Identifying challenges with current soil sampling, testing and interpretation protocols and developing soil-type specific protocols
Reviewing current soil sampling, testing and interpretation protocols used in Pacific Islands
Background
Addressing soil fertility problems in the Pacific Islands requires effective nutrient management, which requires accurate measurement and interpretation. While this principle is in theory easy to adopt, there are many challenges in developing practical systems of soil sampling, measurement, and diagnosis, especially for complex smallholder systems such as those in the Pacific. Lack of guidelines on soil sampling or outdated soil analysis protocols and the limited knowledge of agricultural research and extension officers to undertake correct soil sampling and analysis need to be assessed and addressed. This research reviewed the soil sampling, testing and interpretation protocols currently used in Pacific Island countries.
Key results
Soil sampling
The survey and consultation outcomes showed that the Fiji Agricultural Chemistry laboratory, the Scientific Research Organisation of Samoa, and University of the South Pacific in Tonga laboratories have their own soil sampling protocols, which resulted in both similarities and differences in soil sampling and analysis.
Similarities included:
- Soil was sampled from the surface horizon at 0-15 cm depth to determine soil fertility for field/root crops.
- A commonly used soil sampling tool is the screw auger, while a spade or knife are only used occasionally.
- Soil samples are oven-dried followed by sieving with a 2mm sieve. However, they do not have any protocol to discard excess soil being collected.
Differences included:
- To make up a composite sample, the Fiji Agricultural Chemistry laboratory collects less than 10 subsamples when sampling. In comparison, the Scientific Research Organisation of Samoa and the University of the South Pacific in Tonga laboratories use more than 10 subsamples.
- Zigzag and grid sampling methods are used by the Fiji Agricultural Chemistry laboratory to collect subsamples in a uniform area, while the Scientific Research Organisation of Samoa and the University of the South Pacific in Tonga laboratories only use the zigzag method.
- To label soil samples for identification, the Fiji Agricultural Chemistry laboratory records only the geolocation/GPS readings without inclusion of details such as sampling depth, field history and crop information, village name and the names of farmer or researcher. In comparison, the Scientific Research Organisation of Samoa and the University of the South Pacific in Tonga laboratories record the information that are excluded by Fiji, but they only occasionally record the geolocation of soil samples.
Soil testing
The three laboratories analysed their soil samples following the methods of Daly and Wainiqolo (1993), a manual written under a Pacific regional network in the 1990s called South Pacific Agricultural Chemistry Laboratory Network. The University of the South Pacific in Tonga consultations identified that none of the laboratories followed the soil analysis methods of Rayment and Lyons (2011). The staff pointed out that procedures in South Pacific Agricultural Chemistry Laboratory Network manual were simply written, which facilitated their use, but there is interest in using the Australasian Soil and Plant Analysis Council methods of Rayment and Lyons (2011) for future soil analysis. However, those procedures can only be adopted if written in simple English that can be easily understood by Pacific Islands laboratory technicians, like the procedures of Daly and Wainiqolo (1993).
Interpretation of results
Staff of the Fiji Agricultural Chemistry and University of the South Pacific laboratories possess knowledge and skills to interpret soil analysis results, however the Scientific Research Organisation of Samoa staff lack this knowledge and skills. Consequently, the Scientific Research Organisation of Samoa have requested for staff training(s) on interpretation. Currently, interpretation of soil analysis results in Pacific Islands are using interpretation values in the manual by Daly and Wainiqolo (1993).
Gaps
It was pointed out during the consultations that Pacific Island laboratories have a renewed interest in reviving the South Pacific Agricultural Chemistry Laboratory Network due to its role in addressing the need for quality assurance and soil analysis in the Pacific region. When South Pacific Agricultural Chemistry Laboratory Network was functional, laboratory technicians in the region underwent trainings to build their capacity on simple procedures in documenting standard methodologies for soil and plant analysis and quality control, strengthening soil and plant exchange network, and sharing of knowledge and skills among the soil laboratories in Pacific Islands. Services mentioned have been identified as existing gaps in the Pacific Islands laboratories consulted in this review.
