Content uploaded by Jean-Luc Chotte
Author content
All content in this area was uploaded by Jean-Luc Chotte on Dec 25, 2016
Content may be subject to copyright.
A preview of the PDF is not available
The fate of fertilizer N in field studies in the volcanic Lesser Antilles with 15N-urea
Abstract
In the Lesser Antilles, recent replacement of old rural systems by intensive market gardening (MG) has led to a decrease of the soil organic matter (OM) content. A field experiment was designed on different types of soils (vertisol, ferrallitic soils, and andisol), to determine the effect on soil OM content of previous land use, either fallow (F) and pasture (P) or MG and banana plantations (B). The fate of 15N-urea applied to a maize crop was studied under the different combinations of soil type and previous land use history. Plant productivity of maize crops reflected N uptake, but not the N levels in soil OM. In the Lesser Antilles, soil OM does not limit plant productivity. However, immobilization of fertilizer N in soil (and thus a reduction of N losses) was positively linked to soil OM status. Losses, except for the andisol, were less than 30% of applied urea N, an observation which favours the use of urea in this humid tropical area.
... Meireles et al. (1980) observed that only 1.35% of the N (in the form of ammonium sulfate) applied to an Oxic Paleudalf planted with bean (Phaseolus vulgaris) leached in the year following application. In contrast, in a field experiment conducted by Chotte et al. (1998) in the Lesser Antilles, 40 to 45% of the N applied as urea to corn crop plots was lost during the growing season. For an Alfisol from Venezuela, where yearly rainfall is 1700 mm, Hetier et al. (1989) found that only 2% of fertilizer N was lost by leaching, whereas total N losses amounted to 30%. ...
... The value of the C:N ratio in tropical soils was evaluated using published data from 13 studies. Data of organic C and total nitrogen from Brazil (Araujo et al., 2001;D'Andrea et al., 2004;Sampaio et al., 1990;Van Wambeke, 2003), Dominica, Guadeloupe and Santa Lucia (Chotte et al., 1998), Malaysia (Mubarak et al., 2001), ...
Nonpoint source (NPS) water pollution originates from a broad range of human activities for which the pollutants have no obvious point of entry into receiving waters. Agricultural activities are a major source of NPS pollution and the release of sediments, pesticides, nutrients, and microorganisms can contribute to water quality deterioration. NPS pollution is a major environmental problem in developed countries, and has become a problem of even greater concern in tropical developing countries as areas used for subsistence or low intensity agriculture have transitioned to high technology agriculture with increased use of fertilizers and pesticides.
The tropics comprise about 36% of the earth surface and its characteristics differ from temperate areas in a number of aspects. These differences must be accounted for when practices developed in temperate areas to control NPS pollution are applied in the tropics. Modeling is an important tool used to evaluate the effectiveness of NPS pollution control measures and NPS models are commonly used in developed countries for environmental evaluation and management. Although used widely in temperate zones, the use of NPS models is limited in the tropics and usually involves the application of models developed for temperate conditions with only minor--if any--adaptation to tropical conditions.
The overall objective of this research was to develop a field-scale agro-ecosystem model for humid tropical conditions that can be used as a NPS pollution management tool. The specific focus of the present research is on the fate and transport of nitrogen and phosphorus and prediction of losses in surface runoff and leachate from the root zone. Existing “temperate” NPS models were evaluated and the GLEAMS model was selected for modification. Changes to GLEAMS include incorporation of a new potential evapotranspiration model, changes in initial and default parameter values, and different relationships for selected nitrogen and phosphorus transformations.
The FAO Penman-Monteith model was incorporated for predicting potential evapotranspiration (ET) to overcome weaknesses found in the GLEAMS ET models. More appropriate initial or default values were determined from the literature for C/N ratio, soil nitrate and ammonium, soil P sorption (PSP), and the equation to calculate the initial concentration of soil organic P. Models for carbon, nitrogen, and phosphorus mineralization, nitrification, nitrogen immobilization, and the transformation of labile P to active mineral P as well as models that simulate the effect of temperature in C, N, and P mineralization and nitrification were modified according to equations derived in the tropics. A nitrate retardation factor (Ncrit) and pH were added as input variables. Ncrit is a critical NO3-N concentration and downward and upward nitrate movement in soil occurs only when its concentration is above Ncrit. pH is used in the simulation of nitrification.
