Impact case study database

This Impact Case Study focuses on Ukraine, which, according to the latest data from the UN Food and Agriculture Organisation (Statistics Division), is the world’s largest producer of sunflower seeds, and in the top six producers of potatoes, wheat, barley and corn globally. In spring 2019, Ukraine overtook China to become the third largest exporter of agricultural products to the EU after Brazil and the USA. Translation of research results [R1-6] has resulted in a better-informed practices around using fertilisers and pesticides to protect crops against pests, and in the development of natural, environmentally-safe crop biostimulants. This has led to the increase in crop yields by 7-12% with an associated reduction in expenses for crop protection of 15-20% in the Prydniprovsky region of Ukraine, which contains 20% of the total Ukrainian sown area and produces 8% of the world’s sunflower. Natural microbial biostimulants developed through translation of theoretical results have already proved effective in increasing crop yields in major cereals, such as wheat, barley and sunflowers by 11-21%, 9-18% and 19-63%, respectively, while for vegetables, the yields for tomatoes, potatoes and cucumbers have increased across the country by 20-66%, 10-12%, and 23-26%.

Changing farming practices to improve crop viability and reduce pest control costs

In 2017, Blyuss visited Ukraine and signed a Memorandum of Understanding between the University of Sussex and the Prydniprovsky Scientific Centre of the National Academy of Sciences of Ukraine (PSC). In collaboration with colleagues from the Dnipro State Agrarian and Economic University and the PSC, during 2017-2018 Blyuss translated theoretical results from R1-3 into practical recommendations for farmers on optimal strategies of using nutrients, pesticides and microbial biostimulants for protecting crops against pests, based on monitoring the levels of infection at crop plantations. These practical recommendations, focused on crop pests native to the steppe region, were subsequently distributed to farming companies and agribusinesses in the Prydniprovsky region of Ukraine, with 40% of farmers in the region adopting these recommendations into their farming practices. In terms of reach, the Prydniprovsky region contains almost 20% of the total Ukrainian sown area of agricultural crops and occupies a third of the country’s steppe. It has an abundance of fertile chernozem soil, and, according to latest data from the Ukrainian National Statistics Service, produces 20.2% of the total Ukrainian harvest of wheat, 21.8% of barley, 13.3% of vegetables, and 26.3% of sunflower [S1, link 3]. In fact, in 2019 that region alone produced 8% of the total amount of sunflower produced worldwide.

Professor Anatolii Bulat, President of the Prydniprovsky Scientific Centre, has noted:

“Currently, monitoring and control of the processes of growth and development of agricultural crops on farming lands of various agribusinesses in the Prydniprovsky region are performed in accordance with the results of scientific research and technical recommendations developed by Dr K.B. Blyuss…. Through close collaboration with farmers and agribusinesses, … currently around 40% of farmers in Prydniprovsky region are already using above-mentioned technical recommendations. This has allowed farmers and agribusinesses to increase the survivability of crops by 7-12%, to improve quality and yields of cereal crops, and to reduce expenses associated with control of crop pests by 15-20%, compared to 2015-2017 when these techniques were not used. Reductions in expenses were associated with reduced fuel and personnel costs, as well as a substantial reduction in the amount of nutrients, fertiliser and pesticides being used. The latter factor is a particularly important achievement of the newly developed technical recommendations from the perspective of reduction and minimisation of the negative impact of pesticides on environment and biodiversity.” [S1].

With 18% of the working population of Ukraine being involved in agriculture, new technical recommendations on monitoring of crop infestation and the use of pesticides have had a direct impact on performance of local agribusiness in the Prydniprovsky region.

Mr N. Nodzrin, Chief Agronomist of the company “Pisarevsky Oleksandr Stepanovych” based in Vodyane, has said: “As a result of applying practical recommendations regarding monitoring the levels of crop infestation with pests, as well as optimal use of fertilizers and biostimulants, that have been developed by Dr K.B. Blyuss together with colleagues, our expenses associated with control of crop pests have reduced by 12.5-18.0%, while the viability of wheat and barley plants has increased by 7.5-10%.” [S2].

