ABSTRACT The Eurasian beaver Castor fiber suffered a drastic reduction in both geographical range and population size, due to human persecution, until the end of the 19th century. After the adoption of protection measures, natural expansion and reintroductions led to the recovery of this species over much of its European range. We review historical events that led to the recovery of beavers in France, and summarize the status of beavers in various river systems. Beaver establishment in France is a story of overall success: several major river systems are presently occupied, so that the species is no longer at risk in France. However, beaver recolonization took place in parallel with increasing human impacts on the environment. In addition to natural limiting factors, anthropogenic factors impeded beaver settlement in many areas. Today, beavers often occupy suboptimal habitats and, as a consequence, come into conflict with human activities. Effective solutions for preventing beaver damage include the restoration of riparian habitats to discourage crop damage and the provision of physical barriers to protect crops. Beaver populations reintroduced into France all originate from the relict Rhône population. However, in recent years, beavers from populations in neighbouring countries have been expanding into north‐eastern France. Therefore, our review of beaver origin and distribution in these countries may contribute to the development of appropriate national management strategies and towards important decisions, e.g. the decision to try to keep Rhône beavers genetically isolated, or to allow populations to mix. The recently discovered presence of North American beavers Castor canadensis in three countries surrounding France has raised an important issue. This species may out‐compete C. fiber in places where the species come into contact. A programme based on field‐trapping sessions and genetic analyses has recently been initiated in some western countries in order to eradicate this non‐native species.

Mammalian neonates depend on their mother's food supply and use a defined sequence of actions to find her mammary area. Their behavior is initially uncertain and demanding but rapidly becomes optimal. Efficient learning is thus operating in newborns. For instance, European rabbit (Oryctolagus cuniculus) pups localize the nipples through typical orocephalic movements. These movements are released by the mammary pheromone secreted in milk or by prenatally learned odor cues. During daily nursing, they also learn odors associated with the mother, supposedly with sucking as the main reinforcer. We here investigate the role of the mammary pheromone as an enforcer of early olfactory learning in newborn rabbits. In testing more than 950 pups, we show that the mammary pheromone promotes learning of neutral odorants paired with the pheromone in single and short trials. The pheromone-induced learning is efficient from birth and supports successive acquisition of distinct odorants. This reveals that a mammalian pheromone can function as a "cognitive organizer" that promotes early learning of relevant environmental cues.

Since 1994, there have been nine human Ebola-Zaire virus (EBOV) outbreaks in eastern Gabon and northwestern Congo [1Georges A.J. Leroy E.M. Renaut A.A. Benissan C.T. Nabias R.J. Ngoc M.T. Obiang P.I. Lepage J.P.M. Bertherat E.J. Benoni D.D. et al.Ebola hemorrhagic fever outbreaks in Gabon, 1994-1997: Epidemiologic and health control issues.J. Infect. Dis. 1999; 179: S65-S75Crossref PubMed Scopus (221) Google Scholar, 2Leroy E.M. Multiple Ebola virus transmission events and rapid decline of central African wildlife.Science. 2004; 303: 387-390Crossref PubMed Scopus (474) Google Scholar, 3World Health Organization. http://www.who.int/csr/don/2005_05_25/en/.Google Scholar]. A majority of them originated from the handling of ape carcasses found by local hunters [4Rouquet P. Froment J.M. Bermejo M. Kilbourn A. Karesh W. Reed P. Kumulungui B. Yaba P. Delicat A. Rollin P.E. Leroy E.M. Wild animal mortality monitoring and human Ebola outbreaks, Gabon and Republic of Congo, 2001-2003.Emerg. Infect. Dis. 2005; 11: 283-290Crossref PubMed Scopus (200) Google Scholar]. The impact of Ebola-Zaire virus on great ape density is suspected to be high [2Leroy E.M. Multiple Ebola virus transmission events and rapid decline of central African wildlife.Science. 