In memoriam Professor Jean-Antoine Rioux (1925-2017)

René Houin, Nicole Léger, Jean Dupouy-Camet, Patrick Bastien, and Gérard Luffau 1 Académie Vétérinaire de France, 34 rue Bréguet, 75011 Paris, France 2 UFR de Pharmacie, Université de Reims Champagne-Ardenne, 51096 Reims, France 3 Faculté de Médecine Paris Descartes, Paris, France 4 Département de Parasitologie-Mycologie, CHU de Montpellier/Faculté de Médecine, UMR “MiVEGEC” (CNRS 5290/IRD 224 /UM), Université de Montpellier, Av. Charles Flahault, Montpellier Cedex 5, France 5 Unité de virologie et immunologie moléculaires, Inra, Jouy-en-Josas, France

Born in Naucelle (Aveyron, south-central France), in a family firmly rooted in the Limousin region, Jean-Antoine Rioux ( Figure 1) spent his childhood in Le Vigan, in the heart of the wild Cevennes mountain range. It is undoubtedly there that he acquired his passion for studying nature, which underpinned his activities during his whole life. This fascination led him to deepen the study of botany, a skill which gave him the opportunity later on, between 1977 and 1993, to become the director of the Botanical Garden ("Jardin des Plantes") of the city of Montpellier [12]. He devoted many years to this task with talent and perseverance, preserving and enriching the plant collections, renovating the greenhouses, promoting the garden's image, and ensuring its protection through its classification among "Noticeable Sites and Landscapes" in 1982 and "Historical Monuments" in 1992. To this initial passion was added a strong medical vocation. He was a brilliant medical student during his studies in Montpellier, and graduated top of the class in the "Concours de l'Internat" (a competitive examination to recruit hospital residents) in 1951. He was then appointed as Assistant Professor in Montpellier University Hospital and specialized in dermatology and lung diseases. But in 1952, he soon found the area in which he reached his greatest achievements, by joining the Laboratory of Medical Natural History headed by Hervé Harant, skillful naturalist and parasitologist with great scientific understanding. Jean-Antoine Rioux then specialized in parasitology, and he continued his whole career in this organization termed the "Laboratory of Parasitology" which later became the "Laboratory of Medical Ecology and Parasite Pathology". Jean-Antoine Rioux built a remarkable establishment, recognized and supported by research authorities, which allowed him to conduct research work leading to more than 500 publications.

Early work on malaria and mosquitoes
In French faculties of medicine, teaching the epidemiology of malaria has always been one of the major goals of the parasitology curriculum. The young Jean-Antoine Rioux aimed at transforming this teaching topic into a research activity. During the preparation of his PhD on the "Culicidae du Midi Méditerranéen", he had acquired strong entomological training, which he used to study malaria in Languedoc-Roussillon, proposing the concept of "unstable malaria". This concept would be taken up later, on the occasion of several epidemiological missions in North Chad (Borkou-Ennedi-Tibesti). The persistence, even temporary, of malaria parasites in oasis inhabitants especially in sedentary children, was indicative of in situ transmission and of the real presence of vectors. In fact, he found larvae of Anopheles gambiae complex in the overflow sites of the Faya-Largeau region and one year later, the adults were captured in an oasis dwelling. During these surveys, 25 Culicidae species were identified, providing important bio-geographical clarifications. In the Mediterranean South of France, he also studied variable autogenesis of halophilic sublittoral Aedes (Ochlerotatus) detritus. This variability suggested either simple genetic polymorphism or a true segregating population (cryptic species). Analysis of allelic isozyme frequencies in adults from the same larval breeding habitat (in collaboration with Nicole Pasteur) confirmed the second hypothesis. This twin species was described under the binomial appellation Aedes (Ochlerotatus) coluzzii Rioux, Guilvard & Pasteur [10]. A. coluzzii could be bred without any difficulty in an insectarium, whereas