Similarly, Mr Viktor Lisnyi, the Director of “Lisnyi Viktor Mykolayovych” company, specializing in producing cereal grains, sunflower and corn on its 11ha farm in Promin’, has noted:

“This year we have used a new approach to monitoring of plant growth and protection against pests using pesticides and fertilizers, developed by Dr K.B. Blyuss together with colleagues ... This has allowed us to not only increase the viability of wheat, barley, oat and rapeseed plants by 8-10% on average, but it has also reduced our costs for control of pests of these plants by 15-18%.” [S3].

Enabling the development of successful organic biostimulant products for improved crop protection and yields

In R3-5, Blyuss and Kyrychko proposed and analysed mathematical models of RNA interference, explaining quantitatively how this process develops in plants, and how it can be used to target plant parasites. Results of these models helped their colleagues in the Institute of Microbiology and Virology of the National Academy of Sciences of Ukraine select metabolites of soil bacteria that result in the highest levels of within-plant production of siRNAs protecting plants against parasites, and it was these metabolites that formed the basis of natural microbial biostimulants. After testing in laboratory, greenhouse and field conditions, in 2018 Ukrainian agro-investment and agrochemical companies “Bioinvest-agro” ( http://www.bioinvest.com.ua) and “Imptorgservice” ( imptorgservis.uaprom.net) produced these biostimulants and sold them to farmers and over 100 major agribusinesses around Ukraine and in Kazakhstan. The latter of these companies has annual sales of $500,000 - $1million. Besides direct economic impact in terms of profits to these agrochemical companies, the biostimulants have also resulted in a substantial improvement of crop performance for individual farmers and agribusinesses using them. More specifically, across different regions of Ukraine, these biostimulants have increased yields of winter wheat, barley and sunflowers by 11-21%, 9-18% and 19-63%, respectively, while for vegetables, the yields for tomatoes, potatoes and cucumbers have increased across the country by 20-66%, 10-12%, and 23-26% due to using natural biostimulants [S4, S5].

By virtue of being produced from natural products of soil bacteria, natural microbial biostimulants can be used for organic farming. This is extremely important, since organic farmers suffer substantial losses due to inability to protect their crops against pests using conventional pesticides. In this respect, these natural microbial biostimulants are providing a viable alternative, which allows to protect crops against parasites, while minimizing negative effects on the environment, and maintaining standards of organic farming. One of these biostimulants – “Phytovit”, produced by “Bioinvest-agro” [see R6] – has been approved for organic farming and even received international “Organic standard” certification recognized by the EU and Switzerland [S4] and is currently used by some of the largest organic farming companies in Ukraine.

Results of theoretical and experimental work on natural biostimulants have been published in a joint research publication [R6] in Frontiers in Plant Science, the world’s most cited journal in the field of plant science. Due to major significance of this work in terms of providing effective, and at the same time, environmentally safe, tools of pest control, this work has been covered by a large number of global media specializing in farming and agriculture. This includes articles in English, French, Italian, Spanish, Portuguese, Russian, and Turkish [S6].

The significance of this work has been recognised by an award to Blyuss and Kyrychko of Letters of Gratitude from the PSC, which is the highest honour that can be awarded (only once) to any scientist for their major scientific contribution to solving problems of significant practical importance. These awards were given to Blyuss and Kyrychko on the 17 May 2019 at the annual Day of Science ceremony, organised by the National Academy of Sciences of Ukraine [S7].

Natural microbial biostimulants are already widely used in Ukraine and Kazakhstan. The next phase of impact will concentrate on certification of other natural microbial biostimulants for the purpose of their use in organic farming. The largest Ukrainian producer of organic commodities, which is also the second largest organic grower in Europe and the biggest exporter of organic soybean and corn from Ukraine to the EU, is currently testing these biostimulants for the purpose of adopting them as main tools of pest control in its agricultural lands.