2004; 303: 387-390Crossref PubMed Scopus (474) Google Scholar, 5Huijbregts B. De Wachter P. Obiang L.S.N. Akou M.E. Ebola and the decline of gorilla Gorilla gorilla and chimpanzee Pan troglodytes populations in Minkebe Forest, north-eastern Gabon.Oryx. 2003; 37: 437-443Crossref Scopus (57) Google Scholar, 6Walsh P.D. Abernethy K.A. Bermejo M. Beyersk R. De Wachter P. Akou M.E. Huljbregis B. Mambounga D.I. Toham A.K. Kilbourn A.M. et al.Catastrophic ape decline in western equatorial Africa.Nature. 2003; 422: 611-614Crossref PubMed Scopus (419) Google Scholar], but neither the demographic consequences of outbreaks nor the way the virus spreads within an ape population are well known. The large population of western lowland gorillas, Gorilla gorilla gorilla, monitored since 2001 at the Lokoué clearing, Odzala-Kokoua National Park, Congo, was affected in 2004, providing us with the opportunity to address both questions using an original statistical approach mixing capture–recapture and epidemiological models. The social structure of gorillas strongly influenced the spread of EBOV. Individuals living in groups appeared to be more susceptible than solitary males, with respective death rates of 97% and 77%. The outbreak lasted for around a year, during which gorilla social units (group or solitaries) got infected either directly from a reservoir or from contaminated individuals. The swampy clearing of the Lokoué site (0°54.38N, 15°10.55E) is exceptionally attractive for gorillas. During a 17 month study in 2001–2, 377 gorillas, of which 92% lived in groups and 8% were solitary males, were individually identified [7Gatti S. Levrero F. Menard N. Gautier-Hion A. Population and group structure of western lowland gorillas (Gorilla gorilla gorilla) at Lokoue, Republic of Congo.Am. J. Primatol. 2004; 63: 111-123Crossref PubMed Scopus (65) Google Scholar]. The first evidence for the presence of Ebola among Odzala apes was the discovery of an EBOV-positive gorilla carcass in June 2003, 60 km southwest of Lokoué [4Rouquet P. Froment J.M. Bermejo M. Kilbourn A. Karesh W. Reed P. Kumulungui B. Yaba P. Delicat A. Rollin P.E. Leroy E.M. Wild animal mortality monitoring and human Ebola outbreaks, Gabon and Republic of Congo, 2001-2003.Emerg. Infect. Dis. 2005; 11: 283-290Crossref PubMed Scopus (200) Google Scholar]. On October 13, 2003, two villagers from Mbandza hunting at an undetermined site inside the park got contaminated and became index cases of an outbreak that killed 29 people in 7 weeks. Between January and June 2004, 6 ape carcasses were found within a 4 km distance from the Lokoué clearing. Considering the epidemiological context, there is little doubt Ebola virus is responsible for this die-off. The observation of gorillas in the clearing was maintained during and after the outbreak until the end of June 2005. Overall, 109 distinct gorilla social units visiting Lokoué were reliably identified and monitored during a 1360 day period. We developed two open capture–recapture statistical models in which survival of group-living individuals and solitary individuals was constrained by epidemiological models both to estimate EBOV-induced gorilla mortality and to investigate the transmission of the virus. The first one, model Spillover2, assumed that the outbreak originated in multiple transmissions of the virus from the reservoir to social units [2Leroy E.M. Multiple Ebola virus transmission events and rapid decline of central African wildlife.Science. 2004; 303: 387-390Crossref PubMed Scopus (474) Google Scholar, 4Rouquet P. Froment J.M. Bermejo M. Kilbourn A. Karesh W. Reed P. Kumulungui B. Yaba P. Delicat A. Rollin P.E. Leroy E.M. Wild animal mortality monitoring and human Ebola outbreaks, Gabon and Republic of Congo, 2001-2003.Emerg. Infect. Dis. 2005; 11: 283-290Crossref PubMed Scopus (200) Google Scholar], with ape-to-ape transmission occurring only within groups. The second one, model SEIR2, assumed by contrast that ape-to-ape transmission of EBOV was prominent (Figure 1 and see Supplemental Data published with this article online). Both models adequately fitted the data, without overdispersion (parametric bootstrap, model Spillover2: ĉ = 0.99; model SEIR2: ĉ = 1.00) (Figure 2A). Comparison of the models did not reveal any clear differences, precluding the rejection of one of them (model Spillover2: AIC = 3465, model SEIR2 : AIC = 3468, see Supplemental Data).Figure 2Impact of the Lokoué Ebola outbreak on gorillas.Show full caption(A) Cumulative survival of gorillas during the study, corresponding to the probability that a gorilla alive