TROPGLEAMS was evaluated through comparison with data from field studies in the tropics, through comparison with the original GLEAMS model, and through sensitivity analysis. Model validation included the application of the GLEAMS and TROPGLEAMS to three areas and the comparison of measured and simulated values of selected output variables. Measured values were obtained from three studies. The first study, the Piracicaba lysimeter study, was carried out in Piracicaba, SP, Brazil, using lysimeters planted with sugarcane in which 8 treatments with different doses of mineral N and crop residue were applied. Water in runoff and mineral N in leachate were the assessed output variables. The second study, the Piracicaba plot study, was also located in Piracicaba and included plots planted with sugarcane with four treatments (three doses of sewage sludge and one dose of mineral N). For this study, the assessed output variables were NH4-N and NO3-N in soil solution and TKN and mineral nitrogen (NH4-N + NO3-N) in soil. The third study, the Lages plot study, was carried out in Lages, SC, Brazil and included plots planted with a wheat-soybean rotation with three different soil management systems: Conventional tillage, Chisel+disk, and No-tillage. The assessed output variables were NH4-N, NO3-N, and PO4-P in runoff. The comparison of results of the original and modified models was performed through the analysis of graphics with model results and the comparison of root means square error (RMSE) between measured values and that simulated by both models. The sensitivity analysis included the comparison of the sensitivity of GLEAMS and TROPGLEAMS to changes in temperature for the variables NO3-N and NH4-N in leachate and in runoff, and NH4-N in sediment and the sensitivity of TROPGLEAMS to pH and the retardation factor of nitrate (Ncrit). NO3-N in leachate and in runoff were the variable assessed for pH and Ncrit.
Results of actual evapotranspiration from TROPGLEAMS were more stable than that from GLEAMS resulting in a better prediction of actual ET. TROPGLEAMS is very stable in simulating nitrogen and phosphorus kinetics and the influence of environmental factors (temperature, pH, and soil water content) in the transformations between their several pools. In terms of mass balance of nutrients, TROPGLEAMS showed to be as accurate as GLEAMS, since, for three simulations of three years, a balance of zero for all years was obtained, except the balance of nitrogen in one year, in which both models resulted in negative values, suggesting that the error is in a subroutine common to both models that was not modified in this study. For the Piracicaba lysimeter study, while GLEAMS overpredicted the rates of nitrogen transformations, resulting in a significant overprediction of leached N, TROPGLEAMS results were very close to measured values. For the Piracicaba plot study, TROPGLEAMS was superior to GLEAMS in simulating ammonia and nitrate concentration in soil solution, and mineral nitrogen and TKN concentration in soil. While nitrate and ammonia concentration simulated with GLEAMS resulted in zero values in most of the simulation, TROPGLEAMS resulted in positive values of these variables during most simulations. For NH4-N concentrations in soil solution, RMSE for TROPGLEAMS was smaller than that for GLEAMS. For NO3-N concentration in soil solution, RMSE for TROPGLEAMS were higher than that for GLEAMS, what does not indicate a better performance, since GLEAMS results were zero in almost all simulation.
For the Lages plot study, both GLEAMS and TROPLEAMS highly underpredicted NO3-N in runoff. However, TROPGLEAMS predicted NH4-N and PO4-P in runoff better than GLEAMS.
TROPGLEAMS simulated better than GLEAMS the effect of tillage on losses of NH4-N in runoff in the Lages plot study. Since the incorporation of surface material is more complete in the Conventional tillage treatment, while no incorporation and intermediate incorporation is made in No-tillage and Chisel+disk treatments, the expected tendency is more superficial losses in plots with No-tillage, followed by Chisel+disk and Conventional tillage. The outcome of TROPGLEAMS followed this tendency for average NH4-N in runoff, while GLEAMS simulated greater values for the Chisel+disk treatment, followed by Conventional tillage and No-tillage.
For the range of annual temperature between 16.12 to 21.2oC, TROPGLEAMS is more sensitive than GLEAMS for NO3-N in leachate and in runoff and NH4-N in runoff and in leachate, while GLEAMS is more sensitive than TROPGLEAMS for NH4-N in sediment. For PO4-P in runoff, both models showed the same absolute value of relative sensitivity, while for PO4-P in sediment, sensitivity of TROPGLEAMS was almost three times that of GLEAMS.
It can be concluded that GLEAMS was a temperate NPS model very appropriate to be adapted in this study and the equations and models identified as appropriate to be used in the adaptation, along with the initial or default values considered more representative to tropical soils resulted in a NPS model more accurate than the original model in the simulation of transport and losses of N and P under tropical conditions. Overall, TROPGLEAMS simulate losses in leachate better than simulate losses in runoff and changes in the algorithms related to losses in runoff need to be made to improve model performance. The introduction of Ncrit and pH improved the accuracy of the simulation of nitrification and nitrate movement in tropical soils and the model is reasonably sensitive to them.