The work under Lakkis's leadership featured in this case study – though inspired by initial discussions with Ambiental's staff – was carried out independently of Ambiental and without sight of the problematic [text removed for publication] code. Nevertheless, it was natural that, once Lakkis produced and openly circulated his findings, they would be relevant and potentially very useful for the company, which proved to be the case. To maximise the impact, an intern student, working under Lakkis, was tasked with integrating those components identified and derived from the existing research [3.2] into the specific [text removed for publication] commercial code. A PhD student of Lakkis, working part-time for Ambiental, further improved the code and worked with Lakkis on the mathematical justification of these improvements.

Lakkis identified a problem with the way the shallow water equations were implemented and then they ran on real-world scenarios such as periods of heavy rain or sudden addition of water (non-conservative regime). Lakkis's team saw that by modifying the underlying kinetic equations (a topic Ambiental's developers were unfamiliar with) to cover the non-conservative equations, there was a possibility of importing stable kinetic schemes (previously only tried for conservative regimes of shallow water) to model the non-conservative scenarios.

In 2014, Ambiental developers integrated the method established by Lakkis in the form of [text removed for publication] fluxes via kinetic averaging as described in section 5.6 of [3.1] within the company’s code. The main import of these novel fluxes is extra stability, and thus robustness, for each single simulation of a flow. Since these simulations could be run hundreds of thousands, or even millions, of times in Monte Carlo runs, this robustness is crucial to reduce, or even eliminate, human intervention in the form of postprocessing of runs and the development of multilevel Monte Carlo methods [3.3].

According to [text removed for publication], the results of the academic work “impacted the whole process”. Previously, filtering out the unstable mathematical simulations from the company’s flood predictions was a time-consuming manual task. It’s “actually the single longest process in building flood maps,” [text removed for publication] [5.1].

After integrating Lakkis’s findings into the [text removed for publication] code, the company [text removed for publication] the amount of time and resources needed to root out instabilities.

[Text removed for publication].

The work of Lakkis and his team has thus benefitted Ambiental’s commercial operations via computational and quantitative products, comprising:

1. The delivery of FloodMap products is significantly lower effort since 2013 – given the considerable reduction in required re-runs. "Bearing in mind that we run 100’s of thousands of simulations to create a national model – even a reduction of a small percentage of unstable simulations can have a significant uplift in efficiency". [5.1]

2. The developed FloodCat processes, which "are still in use today – and will help with future product builds". [5.3]

The impact and growth enabled Ambiental to expand its business; it now works with 50% of the UK insurance market, as well as administrations at various levels in [text removed for publication] [5.1].

[Text removed for publication].

As a commitment to further knowledge exchange and benefit from this collaboration, [text removed for publication], the lead software developer of [text removed for publication] at Ambiental since 2008, will join Lakkis’s research team in 2021 for an EPSRC-funded PhD in collaboration with Ambiental, to learn and specialise further numerical and computational techniques to flood-risk modelling.

The mathematical modelling outlined above has supported COVID-19 control measures at the levels of (1) national policy (in Ukraine) and (2) regional healthcare management in East and West Sussex, and Brighton & Hove (in the UK).

4.1 Impact on the design and implementation of COVID-19 control measures in Ukraine

In Ukraine, the majority of scientific research is done in research institutes of the state-funded National Academy of Sciences of Ukraine (NASU). Prior to emergence of COVID-19, there was no dedicated scientific committee in Ukraine that specialised in modelling epidemics and advising government on strategy. To address the growing problem of COVID-19 and to develop strategies for its containment and mitigation, in early April 2020, the National Security and Defence Council of Ukraine (NSDCU) approached the NASU, which then created the Working Group on “Mathematical modelling of problems related to the coronavirus SARS-CoV-2 epidemic in Ukraine” (equivalent of UK’s SAGE group), chaired by Dr Igor Brovchenko. Since April 2020, this Working Group provides weekly or bi-weekly forecasts of cases, deaths and recoveries for each of Ukraine’s regions to the National Security and Defence Council of Ukraine and the Cabinet of Ministers, who then use those forecasts for policy decisions. It is the only scientific committee in Ukraine that directly communicates with the government with regards to monitoring the dynamics of COVID-19, provides forecasts, and models the effects of different types of interventions.