to obtain fertile progeny with A. detritus s. st., artificial fertilization was necessary. Jean-Antoine Rioux's knowledge of entomology led him to perform a major ecological survey on Leptoconops irritans, a particularly aggressive Ceratopogonidae species, found in coastal areas from Camargue to Roussillon [3]. The objective of this action was to identify larval biotopes and to specify seasonal dynamics. A sampling transect was carried out for two years in Middle Camargue. The results clarified the situation and functioning of emergence sites of L. irritans: the most productive phytocenosis was formed by the perennial Chenopodiaceae groups, forming the essential parts of halophilic ecosystems of Middle and Lower Camargue. It was also on this basis that integrated pest management against "nuisance mosquitoes" was proposed (in French we use the word démoustication invented by Hervé Harant). The concept of nuisance concerns hematophagous species responsible only for discomfort and not for the transmission of pathogens (vectors). In order to implement measures to limit these species, it was necessary to obtain ecological data on all stages of insect development. This situation is quite different from that of vector-borne diseases, where the vector is only the entomological component of a pathogenic complex including vertebrate hosts (so-called reservoirs), humans and the pathogen itself. It is then necessary to act on compartmentalized cyclical systems; the control of which implicates very different operational strategies (integrated control). These strategies have later been developed for controlling leishmaniasis and will be detailed below. As mentioned above, mosquito control in Languedoc-Roussillon concerned four highly harmful species: Culex pipiens, A. caspius, A. detritus and A. coluzzii. The initiative, launched by the "Entente Interdépartementale pour la Démoustication du Languedoc-Roussillon" (EID Languedoc-Roussillon) created in 1958, was led by the interdepartmental tourism development authorities of the Languedoc-Roussillon region ("Mission Racine") created in 1963. In sublittoral wetlands, suffering from a high mosquito nuisance, priority was given to anti-larval control and the fight against adult stages was only used in the event of failure of the latter, or to ensure the protection of tourist resorts from the clouds of adults coming from neighboring untreated sites (e.g., natural reserves). This was the case of La Grande Motte and Port-Camargue resorts, regularly invaded by mosquitoes from Camargue. In a few years, this strategy significantly reduced this problem with a minimum number of negative effects on the surrounding flora and fauna. At the same time, the fight against the domestic species C. pipiens was based on the screening and mapping of urban and periurban larval biotopes (cellars, sewers, septic tanks, ornamental ponds, rainwater receptacles, containers, etc.). In most cases, filling of larval habitats (e.g. sanitary voids), their cover by polystyrene beads (e.g. temporally flooded cellars) or even their definitive removal (e.g. abandoned containers) were sufficient to control the problem. Insecticides were used only as a last resort because of resistance phenomena. In addition, education and public awareness played a major role in this control program. Conceived more than half a century ago, the "Entente Interdépartementale pour la Démoustication du Languedoc-Roussillon" has remained remarkably effective and is an excellent example of successful integration between politicians, scientists and operators [14]. It resulted from the competence of the teams gathered by Jean-Antoine Rioux and his successors. This structure became a leading scientific center, both for French and foreign students and, also for experienced researchers. Moreover, after its success in the 1960s, the method was implemented in other regions of France (e.g. EID Atlantique, Rhône-Alpes, Alsace). In Guadeloupe, still in collaboration with the EID, a phyto-ecological map of the mangrovial larval habitats was drawn at a 1:25000 scale. Several countries later asked for assistance from the Montpellier group, including Tunisia, Morocco, Spain, Greece, Cyprus, and Canada.