TROPGLEAMS is the first field NPS model adapted for simulating fate and transport of nitrogen and P under humid tropical conditions. Future works need to be done in order to improve model predictions, between them it can be cited the adaptation of intern database to represent more accurately different crops and varieties usually planted in the tropics; the inclusion of a model to simulate rainfall interception in crop fields; the assessment of the soil temperature model, of the equations that constraints ammonia and labile phosphorus available for plant uptake and leaching, of the model related to the transformation between labile P (PLAB) and active mineral P (PMINP), and of the rate of soil organic matter decomposition; and the improvement of the models related to the processes involving nitrate and phosphorus losses in runoff.
... Concerning nitrogen (N) inputs, banana crops are locally and intensively fertilized at the base of the pseudostem with up to 500 kg N ha −1 yr −1 . For Caribbean conditions, Chotte et al. (1998) and Raphael (2006) found that N leaching may represent as much as one third of the added N fertilizer because of high rainfall intensity in this region. ...
... Concerning nitrogen (N) inputs, banana crops are locally and intensively fertilized at the base of the pseudostem with up to 500 kg N ha −1 yr −1 . For Caribbean conditions, Chotte et al. (1998) and Raphael (2006) found that N leaching may represent as much as one third of the added N fertilizer because of high rainfall intensity in this region. ...
Banana (Musa sp.) residues contain two thirds of the plant nitrogen (N) content at harvest which may represent a large N source for the daughter plant and contribute to reducing fertilizer rates. Our aim was to assess the decomposition kinetics of above- and below-ground banana residues, and to determine the uptake of the released N by the daughter plant. Residue decomposition was analyzed under laboratory and field conditions, and plant N uptake was evaluated in the field using 15N-labelled above-ground residues. The decomposition kinetics was interpreted with a simulation model. In the laboratory, residue decomposition showed an immobilization phase followed by remineralization. The immobilization phase was longer and remineralization was smaller for roots due to their higher C:N ratio and lignin content. The model satisfactorily described both phases and the effect of the residue C:N ratio. Experimental and model results indicated that root residues would be a minor N source for the daughter plant. In the field, above-ground residues decomposed according to first order kinetics. Residue half-life was 2d in the laboratory and 32d in the field, which was due to poorer soil–residue contact in the field. At the time of harvest of the daughter plant, 39% of residue N was recovered in the plant, 54% in the topsoil, and 3% in the remaining residues. About 4% of residue N was probably lost by leaching. Nitrogen derived from residues represented 19% (14kgNha−1) of N in bunches of the daughter plant and 18% (39kgha−1) of the whole plant N. Our study showed that above-ground residue N was an effective N source for the daughter plant, with a good overlap between the period of N release and that of greatest plant demand which reduced the risk of N leaching.
... Thus information about the abundance, diversity and activities of different biotic groups responsible for nitrogen mineralisation is of crucial importance in the management of soil productivity. Soil organic matter could represent the main source of inorganic nitrogen, even in the presence of fertilizer (Guiraud, 1984;Chotte et al., 1998). About 30% of the total inorganic nitrogen mineralised from soil organic matter is a result of microbial consumption by the soil fauna (Verhoef and Brussaard, 1990). ...
The effect of several bacterial-feeding nematodes of the Cephalobidae family (Zeldia punctata, Acrobeloides nanus and Cephalobus pseudoparvus) on the microbial community of a Sahelian soil (Senegal) was investigated in microcosm. The consequences of the activity of these nematodes on the growth and nitrogen nutrition of young maize plants (aerial biomass, root biomass and nitrogen content) were also estimated. Laboratory-cultured nematodes were inoculated into soil containing maize seedlings where the natural nematofauna had been previously eliminated by alternately freezing and defrosting (five cycles). The microbial compartment of the soil community was characterised through total microbial biomass (using fumigation-extraction), density of bacteria (using colony forming units counts), microbial activity (using alkaline phosphatase) and genetic structure of soil microbial community (using denaturing gradient gel electrophoresis) at sowing and at 12, 26 and 47 days after planting. Final nematode densities in the different treatments (between 4 and 20 Ind g−1 dry soil) demonstrated a high level of reproduction. The different types of nematodes tested induced similar trends in changes in the microbial pool of the soil and in maize seedling growth. Compared to control soils, the presence of nematodes led to an increase (+12%) in plant biomass and reduced concentrations of soil ammonium but had no effect on concentrations of nitrate by the end of the experiment. Sixty-three percent of the inorganic nitrogen initially present in the soil was incorporated into the maize plants with nematodes whereas only 47% was incorporated without nematodes. Nematode activity led to a significant decrease in microbial biomass (−28%) and density of cultivable bacteria (−55%), however, nematodes stimulated bacterial activity (+18%). The effects of Z. punctata were weakest compared to A. nanus and C. pseudoparvus. The presence of nematodes modified the genetic structure of the microbial community essentially by changing the relative abundance of dominant bacterial populations. Among nematode species tested, A. nanus modified the structure of the microbial communities the most compared with control soils without nematodes. Overall, results from this study provide evidence for the ability of microbial feeding nematodes to alter microbial activity, microbial community structure, nitrogen mineralisation and growth of maize seedlings in a Sahelian soil from Senegal, West Africa.