in October 2001 is still alive at a given date. Dots are placed according to estimations performed independently for each of 135 10-day intervals (red dots, group-living individuals (Gr); blue dots, solitary males (Sol); sp2: Spillover2). The slow decrease observed before 2004 is due to normal, non-epidemic mortality or definitive emigration. The strong decrease in 2004 corresponds to the outbreak. The solid lines are placed according to the epidemiological models. (B) Instantaneous survival rates, per 10-day period, predicted by the epidemiological models. The epidemic lasted around one year, but comparison with Figure 2A shows that almost all affected gorillas disappeared during the first half of this period. Wet seasons are shown in green. (C) Number of adult males (M), adult females (F) and immatures (I) identified during 30-day periods. Continuous data collection started 2 months before the outbreak and ended 7 months after its end. Note the sex ratio reversal. (D) Number of adult males, adult females and immatures identified during 150 observation days before and after the outbreak.View Large Image Figure ViewerDownload Hi-res image Download (PPT) (A) Cumulative survival of gorillas during the study, corresponding to the probability that a gorilla alive in October 2001 is still alive at a given date. Dots are placed according to estimations performed independently for each of 135 10-day intervals (red dots, group-living individuals (Gr); blue dots, solitary males (Sol); sp2: Spillover2). The slow decrease observed before 2004 is due to normal, non-epidemic mortality or definitive emigration. The strong decrease in 2004 corresponds to the outbreak. The solid lines are placed according to the epidemiological models. (B) Instantaneous survival rates, per 10-day period, predicted by the epidemiological models. The epidemic lasted around one year, but comparison with Figure 2A shows that almost all affected gorillas disappeared during the first half of this period. Wet seasons are shown in green. (C) Number of adult males (M), adult females (F) and immatures (I) identified during 30-day periods. Continuous data collection started 2 months before the outbreak and ended 7 months after its end. Note the sex ratio reversal. (D) Number of adult males, adult females and immatures identified during 150 observation days before and after the outbreak. These analyses reveal that the outbreak started in December 2003 (Figure 2B). The mortality peaked in May 2004. Although the epidemic lasted almost one year, 95% of all affected gorillas had disappeared before late July 2004. Overall, 95% of the gorillas died from Ebola (95% confidence interval (CI): 90–97%). Due to intra-group transmission, the death rate was highest among gorillas living in groups (estimate: 97%, CI: 92–98%). Solitary gorillas were at least two times more resistant to infection (model SEIR2: 2.28 times, = 7.59, P = 0.006; model Spillover2: 2.26 times, = 7.41, P = 0.007), although the virus caused a 77% decrease in their number (CI: 62–87%). Intra-group spread of EBOV was probably rapid since only one partially affected group was observed during the outbreak. All the other groups disappeared as a unit. Whether model SEIR2, model Spillover2, or a mix of these two models corresponds to the evidence is a key point of the epidemiology of the disease. According to model SEIR2, inter-unit transmission of the virus would have been possible provided that, at the epidemic peak, the probability for a susceptible unit to get infected, per 10-day period, reached 0.22 (Supplemental data). This low value is realistic, revealing that this model cannot be disregarded, contrary to what is usually stated [2Leroy E.M. Multiple Ebola virus transmission events and rapid decline of central African wildlife.Science. 2004; 303: 387-390Crossref PubMed Scopus (474) Google Scholar, 4Rouquet P. Froment J.M. Bermejo M. Kilbourn A. Karesh W. Reed P. Kumulungui B. Yaba P. Delicat A. Rollin P.E. Leroy E.M. Wild animal mortality monitoring and human Ebola outbreaks, Gabon and Republic of Congo, 2001-2003.Emerg. Infect. Dis. 2005; 11: 283-290Crossref PubMed Scopus (200) Google Scholar]. The contamination of social units could have occurred during dyadic encounters, for example in the vicinity of fruit trees [8Bermejo M. Home-range use and intergroup encounters in western gorillas (Gorilla g. gorilla) at Lossi forest, North Congo.Am. J. Primatol. 