Plague and schistosomiasis
Before discussing his contribution to the study of leishmaniasis and introducing the corresponding studies on pathogenic complexes, let us mention two programs, not targeting vectors but rather reservoir hosts of the plague and schistosomiasis. These programs were relatively limited in duration, but produced highly significant results. The first, a project carried out in his early career, was conducted with Yves-Jean Golvan and the Pasteur Institute of Teheran, Iran in 1960 [2]. It dealt with the Gerbillidae reservoirs of the enzootic plague in Iranian Kurdistan (Meriones spp.). This study, conducted on an annual biological cycle, identified the biotopes of these major rodent reservoir species (spatial localization of populations, geomorphological characteristics of their burrows, specific trophic preferences, phyto-ecological mapping of risk areas), and defined the roles of susceptible species (Meriones persicus) and resistant species (M. vinogradovi) between which the bacillus circulates. At the end of the survey, traditional agro-pastoral practices (extensive pastures, "dry" grain farming) were considered to be major determinants of plague foci. The second study began in 1972 under the authority of the Délégation Générale à la Recherche Scientifique et Technique (DGRST). Jean-Antoine Rioux had the operational responsibility of a project focused on the integrated control of intestinal schistosomiasis in Guadeloupe, French Caribbean. This program involved multiple teams of specialists: ecologists, epidemiologists, parasitologists, malacologists, water biologists, sociologists, and physicians [6]. The watershed, with its hydrographic network, its vector snails and its human occupation, constituted the unifying element (epidemiology of the landscape). After ten years of research in natura, accompanied by integrated interventions (management of contaminated watercourses, detection and treatment of clinical cases and healthy carriers, sanitation, and health education), anthroponotic transmission was successfully interrupted. However, some sylvatic micro foci (e.g. mangroves), involving black rats as reservoirs, remained active (zoonotic transmission). As a result, continued surveillance was recommended. Despite their apparent diversity, all these research themes are underpinned by the concepts and methods of general ecology applied to classical epidemiology. The work carried out has led to the individualization of a new discipline: eco-epidemiology, now recognized by several organizations in France, such as the CNRS, and international bodies including the WHO, FAO, and OIE. Most of his research studied only one of the members of the "pathogenic complex", either the pathogenic agents themselves or their vectors or reservoirs. Other works have dealt with more finalized research, such as the "integrated control" of nuisance mosquitoes. A major contributor to the study of leishmaniasis However, from the 1960s, the dominant theme of J.A. Rioux's work was the area that enabled him to express the fullness of his talents and earned him international recognition: the in depth study of the pathogenic complex of leishmaniasis, a major public health problem in the whole Mediterranean region in particular. Three epidemiological types of leishmaniasis were and are still present in this area: the zoonotic visceral form due to Leishmania infantum, the anthroponotic cutaneous form due to L. tropica and the zoonotic cutaneous form due to L. major.
In France, L. infantum was already known in humans and dogs in the South of France from the Spanish to the Italian borders. However, in the peri-Mediterranean foci, the 'true' vectors remained to be discovered. The presence of abundant sand fly populations led him to address some of the major issues that are fundamental in parasitology such as Leishmania behavior in vectors, vicariant vectors (respective roles of usual vs. accidental vectors) and the systematic and eco-physiological specificity of the vector/ vertebrate and vector/parasite pairs. Finally, the research conducted in active leishmaniasis illustrated particularly well the "eco-epidemiological" approach dear to the neo-Hippocratic School of Montpellier University. From the beginning of the 1960s, under the auspices of INSERM, a research project on leishmaniasis epidemiology was initiated in the French Mediterranean area. The project included the study of four components: ecology of the parasite cycle actors, ecology of transmission, identification of leishmaniasis risk factors at the spatiotemporal and population levels, and proposals for "integrated control" adapted to each cycle and to each focus. The geographical area of the study ranged from the Cevennes in the North of the focus (Aigoual-Lozère-Espinouse) to the South (coasts of Camargue and Bas Languedoc). The geographical area was divided into epidemiologically homogeneous strata, defined by "zoning indicators" such as altitude, phytoecology, and bioclimatic data. In each stratum, simultaneous sampling of the organisms involved in the cycle was carried out: vectors (Phlebotomus spp.), reservoirs (domestic and wild Canidae) and parasites (Leishmania spp.). During the three decades of this program funded by the INSERM and CNRS, four features