Efforts of researchers and other stakeholders to manage an unfolding epidemic: Lessons from potato purple top in Ecuador

ABSTRACT Purple top is a phytosanitary problem affecting potato farmers in Ecuador. Since its emergence, stakeholders implement – with limited success – isolated efforts to reduce the spread of this problem. Hence, this study aimed to derive stakeholders’ enabling responses to manage the spread of emerging plant diseases and pests, and to analyse the spread of purple top in Ecuador together with the responses of the stakeholders. We review 108 articles on human, animal, and plant epidemics to derive enabling responses to manage the spread of emergent diseases and pests. Then, we describe the spread of purple top and stakeholders’ responses using an innovation history analysis approach. Our results show that managing epidemics calls for collaboration and coordination, communication and raising awareness, design and implementation of interventions, funding, supporting regulations and policies, research, and surveillance. We show that the complexity of the diagnosis of causal agent(s) of purple top caused uncertainty among stakeholders. Each actor experienced the purple top’s epidemic differently depending on her or his role. Poor coordination between scientists and other stakeholders, and lack of political lobby hindered the diagnosis and development of management strategies for purple top. A delayed identification, due to the complexity of the causal agent and poor coordination, can further hinder the mobilisation of necessary resources. Our results together indicate that managing epidemics requires more than identifying the causal agent(s) and vectors. We expect that the findings of this study are useful to improve preparedness for other plant disease epidemics affecting food security crops.


Introduction
Current potato production in Ecuador is strongly affected by purple top (Bolaños et al., 2019;Pérez et al., 2020).It is a phenomenon that is commonly labelled as an epidemic and directly threatens the food security of many smallholder farmers in the Andes.Purple top was first reported in Ecuador in 1986, but its presence was sporadic and did not lead to any response from farmers or researchers.However, in 2012 several observations were reported in the northern and central provinces of the country (Bolaños et al., 2019; Ministerio de Agricultura y Ganadería del Ecuador, 2019a).From there, purple top has been spreading to the south.
Purple top in potato plants presents itself as purple and yellow discouloration of the upper leaflets and a typical development of aerial tubers.However, these are symptoms, and do not directly point to a causal agent: they only indicate clogged assimilate flow from leaves to below-ground tubers.Such symptoms can be caused by a range of different pathogens (J.E. Munyaneza, 2005).Currently, researchers in Ecuador associate the symptoms with four possible pathogens or pests: the bacterium Candidatus Liberibacter solanacearum (hereafter Lso, which also causes Zebra chip) (Caicedo, Simbaña, et al., 2020), two types of phytoplasmas (Caicedo et al., 2015;Castillo et al., 2018), and the potato psyllid (Bactericera cockerelli) which also is vector of Lso (Castillo et al., 2019).Researchers are unclear about spread mechanisms of purple top as some of the associated pathogens may also be seed-borne problems (Crosslin et al., 2011;Pitman et al., 2011).
Farmers who are faced with purple top in their potato fields are reacting by increasing the use of pesticides, abandoning infected fields or replacing potato with other crops, and looking for purple top free seed (Castillo, 2019;Lastres, 2018; first author's personal observations).In addition, farmers have started planting potatoes in the paramos, the neotropical alpine wetland ecosystem located above the timberline (Crespo et al., 2010), since, according to them, the purple top's vector can be controlled at these elevations (Navarrete et al., 2020).Farmers have so far received limited training on the management of purple top -mainly focused on the control of the potato psyllid -from researchers and policymakers.
Experiences show that management of a pest or disease epidemic not only relies on technical solutions but also depends on the interaction between the pest or plant disease and its environment, which includes human responses to its presence (Abbott, 2017;Almeida, 2018;Zadoks, 2017).Studies of the management of potato late blight (causal agent Phytophthora infestans), Ebola (Ebola virus), and foot and mouth disease virus and others have shown the importance of the human responses in the unfolding and consequences of an epidemic (Davies, 2002;Fry & Goodwin, 1997;Liu, 2015;Sohrabi et al., 2020).These human responses relate to different aspects of the disease or epidemic.For example, in the case of potato late blight, Fry and Goodwin (1997) highlight the value of collaboration and communication among multiple actors to rapidly identify new strains and raise awareness among potato growers.Carvajal-Yepes et al. (2019) emphasise the value of including social and political considerations to recommend the establishment of a global surveillance system for plant pests and diseases.
In this paper, we describe the unfolding of the purple top epidemic in Ecuador, and how that represented a challenge for the scientists when the diagnosis of the causal agent(s) and vector is still unknown.We also analyse how purple top affected other actors, in particular farmers.A reflection on the way purple top presented itself to researchers, farmers and other actors allows us to draw lessons on the management of an unfolding epidemic.We relate these lessons to the enabling responses which we identified through a literature review.With the lessons we draw, we aim to improve our knowledge on the interaction between plant pathogens and human responses to increase our collective preparedness for future epidemics that could affect crops around the world.

