 0.1M H2SO4-P and DPSox. From that equation the threshold DPSox of 3.10±0.10% would correspond to a range of 85 to 95 mg kg-1 of 0.1M H2SO4-P which is henceforth considered as the threshold range of P in sugarcane soils in Mauritius above which the soil P would become a hazard to freshwater sources. Using this environmental threshold range of soil P values as basis, the soils can be divided into the following four environmental classes namely: [table]. Application of the above criteria showed that in 1997/1998, 58% of the soils did not represent any hazard to freshwater quality in Mauritius. As much as 42% of the sugarcane fields in 1997/1998 had from the environmental viewpoint unacceptably high levels of P (P �?95 mg kg-1) in the soils. After one crop cycle in 2005/2006, the number of fields with unacceptably high levels of P (P �?95 mg kg-1) had risen to 53%. The majority (74%) of the sugarcane fields with an environmentally unacceptable P status were located in the Latosolic Reddish Prairie and Latosolic Brown Forest soils. In extending the scope of the current agronomic soil test P using 0.1M H2SO4 as an extractant into an agro-environmental soil P test, this study demonstrated clearly that the agronomic objectives in P management for sugarcane production in Mauritius are incompatible with the environmental aims of protecting the freshwater resources in Mauritius. With the agronomic threshold range of 80 to 100 mg kg-1 P overlapping the environmental range of 85 to 95 mg kg-1 P, soils in Mauritius that are agronomically suitable for sugarcane cultivation are on contrary unsafe from the environment protection viewpoint and vice versa.

[发布机构] University of the Free State