highlight this research particularly well. The first was the ecological analysis of the sandfly vectors, involving systematic identification of species, development of trapping techniques to establish their relative densities by biotope, spatial dispersion, trophic behavior and physiological age; the detection of natural infections and intra-vector dynamics of experimental infection; and finally the establishment of laboratory breeding. The second feature was the analysis of disease distribution and prevalence in the canine population by immuno-fluorescence assays. The third was the search for wild animal reservoirs, for example in foxes or rodents. And last but not least, the identification, classification and phylogenetic taxonomy of Leishmania based on phenetic, particularly isoenzymatic, analysis. In the Cevennes, one of the main results was the demonstration that the vector role was usually assumed by P. ariasi, and not by the locally more abundant species, P. perniciosus. The installation and persistence of L. infantum was shown to be based on the existence of bioclimatic conditions favorable to P. ariasi. One of the great advances, indeed, of this research, was the use of phyto-ecological analysis, where the vegetation layers and bioclimatic variables were essential witnesses of the presence of the vector. These concepts, logically arising from the multidisciplinary knowledge (botany, systematics, and eco-epidemiology) of the eminent naturalist that Jean-Antoine Rioux was, constituted a major breakthrough in the knowledge of the leishmaniasis ecoepidemiology [5,13]. Subsequently, these conclusions could be used to analyze many peri-Mediterranean foci in the Maghreb, Middle East and South America. The natural infection of the vectors was detected from samples taken from areas where canine disease was detected. On this occasion, the delicate techniques of sterile dissection and cultivation of the parasite directly from sandflies were developed in the field. These results were of great interest for the identification of the "true" vectors of Leishmania in numerous Mediterranean foci: L. infantum (Pyrénées-Orientales, Spain, Algeria, Syria, Crete), L. donovani (Syria), L. tropica (Morocco) or L. major (Morocco). The trophic preferences of P. ariasi were determined, demonstrating exophagous and anthropophilic behavior. At the same time, the cynophilic preference of this species was demonstrated, as dogs held on a leash near the bivouacs were regularly targeted. The identification of blood meals completed these results. Considered as unable to fly long distances, the role of sandflies in the dissemination of leishmaniasis was thought to be negligible. Nonetheless, several "mark-release-recapture" experiments unambiguously established that, in Languedoc, dogs were not solely responsible for long-distance transport of L. infantum but shared this role with the vector P. ariasi. A survey was also performed on the vertebrate reservoir to establish the regional prevalence of canine leishmaniasis in the Languedoc-Roussillon area, in comparison to the distribution of the vector P. ariasi, and to track possible vulpine or murine reservoirs. In addition, the contamination of a healthy dog through the blood meal of infested sandflies was carried out at the end of the 1970s, definitively establishing the complete L. infantum cycle in a laboratory. To confirm the hypothesis of Percy Cyril Claude Garnham on the possible sylvatic origin of L. infantum, an investigation was conducted on the spontaneous and experimental infection of the red fox (Vulpes vulpes). Spontaneous infection was detected as early as 1968. The experimental infestation was carried out on two foxes bred in a non-endemic region. Subsequently, foxes infected by L. infantum were found in several southern European countries. Therefore, in the area of L. infantum, the cycle of the parasite could be regarded as primitively sylvatic (Garnham's "primary cycle"), with a wild canid as a reservoir. Secondarily, this cycle was anthropized through a passage to the domestic dog (Garnham's "secondary cycle"). An intermediate stage was readily imagined, involving both wild and domestic hosts (Garnham's "primary-secondary cycle"). Finally, the human host usually remains a "parasitic dead-end" because leishmaniasis has a low prevalence in human populations and because sandflies cannot readily be infected directly by biting humans (except when co-infected with HIV). Finally, the study completely met its objectives, entirely elucidating the zoonotic cycle of L. infantum, which then appeared as an excellent model of parasitic ecology. Many basic but novel concepts had been developed during this research project: pathogenic complex, landscape epidemiology and phyto-ecological zoning, compartmentalized cyclic system, and multifactorial and spatial-temporal evolution of foci (changes in climate and socio-economic conditions). Moreover, for the first time in field parasitology, the concept of a transdisciplinary approach had proven its efficiency [13]. Nevertheless, the precise identification of the pathogen, a fundamental element, was lacking. Leishmania species are morphologically indistinguishable but show quite different nosological and epidemiological features. Therefore, to