Building on their previous work on COVID-19 dynamics in the UK [R1], Blyuss and Kyrychko worked closely with Dr Brovchenko to develop a mathematical model of COVID-19 dynamics [R3] that was based on detailed clinical and epidemiological data collected by the Ukraine’s Ministry of Health. Some of the early results of this work featured in a national press conference given to Ukrainian journalists by the Working Group of the National Academy of Sciences of Ukraine on 28 May 2020, where the Chair of the Working Group explicitly stated that the model developed jointly with researchers at Sussex is far superior and more accurate [S1] compared to the basic model used at that point. This new model allowed its users to (i) account for regional differences in population age structure and mixing, and (ii) identify which control measures – such as lockdowns or targeted actions to reduce mixing among the working-age population – lead to the most significant reduction in disease burden. The President and the Head of the Working Group of NASU confirm in their gratitude letter [S5] that the set of “Specific measures that were subsequently introduced on the basis of these results” included:

1. On 13 June 2020, Ministry of Health introduced a requirement for local authorities to modify working hours at local companies to reduce mixing between working-age people during rush hour [S2].

2. On 20 June 2020, the Cabinet modified its guidance and allowed regional administrations to make local decisions on anti-epidemic measures and local lockdowns for different regions depending on their current levels of infection, starting from 22 June 2020 [S3, S5].

3. On 22 July 2020, the Cabinet introduced the tiered system (“green”, “yellow”, “orange”, “red”) for characterisation of severity of epidemiological situation in each of Ukraine’s 25 major administrative regions, starting from 1 August 2020. This tiered system was designed based on the following quantitative measures: the percentage of hospital bed occupancy, the rolling average of the percentage of positive PCR tests per capita, the local growth rate of cases, and the per capita weekly number of administered PCR tests. On the basis of these quantitative measures, restrictions and/or closure of shops, hospitality, sports venues, schools, universities, colleges, public transport were instigated [S4-5].

These measures meant that the Ukrainian Government could:

1. implement epidemic control measures without a nationwide lockdown; and

2. make the best-informed policy decisions they could, and that, in the government’s opinion, their interventions were as effective as possible in slowing the spread of the pandemic.

The President and the Head of the Working Group of NASU confirms these impacts and the work’s continuing legacy: “The results of this research and the associated recommendations have been and are being used by the Government of Ukraine and the National Security and Defence Council of Ukraine when making decisions on the strategy of containment of [the] COVID-19 epidemic in Ukraine” [S5].

4.2 Impact on decision making and care provision in the Sussex region

In response to the COVID-19 pandemic in the Sussex region, the Sussex Health and Care Partnership developed a Gold Command structure which posed modelling questions of strategic operational significance to the Local Health Resilience Partnership (LHRP) covering East Sussex, West Sussex, and Brighton & Hove (with a combined population of 1.7 million).

While national-scale planning for COVID-19 is underpinned by the modelling conducted by SAGE-NHSE, local decision making and planning requires finer resolution approaches. As a result, leaders from Public Health Intelligence approached Madzvamuse and his team to create a consortium, known as the “Sussex Modelling Cell (SMC)”, to undertake COVID-19 epidemiological modelling that is specific to the region [R5]. The work addressed questions of strategic and operational significance for local decision making throughout the pandemic.