Based on the outcomes of the survey and consultations, the following recommendations are made:
- There is a need to develop a soil sampling protocol that can be used by Pacific Island laboratories to allow the laboratory technicians, extension workers and farmers to correctly sample, prepare, bag and tag soil samples to achieve uniformity in soil sampling techniques. Further, this will also address the differences in sampling protocols in Pacific Island countries, which causes inconsistency in results.
- An urgent need exists to simplify soil sampling methods, such as those of Rayment and Lyons (2011), by writing them in a format and language that Pacific Island soil laboratory technicians and managers can easily understand.
- The Pacific Soil Laboratory Network was established under the existing Australasian Soil and Plant Analysis Council (ASPAC) in 2019. The inclusion of Pacific Island and Papua New Guinea laboratories in the Australasian Soil and Plant Analysis Council ensures the long-term sustainability of the network, granting laboratories local technical assistance and increasing their ability to participate in Global Soil Laboratory Network activities. The project has promoted the Pacific Soil Laboratory Network. However, extension and outreach are needed to inform all sectors of the existence of the body. Further development is needed to improve laboratory performance:
-
- Enhance the quality of soil and plant analysis of the region’s participating laboratories with specific objectives of developing quality assurance programmes.
- Document standard methodologies for soil and plant analysis, and quality control.
- Organise training of technicians of Pacific Soil Laboratories, where and when required.
- Strengthen soil and plant exchange network if the Australasian Soil and Plant Analysis Council does not cover this area. At present Australasian Soil and Plant Analysis Council is sending soil and plant samples to the Pacific Soil laboratories.
- Develop links with a laboratory in a metropolitan country to support the quality assurance programmes.
- Identify laboratories within the Pacific region to serve as centres for the provision of regional analytical services.
- Coordinate laboratory upgrades and capacity building with government, aid schemes and non-governmental organisations.
Click here to see the review presentation.
Utilising infrared technology to make rapid assessment of calcareous and volcanic soil fertility
Background
Workflow for soil property estimation using mid-infrared spectroscopy.
The use of mid-infrared spectroscopy under laboratory conditions to rapidly and cost-effectively characterise soil properties of agronomic importance is well-established. Mid-infrared spectroscopy has been demonstrated to be used for the simultaneous estimation of soil physical, chemical and biological properties including particle size – clay and sand content – soil mineralogy, soil organic carbon, total carbon, total nitrogen, soil pH, electrical conductivity, cation exchange capacity, exchangeable cations such as calcium, magnesium, and potassium and available phosphorus. Mid-infrared spectroscopy has also been used with success for determining soil chemical properties of Columbian and Hawaiian variable charge soils.
Previous ACIAR-funded research in the Philippines (SMCN/2009/031) and Myanmar (SMCN/2014/075) demonstrated that adoption of mid-infrared spectroscopy can be achieved through standard protocol development and training. Building on this work, our aim was to introduce this new rapid soil measurement technique to the Pacific Island Countries and Territories through building infrared spectroscopic capacity at the Agricultural Chemistry Laboratory of the Ministry of Agriculture in Fiji, based at Koronivia research station.
The aims of this research were to:
- Introduce mid-infrared spectroscopy soil sample preparations and measurements, and soil spectral inference, and quality control analysis protocols.
- Test mid-infrared spectroscopy for estimating soil properties of agronomic importance for calcareous and volcanic soils of the Pacific Islands.
- Build pilot soil reference spectral libraries to form the basis for calibration model generation to rapidly estimate soil fertility using mid-infrared technology in the Pacific Islands.
Key results
The CSIRO project team trained Fiji Agriculture Chemistry Laboratory staff in the use and operation of mid-infrared technology to facilitate uptake of the technology as a valuable tool for rapid and cost-effective assessment of soil.
Capacity building
- Purchased a Bruker Alpha-II-FT-IR (mid-infrared) instrument and installed it at the Fiji Agriculture Chemistry Laboratory.