further complete the "bio-geographical" analysis of transmission foci, a rigorous and unmistakable identification of circulating parasite populations was required. Morphologically identical, parasites could not be distinguished in culture, as is the case for bacteria, for example. Approaches based on inoculation in animals did not lead much further than those using monoclonal antibodies. However, a new tool based on isoenzyme analysis made it possible to clarify the systematics of the large Leishmania genus. Immediately after the publication of this technique, the most robust method for the time, by Michael Chance [1], Jean-Antoine Rioux developed it, with the help of Francine Pratlong, to make Montpellier the world reference center in this field. The acronym MON, for Montpellier zymodeme, remains until today a WHO recognized reference for the classification of Leishmania strains and species. Thus, for a decade, using this novel and rigorous tool, his team deepened the taxonomy and systematics of Leishmania, culminating in 1990 with the publication of a complete revision of the genus, which still remains the reference today, with more than 500 citations [9]. An "International Center of cryopreservation, enzymatic identification and taxonomic study of Leishmania" was also created, which, without going into details, consisted of four interconnected units devoted to cultures, preservation of cryostabilates, identification and classification. Finally, a biobank of isolates was created, supported by the CNRS and officially recognized by the WHO. When Jean-Antoine Rioux retired, nearly 2,500 strains of Leishmania were preserved and rigorously identified and designated by official codes (LEM and WHO). These expert activities led to the creation in 1998 of the National Reference Center for Leishmaniasis by Jean-Pierre Dedet, his successor. However, long after his retirement, taxonomy and systematics remained Jean-Antoine Rioux's preferred disciplines and, until his last days, he continued his studies and deepened his thoughts on the topic and continued to publish on systematics, epistemology, and the history and philosophy of science. Open-minded and strongly interested in the advances of molecular biology, he remained hopeful that a definitive classification of Leishmania, bringing together the inputs of isoenzymatic phenotypes and DNA data would be developed.
The study of two nearby foci, Corsica and Catalonia, confirmed the validity of the concepts implemented in the Cevennes. The first was known to house enzootic canine leishmaniasis and sporadic cases of human visceral leishmaniasis. The canine infestation was systematically searched for in about 30,000 animals, and the Leishmania species cultivated and identified. At the same time, sandflies were screened throughout the island. The results were very similar to those found on the French mainland. These studies in Corsica enabled fruitful and frequent exchanges with the Italian team of Ettore Biocca and Alberto Coluzzi working in neighboring Sardinia. However, in Catalonia, the Leishmania "pathogenic complex" was different from that of Languedoc by the relative abundance of P. perniciosus compared to P. ariasi, and by the presence of P. sergenti, a recognized vector of L. tropica. A few years later, strictly cutaneous human leishmaniasis, known as "autochthonous oriental sore", proved to be common in the Pyrénées-Orientales department of France. Moreover, the enzymatic taxonomy, developed in the meantime, linked these lesions not to L. tropica, as many assumed, but to L. infantum. More importantly, besides the presence of the classic zymodeme MON-1, several other zymodemes were detected, some unknown until then such as MON-29, MON-11 and MON-33, particularly related to cutaneous forms. Several field surveys, on both sides of the French-Spanish border, clarified the characteristics of this focus. As with the Cevennes survey, the prevalence of canine leishmaniasis was established. The canine infestation was confirmed by the isolation of 27 strains of Leishmania, all reported to belong to the MON-1 zymodeme. In addition, 13,635 sandflies were harvested by adhesive traps following different transects and identified as belonging to the genera Phlebotomus and Sergentomyia. The prevalence in humans was determined using the Montenegro test which was carried out in 718 children. Statistical analysis showed a good correlation among the altitudinal prevalences of the three components of the leishmanial complex. These prevalences were highest between 300 and 600 m above sea level and lowest below 150 m and above 600 m. In the field, the maximum prevalence corresponded to the Quercus ilex and Q. pubescens mixed forest level which was then considered in the Cevennes as a "risk zone". Dissections of P. ariasi and P. perniciosus carried out in the vicinity of Céret (Pyrénées-Orientales) led to the isolation and identification of two of the zymodemes (L. infantum MON-1 and MON-29) already observed in autochthonous human cases of cutaneous leishmaniasis. The same results were obtained in a comparable survey in Spanish Catalonia. However, the search for Leishmania in small mammals (such as Rattus rattus, a species implicated in the L. infantum cycle in Italy) remained negative.