Methodology
We present this study in two parts.First, we report a review of the literature in search of the lessons learned from past human, animal, and plant epidemics.
In the second part we present information on the spread of purple top and the responses of multiple actors in Ecuador from 2012 to 2020.Finally, we present the lessons from the experiences with purple top in Ecuador, using a framework derived from the literature review.
The literature review was carried out in October and November 2019.Using Google Scholar, we searched the terms "Epidemics", "Outbreaks", "Lessons", and "Insights" in combination and downloaded 305 articles.From this set of articles, we selected those that formulated lessons from the epidemic they reported; this narrowed our pool to 108 publications (see Supplementary material 1).With the use of the software NVivo 12 (QSR international), we labelled the lessons to cope with ongoing or future epidemics.We then grouped the labels according to their similarity.This resulted in seven clusters which we called "enabling responses".Together they shape our framework for the analysis of our experiences with purple top in Ecuador.We describe these seven enabling responses.
For description of the experiences with purple top, we adapted the methodology of innovation history analysis or innovation journey analysis (Douthwaite & Ashby, 2005;Spielman et al., 2009).This methodology consists of a construction of a timeline and the identification of the stakeholders during each event.Then, the history is used for further analysis.We used data from our own work as potato agronomists and plant pathologists, in which roles we visited farmers and actively participated in meetings about purple top with peers, colleagues, and policy makers.We talked to key peers about purple top in Ecuador to identify the most important events (e.g.situations where critical decisions were taken).We also made an inventory of relevant information on purple top that was made available over time, highlighting from whom and where the information was derived.We ordered all our information of meetings and information on purple top chronologically on a timeline to show the main events in the spread of purple top (Figure 1).This step also included the analysis of surveillance data that was kindly provided by the Agency of Plant and Animal Health Regulation and Control of Ecuador (Agrocalidad).The surveillance data were collected when technicians of Agrocalidad visited farmers' fields.This dataset included the presence or absence of purple top disease and the psyllid, the location, and date of data collection.The analysis of surveillance data consisted of establishing the frequencies of infected/infested fields with purple top or the potato psyllid.Then, we aggregated the frequencies as monthly percentages of the total observations collected from January 2018 to October 2019 in each province.

Enabling responses: The lessons from literature on human, animal, and plant epidemics
We identified seven enabling responses through our literature review (in alphabetical order): (1) collaboration and coordination, (2) communication and raising awareness, (3) intervention design and implementation, (4) funding , (5) policies and regulations, (6) research, and (7) surveillance (Table 1).Each of these enabling responses offers cross-cutting lessons to manage human, animal and plant epidemics (Table 1).The literature shows that proper management of epidemics requires good leadership and early engagement of multiple actors.Also, constant and contextualised communication among multiple stakeholders is needed to design and implement early and timely interventions.On the other hand, the design and implementation of these interventions require reliable information -such as identification of the causal agent(s), accurate estimation of yield losses, and negotiation with farmers and decision makers -to strengthen existing knowledge and coping mechanisms to manage epidemics.Furthermore, availability of funds is critical to implement interventions and to strengthen local capacities and facilities.Supporting regulations and policies are required to facilitate the process of managing epidemics.Research needs to explore collaboratively a diversity of aspects that are related with the epidemic, such as its environmental and economic consequences.Finally, surveillance should monitor multiple places and events, analysing and sharing the information openly with stakeholders.
We found that these enabling areas are complementary and interdependent: most articles in the literature review described the lessons to manage epidemics combining several or all enabling areas.Designing good interventions, for instance, requires the use of information coming from research and surveillance.Moreover, the management of an epidemic needs funding and a responsive political environment.

The case of potato purple top in Ecuador
The purple top first presented itself in Ecuador in 1986: its incidence was sporadic, and it was not yet considered a threat for potato production.Back then, phytoplasmas were the suggested causal agents of purple top (Ministerio de Agricultura y Ganadería del Ecuador, 1986).However, the validity of this report has been questioned because it did not comply with the "International Standards for Phytosanitary Measures" No. 8 (International Plant Protection Convention, 2006).The observations in this report contrast strongly with today's perception that purple top is a major threat for potato production.• Key local and international stakeholders need to be involved.
For instance, farmers' organisations, policy makers, and NGOs.
• Collaboration needs to take place as early as possible in the epidemic.
• It is necessary to establish a negotiation process to realise concerted actions.Communication and raising awareness • Share a clear and timely message and reports of the epidemics including yield losses.
• It is vital to foster regular meetings to keep stakeholders updated.
• Raising awareness about the epidemics needs to consider the actor and the local context.For instance, local languages and appropriate communication channels.Design and implementation of interventions • Educated interventions need to be designed as early and as timely as possible considering existing information.
• Measures should aim to support social adaptation and transfor- mation in the future.
• Preparedness and contingency plans should be reviewed con- sidering the lessons of previous epidemics, capacity assessment theory and emergency management theory.Funding • Emergency, medium-, and long-term funding should be made available to: (1) develop rapid and effective control strategies (e.g.identification of causal agent, development of diagnostic tests); (2) implement/strengthen preparedness and contingency plans; (3) improve facilities and capacity building and (4) promote research.