2004; 64: 223-232Crossref PubMed Scopus (57) Google Scholar] or forest clearings [9Levrero F. Gatti S. Menard N. Gautier-Hion A. Inter-unit encounters in western lowland gorillas (Gorilla g. gorilla).Folia Primatol. 2002; 73: 157Google Scholar], or during contact with infected carcasses. Alternatively, the Lokoué outbreak could also have been driven by a massive spillover from the reservoir host, provided that this phenomenon lasted around 10 months (estimate: 322 days, CI: 130–539 days). This duration exceeds that of dry seasons previously proposed to promote reservoir-to-ape transmission (Figure 2B) [2Leroy E.M. Multiple Ebola virus transmission events and rapid decline of central African wildlife.Science. 2004; 303: 387-390Crossref PubMed Scopus (474) Google Scholar, 4Rouquet P. Froment J.M. Bermejo M. Kilbourn A. Karesh W. Reed P. Kumulungui B. Yaba P. Delicat A. Rollin P.E. Leroy E.M. Wild animal mortality monitoring and human Ebola outbreaks, Gabon and Republic of Congo, 2001-2003.Emerg. Infect. Dis. 2005; 11: 283-290Crossref PubMed Scopus (200) Google Scholar]. These results provide new insights into the epidemiology of a still largely unknown disease. In an evolutionary perspective, this study provides direct evidence that, in hominoids other than humans, group individuals face a higher disease risk. This cost has probably been an important constraint to sociality evolution in early humans [10Altizer S. Nunn C.L. Thrall P.H. Gittleman J.L. Antonovics J. Cunningham A.A. Dobson A.P. Ezenwa V. Jones K.E. Pedersen A.B. et al.Social organization and parasite risk in mammals: Integrating theory and empirical studies.Annu. Rev. Ecol. Evol. Systemat. 2003; 34: 517-547Crossref Scopus (509) Google Scholar]. In a conservation perspective, the demographic impact of Ebola virus is dramatically enhanced since it disproportionately affects females and young individuals, which are essential for population recovery (Figure 2C,D). Censuses conducted in 1994–5 revealed that Odzala-Kokoua National Park gorilla density was the highest ever recorded, averaging 5.4 ind/km2[11Bermejo M. Status and conservation of primates in Odzala National Park, Republic of the Congo.Oryx. 1999; 33: 323-331Crossref Google Scholar]. Preliminary surveys we conducted show that EBOV may have affected this population heterogeneously, with some large areas being now almost devoid of gorillas and others seeming intact. Thousands of gorillas have probably disappeared. As the impact of EBOV on apes is still difficult to control, reinforced protection of gorillas and chimpanzees is required throughout their range, especially against poaching and logging, the two major additional threats to these species [6Walsh P.D. Abernethy K.A. Bermejo M. Beyersk R. De Wachter P. Akou M.E. Huljbregis B. Mambounga D.I. Toham A.K. Kilbourn A.M. et al.Catastrophic ape decline in western equatorial Africa.Nature. 2003; 422: 611-614Crossref PubMed Scopus (419) Google Scholar]. We thank M. Bermejo, P. Labbé, E. Petit and P. Rouquet for helpful discussions, the ECOFAC program (UE), and all our field assistants for their personal investment all along this study. This work was funded by Espèces-Phares program (DG Environnement, UE), Institut Français de la Biodiversité and the National Geographic Society. Contribution ISEM No. 2006-042. Download .pdf (.3 MB) Help with pdf files Document S1. Supplemental Background and Supplemental Experimental Procedures

Since 1994, there have been nine human Ebola-Zaire virus (EBOV) outbreaks in eastern Gabon and northwestern Congo [1–3]. A majority of them originated from the handling of ape carcasses found by local hunters [4]. The impact of Ebola-Zaire virus on great ape density is suspected to be high [2,5,6], but neither the demographic consequences of outbreaks nor the way the virus spreads within an ape population are well known. The large population of western lowland gorillas, Gorilla gorilla gorilla, monitored since 2001 at the Lokoue clearing, Odzala-Kokoua National Park, Congo, was affected in 2004, providing us with the opportunity to address both questions using an original statistical approach mixing capture–recapture and epidemiological models. The social structure of gorillas strongly influenced the spread of EBOV. Individuals living in groups appeared to be more susceptible than solitary males, with respective death rates of 97% and 77%. The outbreak lasted for around a year, during which gorilla social units (group or solitaries) got infected either directly from a reservoir or from contaminated individuals.