The successful application of the concepts imagined by Jean-Antoine Rioux on the circulation of Leishmania parasites to a geographically close but structurally different focus, confirmed their validity. Their universality was then proven by their use for a large number of other foci and by their popularity in the international Leishmania community. Tirelessly challenged and refined by Jean-Antoine Rioux, they remained the basis of the work that he never stopped publishing until his last months. In the same way, the Center for cryopreservation, enzymatic identification and taxonomic study of Leishmania, the founding element of the National Reference Center for Leishmaniasis, was the subject of fruitful studies throughout the whole career of its designer, and was passed on to his successor at his retirement. Led by a team of skillful scientists and technicians, constantly improving, it was an essential tool of the work carried out by what had become "the School of Montpellier" and by all those who collaborated with it, from France or all over the world. Jean-Antoine Rioux detailed the results obtained in the study of the ecology of leishmaniasis in the South of France in 22 papers published in the Annales de Parasitologie Humaine et Comparée and Parasite from 1967 to 2013 [4,15]. Many research programs were conducted in Europe (France, Spain, Italy, Cyprus), North Africa (Morocco, Algeria, Tunisia, Egypt), the Middle East (Yemen, Syria, Iraq, Oman), Saharan Africa (Chad, Senegal) and Latin America (Colombia, Ecuador) and their analysis confirmed the value of all the concepts developed by J.A. Rioux. It is not possible to detail here all these collaborative research initiatives. Let us illustrate one of them, in Morocco, which lasted for 30 years. It was made possible by an exemplary cooperation which resulted in the understanding of a very complex situation and the implementation of control and prevention measures.
In Morocco, as in most peri-Mediterranean countries, leishmaniasis has always been an important public health problem [11]. Visceral or cutaneous, zoonotic or anthroponotic leishmaniasis are present in Morocco from the North to the South, from the Rif cedar woods to the Anti- Nineteen species were identified, including several so far unknown species in Morocco and at the world level. The densities of each species, expressed in frequency classes per station, were processed by factorial analysis of correspondences. Using thematic maps, each station could be related to its phyto-ecological level and bioclimatic zone. Each of the species considered a vector of Leishmania was accurately located. Once again, the preeminence of the climatic factor in the geographical distribution and abundance of the vectors and, consequently, in the distribution and strength of infection of the foci appeared indisputably. But beyond the entomological survey, a long path remained before it was possible to achieve the structural and dynamic knowledge of the various Moroccan foci. Two decades were required to evaluate the epidemiology of visceral and cutaneous leishmaniasis in Morocco. Most of the local epidemiological cycles were determined after enzymatic identification of Leishmania species in vectors, animal reservoirs and humans. For each focus, potential risks were assessed, including those related to potential climate changes. Finally, "integrated control" was implemented in several pilot sites. Led by the Moroccan Ministry of Public Health, these operations were enhanced through the establishment of an epidemiological monitoring observatory. Each of the three Leishmania species found in Morocco was specifically studied. The first Moroccan outbreak of human cutaneous leishmaniasis due to L. major was reported at the end of the 1970s, in the south of Anti-Atlas, in the district of Tata. The 51 isolates obtained from human samples were identified as L. major MON-25, a zymodeme observed throughout Maghreb, from Morocco to Libya. A Gerbillidae reservoir and a vector of the subgenus Phlebotomus were also identified. Out of a total of 484 rodents, only Meriones shawi grandis was infected with L. major MON-25 (12 isolates identified with an overall prevalence of 14%). The infection, strictly cutaneous, could persist for several years without apparently affecting the animals. However, an intense parasitic diffusion was observed at the end of the course. This long skin tolerance, followed by a short phase of high release, gave Meriones shawi the quality of the "real reservoir" of L. major [7]. Its optimum biotope was made up of circular pits dug by man in the vicinity of the dwellings. Originally used for the manufacture of raw clay bricks, these excavations were subsequently used as dumps and latrines. Becoming