Policies and regulations
• Policies and regulations should facilitate the process of manage- ment of the epidemic.
• Regulations should be flexible to work on current and future epidemics depending on the context.
• Reflect on the role of the pest or disease in the design of new regulations or response plans.
• There is a need for coordinated international policies to avoid future epidemics.Research • Implement participatory research strategies such as farmers' field schools or farmers´ research networks.
• Implement and integrate a diversity of approaches to facilitate an integral understanding of the epidemic.These could entail modelling (e.g.spatial or network modelling), genomic, social, economic (e.g.variation in production costs), and political approaches.Surveillance • Early laboratory surveillance is vital for containment of the epidemic.
• Surveillance strategies should consider monitoring disease dis- semination events (e.g.markets or seed exchange fairs) and include secondary databases in a pre-outbreak phase.
• Data analysis should be carried out and shared openly.
• Active (provision of information) and passive (reception of information) surveillance are complementary.Hence, close coordination with farmers should be strengthened.

Region zero
Purple top was again reported in potato fields in Carchi -the northern province of Ecuador -sometime in 2012-2013 (Figure 1) (Bolaños et al., 2014(Bolaños et al., , 2019;;Castillo et al., 2018).Symptoms of purple top were also spotted in farmers' fields in the province of Tungurahua, the centre of the country, but the incidence was restricted to a few dispersed potato plants in fields with the cultivar Super Chola (Bolaños et al., 2019;Castillo et al., 2018).
Two hypotheses circulated among researchers on the appearance of purple top symptoms in the province of Carchi.The first one assumed that it had entered from Colombia with illegal imports of potato (Diego Peñaherrera, personal communication, 2019;INIAP, 2018).This opinion aligned with complaints of farmers and traders about the important competing flow of ware potatoes from Colombia entering the country (e.g.El Comercio, 2012;El Norte, 2013).The second hypothesis was that the disease had been there for a long time, and that changing climate conditions facilitated its spread (Bolaños et al., 2019).In late 2013, only a few reports on purple top incidence appeared in the country.

The first reports and tests in 2014
Nothing hinted at a further spread of purple top until in 2014 when researchers of the Ecuadorian National Agriculture Research Institute (INIAP) noticed an increase in advertisements by agrochemical stores in Carchi for fungicides to manage Rhizoctonia solani.The researchers visited a potato field in Carchi and noted that 5-20% of the plants showed symptoms of purple top (Bolaños et al., 2014;Ochoa et al., 2018) (Figure 1).Yield losses for farmers were estimated to be close to 50% (Bolaños et al., 2014).The symptoms in the field were similar to those caused by R. solani, but the scattered affected individual plants in the field did not match the normal patched pattern of plants in fields affected by R. solani (Castillo et al., 2018;Gilligan et al., 1996;José Ochoa, personal communication, 2019).The researchers therefore thought of a phytoplasma as the potential causal agent considering the distribution pattern of infected plants in the field.To explore this hypothesis, INIAP researchers visited affected fields in Carchi a second time, but they failed to find any potential phytoplasma vector, such as the insects of the family Cicadellidae.Thus, there were no grounds for recommending pesticide applications.Researchers advised farmers to sell the entire harvest as ware potato, discard any tubers kept for seed, and plant certified seed instead (Ochoa et al., 2018).Laboratory analysis of samples from infected potato plants collected in the cantons Montúfar and San Pedro de Huaca in Carchi did test positive for a phytoplasma, but the tests (nested PCRs based on standard PCR to amplify phytoplasma DNA) could not identify which phytoplasma.A genetic sequencing of the DNA extracted from symptomatic plants was necessary to identify the phytoplasma and to rule out the possibility that false positives were found due to amplification of non-specific products of the nested PCR (Baez et al., 2016;Rivadeneira et al., 2015) (Supplementary material 2).