- Delivered a “Mid-infrared spectroscopy” 5-day (5-11 May 2018) training course , which covered soil survey and sampling methodology and a meta-data collection introduction. The course was attended by 11 Fiji Agriculture Chemistry Laboratory staff.
- Introduced laboratory protocols including soil sample preparation, soil scanning protocols, quadruplicate mid-infrared scanning.
- Developed comprehensive step-by-step training materials for soil spectral inference utilising the Bruker Opus software:
- Calibration model generation for soil property prediction.
- Soil property estimation from mid-infrared spectra and calibration.
- Quality control and averaging of quadruplicate soil property predictions using a custom build Excel Macro.
-
A summary of soil samples analysed with mid-infrared at Fiji Agriculture Chemistry Laboratory.
The Fiji Agriculture Chemistry Laboratory team:
- Independently setup mid-infrared operations and successfully conducted mid-infrared soil spectral inference and quality control.
- Adopted mid-infrared soil analysis in the new National Soil Health Card program to build spectral libraries for soils in Fiji.
- Analysed soil samples with mid-infrared and traditional wet chemistry analysis methods, and built spectral reference libraries for Fijian (regional) and Tongan (local) soils of agronomic importance.
- Used mid-infrared scanned soil samples from the ACIAR field trials in Taveuni, Fiji and predicted soil chemical properties from mid-infrared spectra and calibration models developed from Fiji Agriculture Chemistry Laboratory soil archive samples of Taveuni.
- Mid-infrared scanned soil samples from the ACIAR field trials in Tonga.
Mid-infrared spectroscopy calibration results
Regional calibration Fijian islands
Local calibration Tongan islands
- Results showed that mid-infrared spectroscopy can be used successfully to estimate a range of soil chemical properties important for soil fertility assessments for allophanic soils in Fiji and Tonga.
- Results showed that the predictive power of calibration models varies depending on the dataset used for calibration model building.
- Results showed that there is a noticeable difference in relation to the geographic extent of the datasets (local or regional) and distribution of soil chemical properties of the datasets used for calibration building.
- Results are consistent with the general rule that local calibrations tend to give more accurate estimates than regional or national calibrations.
- Some datasets considered in this project were not suitable for calibration model building. Continued building of soil spectral libraries of representative soil types in the Pacific Islands together with laboratory standards and protocols will improve predictive power of calibration models.
During this project a regional survey was conducted by staff from the Tonga Ministry of Agriculture and Food, Forests and Fisheries and a Monash University student during which historic soil sampling sites across Tongatapu island, Tonga, with legacy soil information were revisited and re-sampled together with their geolocation. Soil samples were sent to Australia and analysed by the CSIRO team through various spectroscopic techniques and subsamples were also analysed for a range of soil properties in the laboratory. Click here to see rapid soil analysis results using visible and near-infrared spectroscopy: Utilising rapid spectral techniques to assess the impacts of agriculture on soil function in pacific soils – An example from Tongatapu island, Tonga.
Community impacts
- “The purchase and installation of the new MIR [mid-infrared] instrument at FACL [Fiji Agriculture Chemistry Laboratory] MOA [Ministry of Agriculture] by ACIAR provides Fiji and the PICTs [Pacific Island Countries and Territories] enhanced capability to rapidly and economically analyse soil samples.” Mr Ami Sharma, Ministry of Agriculture
- “MIR [mid-infrared] is a most valuable tool to overcome challenges such as timely throughput of large numbers of soil samples, tight resources and budgets, acquiring large volumes of organic solvents for traditional wet chemistry analysis, which might not be locally available.” Mr Ami Sharma, Ministry of Agriculture
Click here to see the mid-infrared spectroscopy presentation.
References:
Daly, B. K., Wainiqolo, J. L. (1993). Methods of Analysis for Agricultural Samples: Soil, Plant, Animal Feed and Water. Fiji Agricultural Chemistry Laboratory Technical Report 03/93. 115p.
Rayment, G. E., Lyons, D. J. (2011). Soil Chemical Methods – Australasia, Australia: CSIRO Publishing. pp. 100-103.