Nonpoint-source (NPS) or diffuse pollution is a major environmental problem in developed countries, and modeling is an important tool used to evaluate the effectiveness of pollution control measures. The use of NPS models in the tropics usually involves the application of models developed for temperate regions with little, if any, adaptation to tropical conditions. In this article, we provide a synthesis of the literature values from studies in the tropics, using the GLEAMS model as a reference for the comparable values used in representing temperate conditions. We found that values for the carbon to nitrogen (C:N) ratio, potentially mineralizable nitrogen to total nitrogen ratio (N 0/N total), and base NO 3-N and NH 4-N concentrations representative of tropical soils were all different from the values considered appropriate for temperate soils. Relationships between phosphorus pools in tropical soils and in phosphorus sorption parameters likewise were different from those used in GLEAMS, with the exception that the GLEAMS ratio between labile and organic phosphorus in highly weathered soils was found to be comparable to data specific for tropical soils. © 2009 American Society of Agricultural and Biological Engineers.
Incorporation of crop residues after harvest may improve nitrogen (N) retention of tropical soils. A maize‐groundnut rotation experiment was established in an Ultisol to study the recovery by the crop‐soil system of N from mineral fertilizer and from the first maize residue. Out of the recommended N fertilizer for maize (150 kg N ha), 60 kg N ha was applied in the form of ammonium sulphate (9.82 atom percent N excess) to the first maize crop and traced (for four crops) in the subsequent cropping system, with or without incorporation of crop residues. Similarly, N in the labelled residues of the first crop (maize) was also traced in the soil‐crop system with incorporation of unlabelled crop residues. Recovery of fertilizer N by the first crop ranged from 20 to 22%. In the subsequent four crops, this value averaged 2% (but ranged from 0.4 to 6%). The proportion of fertilizer N retained in the soil (0–50 cm) after harvest of the first crop was 35 to 44% whereas after harvest of the subsequent crops, this proportion was 30 (but ranged from 24 to 34%) in crop residue treated plots and 22% (but ranged from 18 to 27%) in plots where crop residues were removed. Maize residue‐N recovery in the subsequent three crops averaged only 5% (but ranged from 1 to 11%), being highest in the crop after crop residue incorporation. The corresponding recovery in the soil (0–50 cm) averaged across crops was 46% (but ranged from 40 to 52%). In tropical regions with high rainfall, conserving and decreasing soil N losses from mineral fertilizers could be achieved (in the long term) by the combined application of crop residues with mineral fertilizer.
The aim of this study was to examine the effects of amendments with leaf biomass on the development of tomato plants in a soil where root-knot nematodes (Meloidogyne mayaguensis) and/or a nematophagous fungus (Arthrobotrys oligospora, strain ORS 18697) had been inoculated. Six origins of leaf biomass were chosen: Acacia mangium, Acacia holosericea, Eucalyptus camaldulensis, Casuarina equisetifolia, Azadirachta indica and Sorghum vulgare. These leaf biomass types inhibited the development of the aerial parts of the tomato plants. This negative effect was not observed when the fungus was inoculated. On the contrary, plant growth was stimulated. Moreover, the antagonistic activity of Arthrobotrys oligospora was strengthened in the presence of ground leaf powder, especially that from Acacia holosericea. The effects of phenolic compounds on fungal growth and predatory activity and on plant growth are discussed.