First outbreak (2015-2016) and the further search for the pathogen and its vector
While the researchers were still not sure about the identity of the causal agent and the possible vector, reports of purple top incidence in farmers' fields rose sharply in 2015 and 2016, especially in the provinces of Carchi, Cañar, and Pichincha (Figures 1, 2).In Carchi alone, researchers estimated that 48% of all potato fields were infected in those years (Caicedo et al., 2015;García et al., 2015) (Figure 2).
To pursue clarity on the aetiology of purple top, INIAP researchers in 2015 sent DNA samples of infected potato plants to laboratories outside the country.However, the test results were confusing because they did not point to the presence of phytoplasmas, but rather suggested the presence of Lso (Baez et al., 2016, further details in Box 1).Towards the end of 2015, researchers from the Central University of Ecuador found a phytoplasma in a sample from Carchi: Candidatus Phytoplasma aurentifolia S16rII.
Box 1.Researchers struggling with the identification of the pathogen INIAP researchers sent samples to laboratories in the USA (27 samples), Mexico (5 samples), New Zealand (8 samples), and Peru (6 samples).The results were confusing: the laboratories in the USA, Mexico, and Peru did not find any phytoplasma in the samples (Baez et al., 2016;Muller, 2016).However, the laboratory in Mexico did find the bacterium Lso in some of the samples.The bacterium Lso causes purple top symptoms in the above ground parts of the plant, and, in the tubers, it produces an internal brown discolouration when sliced and dark stripes when fried (Baez et al., 2016).The eight samples sent to the laboratory in New Zealand were only tested for the bacterium Lso, but all tested negative (Baez et al., 2016).In November 2015, researchers from the Central University of Ecuador finally identified the possible phytoplasma associated with purple top: Candidatus Phytoplasma aurentifolia S16rII was detected in potato samples from Carchi (Figure 2) (Caicedo et al., 2015).Agrocalidad started monitoring the identified phytoplasma using the same set of primers and protocols used by Caicedo et al. (2015).Unfortunately, the phytoplasma was found in both symptomatic and asymptomatic potato plants.This could mean different things.Either the causal agent was still not recognised, the concentration of the DNA of the phytoplasma was too low to be detected, or there were human errors.In the same month, Agrocalidad carried out three different tests on infected potato plants to identify the causal agent: one using the "Fluorescence In Situ Hybridisation (FISH)" technique to identify phytoplasmas, another one using a PCR technique to identify bacteria, and another one using a nested PCR technique to identify phytoplasmas.These tests contributed to the struggle of finding the causal agent of purple top since, in contrast with the previous studies, they excluded the presence of the bacterium Lso but suggested the presence of phytoplasmas in the phloem of the infected plants (García et al., 2015).The genetic sequencing showed no existing relationship with any earlier reported phytoplasma in potato (García et al., 2015) (Supplementary material 2), including the one reported by Caicedo et al. (2015).To identify potential vectors, INIAP's researchers surveyed infected fields and the collected insects were analysed by Agrocalidad technicians.They found 275 different specimens of the family Cicadellidae and one specimen of the family Psyllidae (Baez et al., 2016).
Agrocalidad then began to monitor and to test for this phytoplasma, but again with confusing and disappointing results: the phytoplasma was found in plants with and plants without the symptoms (first author's personal observations).Agrocalidad and INIAP also collected insects to identify potential vectors (Baez et al., 2016).
While there was no clarity on the causal agent and its vector, there was urgency to provide recommendations to farmers who were facing purple top in their potato fields, suffering severe yield reductions. INIAP and Agrocalidad recommended to use certified or healthy seed, and to train technicians and farmers in recognising symptoms of purple top (Supplementary material 2).The application of salicylic acid to the seed was also recommended because researchers found in the literature that it could potentially stimulate the resistance of the plant to phytoplasma infection (García et al., 2015;Wu et al., 2012).

An intermezzo and the continued search by researchers
In 2017, the reported incidence of purple top strongly declined.Nevertheless, researchers continued their search for the causal agent.Studies in INIAP in 2017 showed again that Lso was not present in the newly collected samples (Castillo et al., 2017).These studies also revealed the ambiguity of the PCR tests -with new primers -to detect phytoplasmas (Castillo et al., 2017).Not knowing the causal agent of purple top was a strong limitation for the design and formulation of integrated pest management (IPM) strategies (Castillo et al., 2017).
In September 2017, researchers collected specimens of the potato psyllid in potato fields in Pichincha.The psyllid is an insect that can transmit Lso (Castillo et al., 2019;Castillo, 2018;J. E. Munyaneza et al., 2007;J. Munyaneza et al., 2007) and can induce purple top symptoms in the foliage called "psyllid yellows" (Sengoda et al., 2010).This motivated Agrocalidad to monitor this insect, but evidence that potato psyllid was the causal agent of purple top was still lacking; all explanations to this point had been based on correlated observations.