detritivore and coprophagous, the rodents behaved like a dangerous commensal. At the same time, vector infection was identified 70 km east of Tata. A total of 1,673 Phlebotomus sp. females were dissected and only P. papatasi was found carrying L. major MON-25. To control the disease, it was therefore proposed to physically destroy M. shawi biotopes, to use rodenticide campaigns, and to regularly monitor the density of rodent populations. In parallel, human mass screening and treatments were recommended to accelerate healing of the lesions. In southern Morocco, the range of the other species, L. tropica, covered 40,000 km 2 . On the northern slope of the Upper Atlas, the "risk zone" extended from east to west, in a continuous fringe, between an altitude of 500 and 1,000 m above sea level. This area corresponds to the semiarid layer biotope of the Barbary thuja. In most inventoried sites, human cutaneous leishmaniasis was reported on the "rural mode", i.e. hypoendemic, dispersed and unstable (mean incidence 5%). The importance of enzymatic polymorphism of L. tropica in the Upper Atlas focus (MON-102, MON-107, MON-112, MON-113, MON-122, MON-123) and the distribution of zymodemes in three distinct phenetic sub-groups, suggested successive inputs of Oriental origin, rather than in situ differentiation from a Moroccan ancestor. Diffuse and pseudo-lepromatous infiltrative forms, already observed in Yemen and Syria among the elderly, were also observed in Morocco in the earliest surveys. These lesions, with a prolonged course and rich in amastigotes, constituted an important source of parasites available to the vector. In these specific cases, L. tropica cycle appeared to be of an anthroponotic nature. In the Upper Atlas foci, as in the north of Syria, the domestic dog was seen with strictly cutaneous lesions due to L. tropica. However, the animal did not appear to constitute a "real" reservoir, despite its high infestation rate (12%) and its contamination by MON-102 and MON-11 zymodemes also observed in humans. In dogs, the skin lesions were in fact discreet and transient and no cases of generalization to viscera were observed. It should be noted that in L. tropica Moroccan foci, the dog was often parasitized by L. infantum MON-1. Furthermore, P. sergenti, a recognized vector of L. tropica, was abundant in all Moroccan foci. The isolates obtained from the vectors were typed as MON-102, MON-107, MON-122 and MON-123; the two latter zymodemes had never been observed in humans or dogs. As for L. major and P. papatasi, the period for transmission of L. tropica by P. sergenti was done at the end of the hot season. The recommendations to control cutaneous leishmaniasis due to L. tropica did not include insecticide control because of the hypo-endemic nature of the disease, its "dispersed" distribution, difficulties of access to the contaminated douars and the cost of operations. Therefore, control was based on active case detection and subsequent treatment of lesions. This strategy, which was successfully tested in the Tanat district, could be generalized to all foci. In Morocco, as in the other Maghreb countries, human and canine visceral leishmaniasis had been known for several decades. If the distribution and importance of zoonotic foci were well known, the identification of the "true" vectors and the existence of human skin forms remained debated. Most of them came from moderate-altitude areas of the Rif and of the Atlas. The involved species was L. infantum MON-1. However, L. infantum was not only responsible for visceral forms. In 1990, about 100 cases of human cutaneous leishmaniasis were detected in the district of Taghjicht and were due to L. infantum MON-24, a zymodeme usually responsible for sporadic cases. The entomological investigation showed that P. longicuspis s.l. was the only sandfly of the subgenus Larroussius present at the site infected by promastigotes (strain not isolated). Finally, the presence of L. infantum was proven on almost all the Moroccan territory. This extensive distribution was explained by the presence of three vectors, P. ariasi, P. perniciosus, P. longicuspis, whose bioclimatic preferences and consequently geographical distributions differed. P. perniciosus, was predominantly present in the semi-arid/ sub-humid Rif and the Middle Atlas. P. ariasi, accompanied it in the wet/sub-humid zones of Rif and on the northern slopes of the Upper Atlas. P. longicuspis and its possible twin species, preferring arid/peri-arid zones, cohabited with the previous species, but were observed alone in the arid, warm and Saharan zones. In other words, in Morocco, the North-South succession of vectors and their association with different bioclimatic zones allowed L. infantum to cover the whole territory.