a. Forecasting the impact of COVID-19 secondary waves within Sussex [S6-7, S10]. By manipulating the reproduction rate (R number), the model results [R5] showed that if a second wave was to happen it could cause up to ten times the strain on hospital demand compared to the experience that incited the first national intervention. On 20 July 2020, this was the basis of LHRP’s decision to not reduce the available wards for COVID-19 patients, and for funding to be provided for a one-year postdoctoral research assistant to focus on the future of COVID-19 within Sussex [S6]. Kate Gilchrist, Head of Public Health Intelligence at Brighton & Hove City Council, said “[national modelling] was not an accurate representation of what was happening at the time, as we found out when we started using the modelling outputs the university team [R5] derived, and so without the Sussex Modelling Team we could have easily mis-calculated the necessity for wards and body storage inducing monetary and resource mismanagement” [S10].

b. Translating the reasonable worst-case scenario (RWCS) SAGE national forecast [S8, S10]. With the success of the second wave modelling, on 11 September 2020, Madzvamuse’s team undertook research to aid the translation of the SAGE national RWCS document to reflect the impact in Sussex. By working with the official sensitive SAGE documents and through modelling, they demonstrated that not only should Sussex expect the peak of hospital demand later than nationally modelled, but in fact Sussex was already experiencing less hospital demand than what was expected from the RWCS document. On 6 October 2020, this underpinned LHRP’s decision that healthcare capacity was accounted for until at least the end of January 2021 and no changes were required [S8]. Graham Evans, Head of Public Health Intelligence at East Sussex County Council, referred to the RWCS modelling and said “After reviewing the information and the presentation, Mark Angus (Director of Urgent Care System Improvement, Sussex Clinical Commissioning Group (CCG)) said ‘this work has been fundamental to how we have developed our plans for our system and how we monitor our systems as part of our incident management and resilience management.’.” [S10].

c. Forecasting demand and capacity for the death management cell (DMC) [S10]. Following the RWCS modelling, on 4 November 2020 the DMC, a subgroup of the Sussex Resilience Forum dealing with excessive deaths, approached the SMC to understand the demand for the potential body storage requirements over the winter period. Madzvamuse’s team provided this information and on 7 December 2020 the DMC decided to renew their body storage contracts for the winter period. The forecasting for the DMC is particularly important as they need at least a month’s notice of any large changes in deaths, which would mean the ability to forecast reliably at least two months in advance, something which is not attainable using the national modelling. Regarding the modelling of death management, Jacqueline Clay, Principal Manager, Public Health and Social Research Unit at West Sussex County Council, said “Without the University of Sussex Mathematics Team, we would have used short term monitoring of deaths by underestimating the lag between infection, illness and death; this would not have provided a sufficient time period to adjust plans. However, the [DMC] knows, ... that the research and forecasting done here was reliable, easily understandable and accessible”, and “Halogen [the online tool developed by Sussex based on [R5]] gave me the ability to understand how the mathematical modelling works, something that is crucial for having the confidence that the modelling techniques work…” [S10].

d. Urgent Community Response (UCR) service within West Sussex [S9-10]. With the growing hospital demand throughout the pandemic, there was a need to understand the added burden COVID-19 was having on acute hospital care and post discharge services. The UCR service tries to minimise the burden in acute care by (i) providing potential patients with care at home or in other care establishments (like care homes) and (ii) providing extra support to discharged patients who may need it. The research provided by Madzvamuse’s team projected the number of patients discharged throughout the winter period, which – in combination with the discharge pathway analysis by Phil Allman, Head of Performance, Planning and Intelligence for West Sussex at the Sussex CCG – provided the Sussex CCG with the information to appropriately allocate resources and successfully secure an additional £1.63 million to deal with the burden of COVID-19 within West Sussex [S9]. In addition to the typical winter services needed, this extra grant almost doubled the amount requested last year. Phil Allman said “Without the Sussex Mathematics Team [R5], I would have simply used the national model outputs and scaled them to fit the Sussex regional numbers. This would have been highly ineffective due to the different impact COVID-19 has had nationally in comparison to Sussex, as well as the crude approximation from the reduction in magnitude of numbers. This will not only have either majorly under- or overly-estimated the funding needed, but would have caused the wrong allocation of resources, potentially causing major issues later in the winter period” [S10].