Second outbreak (2018-2019): Further spread of the disease
Research findings and observations on the spread of purple top incidence peaked in 2018 (Figure 1).First, INIAP, in collaboration with the department of Agricultural Science of the University of Bologna, reported the identification of another phytoplasma in samples of infected plants: Candidatus Phytoplasma [aster yellow cluster, ribosomal subgroup 16SrI-F] (Castillo et al., 2018).Thus, there were now two phytoplasmas possibly related to purple top.However, because of limited funding, researchers were not able to confirm -through a modification of Koch's postulates (Box 2) -if phytoplasmas caused purple top.
While uncertainty about the causal agent(s) and possible vectors continued, the purple top disease steadily increased and spread south.The increased reporting of purple top motivated technicians of Agrocalidad to randomly monitor and report fields with symptoms of purple top.In 2018, potato plants with purple top were observed in farmers' fields in the provinces of Carchi, Imbabura, Pichincha, and Cotopaxi (Figure 3).In early 2019, the first reports of purple top infected fields in Tungurahua, Bolivar, Chimborazo, Cañar and Azuay appeared, while the number of reports from Pichincha increased rapidly (Figure 3).There were no fields with purple top reported in 2019 from Imbabura (Figure 3) and Loja (data not presented).
Next to the two phytoplasmas, there was also the potato psyllid that had to be considered as causal agent.The potato psyllid surveillance data of Agrocalidad in 2018 showed that this insect was present in the same provinces as purple top and showing similar trends (Figure 3).However, potato psyllids were not necessarily found in the same fields as purple top.For instance, in Pichincha 66% of the inspected fields were infested with the potato psyllid, of which only 63% showed symptoms of purple top.In contrast, in Carchi 48% of the fields inspected were infested with the psyllid but no plants with purple top symptoms were found in these fields.In 2019, the percentage of psyllid infested fields grew rapidly mainly in Cotopaxi and in Tungurahua (Figure 3).The potato psyllid, also known as "Paratrioza" (the previous taxonomic genus of the potato psyllid and common name in Central America) became so notorious during this time that it became the new reference point in conversations of farmers and technicians: Did they have "Paratrioza" in their fields already?What were the right chemicals to fight "Paratrioza"?and What should be the frequency of pesticide sprayings?By mid-2019, the psyllid had not yet reached the southern provinces (Bolivar, Cañar, Azuay and Loja) in Ecuador (data not presented).

Initial collaboration: Starting to fight back
From the first reports of purple top in 2012 to the end of 2018, biophysical researchers from various institutions had acted individually, mostly Box 2. Description of Koch's postulates for the potential causal agents of purple top.
Koch's postulates are a series of tests aimed at determining the causal pathogen of a disease.It consists of the following steps: (1) identification of material showing symptoms of the disease, (2) isolation of potential causal agent usually in a culture, (3) inoculation of the potential causal agent in healthy material, and (4) monitoring of symptoms in healthy plants.However, with phytoplasmas or the bacterium Lso involved, Koch's postulates are not satisfied since these pathogens cannot be isolated (Secor, 2007;Wang et al. 2017).The use of grafting or "hypothesised vector" are alternatives that could be tested for confirmation as reported by Akhtar et al. (2008), Arocha et al. (2005), and Lee et al. (2004).
communicating with colleagues informally and scarcely interacting with other stakeholders.During this period, there was little involvement of social scientists.In November-December 2018, the increased reporting of purple top in farmers' potato fields motivated the Ecuadorian Ministry of Agriculture and Livestock (MAG) and the agro-chemical company Syngenta to invite Lorena Lastres, a Peruvian expert on IPM, to Ecuador (Giovanna Muller, personal communication, 2018; Ministerio de Agricultura y Ganadería del Ecuador, 2018).After visiting fields in several provinces, Lastres observed that (1) farmers were abandoning potato fields infected with purple top; (2) weeds of the Solanaceae family had symptoms of purple top and were infested with the potato psyllid; (3) there was a high presence of volunteer potato plants in the fields that could act as sources of inoculum; and (4) the potato psyllid was present in potato fields (Lastres, 2018in La Primicia, 2018).
After Lastres' visit, several researchers of INIAP decided to join efforts and formulate a purple top management strategy.This strategy had recommendations on several components: research, seed system development and training, and surveillance.INIAP was to lead the research component, MAG would be responsible for seed system development and training, and Agrocalidad would support the surveillance (Supplementary material 2; Cuesta, 2019;INIAP, 2018).
In early 2019, Dr. Carmen Castillo, a researcher from INIAP, published an article on the evidence of the potato psyllid Bactericera cockerelli (central haplotype) being present in potato fields in Ecuador (Castillo et al., 2019).Her study suggested that the potato psyllid was introduced through imported agricultural produce (Castillo et al., 2019).However, the introduction mechanism was difficult to identify because nearby countries like Colombia, Panama, and Peru had not yet reported the potato psyllid (Castillo et al., 2019;EPPO, 2020).