Field studies were conducted on Indian Vertisol to determine the fate of15N-labeled fertilizers applied to dryland sorghum in two successive rainy seasons. In the 1981 season, a split-band (SB) urea application of 74 kg N/ha, half amounts placed 5 cm deep and 8 cm from opposite sides of plant rows at 4 and 19 days after emergence, was superior to preemergent applications of either surface-applied (S) or incorporated (I) applications at the same rate; 907 mm of rainfall fell during the sorghum growing period. Percentages of applied N recovered in the soil-plant system after the sorghum harvest were 94%, 74%, and 72%, respectively, for the SB, I, and S application methods. Substantial quantities, 39%, 45%, and 42% of the added N for the SB, I, and S tretments, respectively, remained in the soil after the final harvest. Plant utilization of added urea-N was greater in 1980 when rainfall during the growing season was 212 mm less than in 1981. S or I applications of urea at 74 kg N/ha, with above-ground plant15N recoveries of 48.0% and 48.6%, respectively, were also equally as efficient during 1980. Residual soil N derived from fertilizer was of little value for a sorghum crop in the following rainy season and for a safflower crop in the post-rainy season in a double-cropping system.
La présente étude a pour objectif l'étude quantitative du cycle de l'azote dans les systèmes de cultures actuels, arachide-mil et soja-maïs de la zone tropicale sèche ouest africaine. L'auteur fait l'hypothèse que la maîtrise du bilan azoté permettra de dégager des solutions économiquement acceptables. Cette étude est fondée sur des expérimentations au champ et le recours systématique à la méthodologie isotopique, permettant ainsi de quantifier avec certitude des processus connus presque exclusivement sur le plan des mécanismes. Il s'agit en tout premier lieu de la fixation de N2 sur laquelle les informations quantitatives étaient jusqu'à présent absentes en Afrique de l'Ouest. Il s'agit aussi des processus d'immobilisation dans les sols (par voie microbienne et par voie végétale, c'est-à-dire par incorporation dans les résidus de récolte). Les pertes d'azote ont pu être évaluées dans tous les cas mais de façon globale sans qu'il ait été possible de déterminer la part de la volatilisation ou de la dénitrification. Le processus de volatilisation a cependant été étudié en milieu contrôlé. Dans les sols sableux tropicaux de la zone soudanienne et dans les systèmes de culture actuels, le coefficient réel de d'utilisation de l'engrais azoté par la céréale est beaucoup plus faible qu'on ne pouvait l'imaginer puisque la plante utilise seulement 25 % (mil) et 35 % (maïs) de l'azote engrais qu'on lui a apporté. L'apport d'engrais azoté au sol accroît l'absorption par la plante des réserves en azote du sol (désignées par l'auteur sous le terme de "pool d'azote mobilisable"), probablement grâce à l'extension du système racinaire et à un "prime effect". Une fraction importante (de l'ordre de 30 à 50 %) de l'engrais apporté au sol (beaucoup plus importante qu'on ne pouvait le supposer a priori dans les sols sableux étudiés), est immobilisée par voie biologique dans les sols. La volatilisation de l'azote - étudiée en conditions simulant celles du champ - peut représenter jusqu'à 40 % de l'urée apportée en surface (en sols sableux) selon la technique actuellement vulgarisée. La fixation de N2 par l'arachide ou le soja représente de 20 à 80 % de l'azote absorbé par les cultures, les variations dans l'intensité du processus pouvant être considérables. L'incidence des facteurs qui contrôlent cette fixation a été étudiée. Il s'agit des facteurs environnementaux (rhizobium spécifique, endomycorhizes, sécheresse, azote minéral du sol et fertilité du sol) et des facteurs "pratiques culturales" (variétés, engrais-N, engraisP, fumier et chaux). Ces résultats ont conduit l'auteur à proposer des méthodes permettant d'améliorer la nutrition azotée des cultures par une maîtrise du bilan azoté au niveau du système de culture : la validité de ces différentes méthodes, dont les principes sont décrits en détail, a été vérifiée in situ. En particulier, il se confirme que la légumineuse est la clé du maintien du bilan azoté des sols sableux tropicaux en culture pluviale semi-intensive.
Nitrogen behaviour was studied in different types of soil with different management conditions. The quantity of organic matter was smaller in the continuously cropped soils than in those with perennial vegetation. A 15N-labelled fertiliser was applied in the field to assess losses, uptake by plants and storage in the soil. No effect was observed in soils with organic carbon contents greater than 15%. However, soils with higher organic matter contents stored more nitrogen from the fertiliser and losses were greater from the soils with a low organic matter content. Management of soil organic matter is necessary for the conservation of making up of stocks. -from English summary