Conclusions
In the course of these investigations, many new concepts and techniques were developed and clarified such as analysis of temporo-spatial risk, phytogeographical cartography, role of anthropization (impoverishment, desertification) in the emergence of new foci, "integrated control" and, recently, the role of climate change. We must also stress the quality of the scientific and personal relationships established with the countless teams of the countries in which these programs were carried out: the striking enthusiasm of Jean-Antoine Rioux, his charisma and his encyclopedic knowledge compelled admiration of all researchers. His field missions, particularly in the Cevennes, also attracted a number of foreign scientists, on whom he left an indelible imprint. Two international symposia, held in Montpellier in 1974 (under the auspices of INSERM) and in 1984 (under the auspices of CNRS and WHO), crowned these two decades of research in ecoepidemiology and taxonomy of Leishmania. They brought together more than 100 participants from all over the world (UK, USA, Belgium, Germany, Italy, Spain, Portugal, Iran, Israel, USSR, Brazil, Venezuela, Colombia, Morocco, Algeria, Tunisia, and others), among whom were prominent experts either in epidemiology or in biochemical and molecular typing. The proceedings of these two symposia are still today the most important references in the field [5,8]. J.A. Rioux was a member of numerous scientific societies and academies, among which the "Académie Royale de Médecine de Belgique", "Académie Royale des Sciences d'Outre-Mer (Brussels), and the "Académie des Sciences et Lettres de Montpellier". Among the many awards received in recognition of his work, let us quote the Marchaux prize of the French National Academy of Medicine (1969), the Silver medal of the French Society of Parasitologists (1999), and a special award presented during WorldLeish 5 (Brazil, 2013) "in recognition of his support and estimable contribution to the field of leishmaniasis", as well as a posthumous tribute in WorldLeish 7 (Spain, 2017). The results from Jean-Antoine Rioux's research presented here are obviously far from being the only ones this tireless researcher obtained. He published or participated in the publication of more than 500 papers, to which must be added books and the supervision of many PhDs. To use the words of one of us (Patrick Bastien, his current successor at the head of the Parasitology Department in Montpellier University): "Mr. Rioux was one of those men whom one crosses only too seldom and who cannot be forgotten once you knew him". President of the French Society of Parasitology from 1982 to 1999, Jean-Antoine Rioux brought the society to the rank it currently occupies, in particular by providing it with modern bylaws. He was able to motivate the members to organize the 7 th International Congress of Parasitology (ICOPA7) in Paris in 1990, which brought together more than 2,500 participants. To conclude, we should add that this brilliant teacher, an impeccable French language writer, was also a man of culture, and that he never ceased to be, for those who accompanied him throughout these years, a friend equally unforgettable.

Conflicts of interest
The authors declare no conflict of interest A note about the collections of Jean-Antoine Rioux All collections made by Jean-Antoine Rioux have been transferred to the laboratory of Parasitology of Faculté de Pharmacie, Université de Reims Champagne Ardenne, Reims, France, where they will be curated by Prof. Jérôme Depaquit.