Actions taken to mobilise other stakeholders: Speeding up
Towards the end of 2018, researchers at the Potato Centre (CIP, Centro Internacional de la Papa) in Lima and Quito approached officials of the FAO delegation in Quito to discuss potential support mechanisms to mitigate the impact of purple top.CIP's researchers were concerned with the increasing number of reports of infected fields, especially in areas where certified seed was also produced (André Devaux and Peter Kromann, personal communication, 2020).FAO officials indicated the first step was to develop a Pest Risk Assessment (PRA): a process that describes possible risks, dissemination pathways of any pest (e.g.insects, fungi or plant), and identification of available phytosanitary management measures (International Plant Protection Convention, 2017).
In January 2019, MAG organised a two-day workshop to produce a document synthesising the PRA about purple top disease (Ministerio de Agricultura y Ganadería, 2019a).There were 13 persons in the workshop, representing INIAP, CIP, MAG, and Agrocalidad.After this workshop, a PRA document was produced that described the complex interaction between the potato psyllid, Lso, and phytoplasmas, and the hypothesised impact in the country.However, Agrocalidad could not be motivated to declare a "phytosanitary emergency" because such declaration was not possible without research showing the environmental, economic, and social impact (Douthwaite, 2020).In the end, the information in the PRA and the absence of the phytosanitary emergency did not justify FAO to activate supporting mechanisms and provision of emergency resources (Karina Marcillo, personal communication, 2020).
Meanwhile, the increasing number of reports of purple top and the potato psyllid became a concern for multiple actors.This motivated the implementation of different initiatives.INIAP prepared a guide for extensionists to manage this disease (Cuesta et al., 2018) because access to reliable information was also a problem next to the identification of the causal agent(s).MAG, INIAP, and Agrocalidad organised training sessions for extension agents of MAG and farmers as poor identification of purple top and the psyllid in the field would be an issue for the management of these problems.These training sessions aimed to inform better about the symptoms of purple top, the presence of the potato psyllid, and possible measures to reduce incidence and prevent spread, such as rotating the use of different insecticides and removal of foliage after harvesting (INIAP, 2018).Thereafter, extension agents trained potato farmers in different parts of the country supported by agrochemical companies (first author's personal observation).MAG and INIAP also disseminated information about purple top through TV, radio and social media.But still there was no clarity about the causal agent and vector of purple top in the potato fields of Ecuadorian farmers.
In February 2019, the Central University approached MAG to co-organise a two-day meeting to design a "National Strategy to Mitigate Purple Top" (Yandún, 2019, Figure 1) and create a multi-stakeholder platform.Approximately 40 participants took part in the meeting, with representatives from Central University, MAG, INIAP, Agrocalidad, CIP, private companies, but no one representing farmers' organisations.The resulting strategy considered the same components that had been earlier suggested by INIAP and would be coordinated by MAG (Ministerio de Agricultura y Ganadería del Ecuador, 2019b).However, there were not sufficient funds available.Without an explicit coordinator this and other efforts of setting up an multistakeholder process failed.
The multi-stakeholder platform that was created early 2019 did not have an explicit coordinator and the platform did not meet anymore (first author's personal observations).Other meetings involving OIRSA (Regional International Organisation for Plant and Animal Health Protection), Agrocalidad, MAG, FAO and CIP were held to discuss the management of purple top.However, coordinated actions were not further planned (Karina Marcillo, personal communication, 2020).While potato farmers and researchers struggled to manage purple top, in July 2019, Ecuador was confronted with a new epidemic threat in banana (Ecuador's main export commodity): the Fusarium wilt tropical race 4 (Fusarium odoratissimum), initially reported in Colombia (García-Bastidas et al., 2019) and now threatening banana production in Ecuador.The threat motivated Agrocalidad to implement phytosanitary activities to prevent the entrance and dissemination of Fusarium wilt tropical race 4 (Douthwaite, 2020;Registro Oficial del Gobierno Del Ecuador, 2019), with a budget of 18 million dollars (El Universo, 2019).
In January 2020, concerns of purple top entering Peru persuaded CIP scientists in Lima to strengthen and initiate regional collaborations, and to connect Ecuadorian and other researchers (Figure 1).They organised the first regional workshop on potato purple top, inviting 62 people from local and regional agricultural institutions and universities, NGOs, and agro-chemical companies (Pérez et al., 2020).In this meeting, Jorge Caicedo, a researcher from the Central University of Ecuador, officially reported the presence of Lso in Ecuador (Caicedo, Simbaña et al., 2020).An important agreement made in the workshop was to establish a multi-institutional international committee to coordinate and follow up on the actions suggested during the workshop (Pérez et al., 2020).

Observed impacts of purple top and the psyllid on farmers and private companies
As mentioned, at the time of writing, researchers still have no conclusive evidence of which pathogen causes purple top in potato fields in Ecuador.Other aspects of the purple top are however clear.Farmers are responding to purple top by using seed from large-scale farmers rather than saving their own, abandoning infected fields, planting other crops, and in some cases they stop planting potato entirely (Lastres, 2018; first author's personal observations) or are planting potatoes in the fragile ecosystem of the paramo (Box 3, Bert de Bièvre, personal communication, 2020).Purple top has affected the livelihoods of farming households through the reduction of potato production areas, foregone labour incomes, and increased production costs because of increased use of insecticides (Castillo, 2020;X. Cuesta, interviewed by El Heraldo, 2019).
Purple top also presents an anxiety for various other private sector actors.Food industries are concerned about the impact of purple top on the availability of ware potato for their products, e.g.chips.Seed companies have The first author together with FONAG's team of experts followed up on the concerns that were heard about farmers moving their potato production from lower to higher altitudes (~3400 metres above sea level) to avoid potato purple top.They found several potato fields when visiting the paramos near the Cotopaxi volcano (Figure 4).One potato producer looking to rent a piece of land in this area said, "Here, the potato psyllid exists but you can control it".NJAS: IMPACT IN AGRICULTURAL AND LIFE SCIENCES been suffering because they cannot assure the delivery of quality seed anymore.Both are looking for varieties resistant to purple top or sources of clean seed and have sent requests to CIP.Yet, more studies are needed to confirm the impacts that purple top and the psyllid are having on farmers, consumers and companies.

Knowns and unknowns of purple top and the psyllid in Ecuador
Even though the impact of the purple top disease is evident, this is what we know at the moment of writing this article about purple top in Ecuador.First, the potential causal agents of purple top diseases are: the potato psyllid, Lso, and two types of phytoplasmas, probably occurring in mixed infections.However, it is not clear to which agent we can attribute the high incidence in farmers' fields.Second, the potato psyllid that we found in Ecuador is the central haplotype (Castillo et al., 2019) which is associated with the transmission of Lso (J.E. Munyaneza et al., 2007;J. Munyaneza et al., 2007).Third, in Ecuador, the potato psyllid has been spotted in other Solanaceae including chilli pepper (Capsicum frutescens), pepper (Capsicum annum), tamarillo (Solanum betaceum), tomato (Solanum lycopersicum) and golden berry (Physalis peruviana) (Chamorro et al., 2020;Tipán, 2020).Phytoplasmas have not been identified in these crops (Tipán, 2020).But Lso (haplotype A infecting potatoes) was found in tamarillo and golden berry (Caicedo, Vallejo, et al., 2020).Fourth, we know that purple top and the potato psyllid are already present in the southern provinces of Ecuador.At the time of writing, purple top was not reported in the province of Loja (south of Ecuador); however, the potato psyllid has been found already in the provinces of Bolivar, Cañar, Azuay and Loja, and recently in Colombia (Instituto Colombiano Agropecuario, 2021) and Peru (Servicio Nacional de Sanidad Agraria del Perú, 2021).
There are also still important unknowns.For one: what is the predominant causal agent and how is it transmitted?Furthermore, we do not have accurate estimates of how farmers and other stakeholders are affected (e.g.production losses or the number of potato fields affected by purple top) and how their responses to purple top affect other components of the agro-ecological system.To answer these questions, it is necessary to identify or develop reliable diagnostic tests (see Box 2), and consequently, information on transmission, the role of host plants, effects on yield and quality, and the effect of agronomic (IPM) practices.Only then effective management strategies can be formulated to reduce purple top spread.
Different actions are planned for Ecuador.Currently, a project funded by the Spanish Agency of International Cooperation (AECID) is monitoring and studying the ecology of the potato psyllid and raising awareness of the problem.In addition, different agrochemical stores have identified pesticidebased management strategies and started promoting them.However, these actions are isolated initiatives and not part of a multi-actor coordinated programme.At the regional level, the initiatives look promising: the regional multi-institutional committee formed in Peru in early 2020 is having regular update meetings and discussions to explore funding for the jointly defined research action plan (first and third author's personal observations).At the time of writing this article, the One CGIAR Initiative on Plant Health assigned funding to address purple top at regional level.

Discussion and conclusions
Potato purple top has peaked in two periods in Ecuador since 2012, causing significant environmental, economic and social problems.We analysed the unfolding of the epidemic based on an innovation history analysis approach and structured the responses on the basis of enabling responses identified in our literature review.We conclude that in the case of purple top in Ecuador, an effective response of multiple stakeholders has been hampered in the first place by the lack of knowledge on the causal agent(s).The complexity of the pathogen(s), its vectors and the associated myriad of testing options challenged researchers on the diagnosis and the understanding of the disease.However, additional factors have contributed to the ineffective management of the epidemic as shown by the analysis of the spread of purple top and the responses of different stakeholders.
We noticed that purple top represented different problems for each actor: for biophysical researchers, it is a problem related to the potato plant, causal agents and vectors; for socio-economic researchers -largely absent in the unfolding of the purple top epidemic -, it could be a problem related to production costs, incomes and effects on livelihoods; for farmers, it represented a challenge in the use of pesticides, loss of yield, income, and food security; and for decision makers it was and still is a secondary problem causing minor socio-economic impacts.For agrochemical dealers, purple top presented a potential source of additional income through the selling of pesticides.For consumers it may have led to higher prices of their most important staple.While agricultural researchers struggled with the diagnostics of the causal agent(s) and vector(s), the other representations of the epidemic and the associated research areas (e.g.environmental and economic impact) remained largely outside their field of observation and consideration.This study clearly shows that researchers working alone -and/or solely in their laboratories, focusing on their part of the problem -cannot manage a new disease or pest, especially when dealing with complex pathogen(s).The analysis suggests that disconnectedness between the different aspects of the problem and respective stakeholders may explain the limitations to effectively manage the epidemic.
Using the enabling responses (Table 1), there are concrete lessons that potato stakeholders can consider to prevent the spread of emergent diseases taking the purple top case in Ecuador as reference.We learned that the capacity to manage epidemics is highly dependent on collaboration and coordination among stakeholders.Involving farmers' organisations, social scientists, and other stakeholders at early stages of the epidemic could have helped identify opportunities to mitigate the other aspects of the epidemic problem.A neutral actor might be needed for the coordination to create a fair decision-making environment (Thiele et al., 2011).We also learned that raising awareness and communicating available and unavailable knowledge is an essential ingredient of collaboration and decision making under scenarios of uncertainty.Though, without special budgets and resources all efforts to communicate and collaborate are hindered.Another lesson, from the spread of purple top, is that policies and regulations need to enable an environment to support the management of emergent diseases and pests affecting food security crops as occurs for export crops.We also learned that the enabling responses and the corresponding lessons did not stand on their own: they showed a high level of interaction and interdependence.For example, without good diagnostics and research, an effective stakeholder response and communication on management of the epidemic were difficult.Hence, the lessons described cannot be considered isolated but need to be integrated for the management of purple top.
It is important to recognise that collecting the required evidence about purple top and mobilise advocacy are limited by: (1) the complexity of the pathogens that usurp the attention of the researchers, and (2) the fact that potato is a food security crop produced by smallholder farmers who are not well organised and cannot easily influence policy makers.This situation is very different from the situation in the banana sector in which the mobilisation of resources to manage Fusarium wilt tropical race 4 was more successful.In addition, assessing the impact of purple top on the economy is hampered by the high and rapid fluctuation of ware potato prices.At this moment, mid 2022, there is still a need for sound evidence showing the importance of this epidemic, and for advocacy to leverage the evidence.
Finally, we -the authors -realise that we learned much from the analysis and reporting of this case and writing this publication.Although we consulted a wide range of experts and collaborators in order to compile the presented information, we also acknowledge that the interpretation of that information is influenced by our analytical lens.Furthermore, this case makes clear that scientific research does not take place in a vacuum and is subject to societal dynamics; we therefore need to engage with other stakeholders and policy processes when opportune.We are convinced that a multi-stakeholder reflection on the experiences with purple top in Ecuador, using inputs from Douthwaite (2020), would yield additional important insights and lessons that contribute to improving the socio-technical management of plant epidemics, particularly in relation to purple top at the regional level in Latin America.
under the project 16-275.We appreciate the support of Agrocalidad in providing information and participating in discussions about managing potato purple top in Ecuador.We also would like to thank the team of FONAG for supporting the visit to the paramos and the people who shared conversations with the authors.We also thank Claire Nicklin and anonymous reviewers for their insightful comments on earlier versions of the manuscript.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Figure 1 .
Figure 1.Timeline of events characterising the spread of purple top in the tropical highlands of Ecuador.

Figure 2 .
Figure 2. Incidence of potato purple top in 2015 and 2016.

Figure 3 .
Figure 3. Fields infected with purple top and infested with the potato psyllid in the main potato production provinces of Ecuador from 2018 onwards.Shaded areas show data collected in 2019.

Box 3 .
Change of land use in the Ecuadorean Andean paramo due to shift of potato production

Figure 4 .
Figure 4. Change of land use in the paramo to avoid the impact of potato purple top.Left: Natural Andean paramo of the volcano Cotopaxi.Right: The same Andean paramo landscape with potato fields on the other side of the road.Pictures taken by Israel Navarrete (8 th January 2020).

Table 1 .
Enabling responses and associated cross-cutting lessons reported in selected literature on human, animal, and plant epidemics.