The Spix's macaw (Cyanopsitta spixii) is the most endangered parrot in the world. The predictable extinction in the wild of this species is a special loss as it regards a monotypic species and genus. As with most other species of parrots, we have only little, if any, idea about the historical distribution and abundance of Cyanopsitta spixii; however, the species probably never was very common in its apparently restricted range in northeastern Brazil (Ridgely l982). Today, all independent reports concur that the total wild population numbers less than ten individuals and that, most probably, only three or four individuals remain in the wild (Anon. 1986. Roth 1986, Thomsen and Munn unpubl. data).

During our recent field survey of the hyacinth rnacaw (Anodorhynchus hyacinthinus) (see Munn et al. 1987, Munn et ad. in press) for the Secretariat of the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES), we succeeded in obtaining information from Brazilian parrot trappers and traders on the recent catastrophic decline of Cyanopsitta spixii. Independently of each other, four individuals affirned that the species (local name: Ararinha azul) now onlv occurs in one population in one small area in northeastern Brazil, and that onlv four individuals remain in the wild. They agreed that trapping of Cyanopsitta spixii has been the main reason for the decline, but added that not many offspring have been produced in recent years. They maintained that the lack of reproduction is not because of the trade, but because the introduced African bee has taken over more and more of the existing potential nest sites in the area.

During the last ten years, the trade in Cyanopsitta spixii from the only known breeding area has been controlled by two dealers living in the Brazilian states of Piauí and Pernambuco. From 1977 to 1985, the dealer in Piauí traded 15 individuals of which all but two were adult birds.

He claims that he was the first dealer to arrange for nestlings to be taken out of the remaining breeding area and that was during the breeding season of 1984/1985; he asserted that these two nestlings are now kept by an aviculturist in the vicinity of São Paulo. In addition, from 1984 to 1987 the dealer from Pernambuco moved eight individuals from the known breeding area; four of these were nestlings taken during the breeding seasons 1985/86 and 1986/87 from the same nest that the 1984/5 young came from. Unfortunately we did not succeed in obtaining information from the dealer in Pernambuco about his trading in Cyanopsitta spixii prior to 1984. Based on these reports, the total number of individuals taken out of the known breeding population during the past ten years is a minimum of 23 birds.

Reports by Roth (1986,1987) that the only known breeding pair in the wild did not reproduce during the breeding seasons 1985/6 and 1986/7 conflict with the detailed information provided by these individuals, all of whom claim that the young birds traded since 1984 came from the same nest in the area investigated and described by Roth. In fact, the two young Cyanopsitta spixii that were recovered by TRAFFIC (Sudamerica) in Asunción, Paraguay, in April 1987 (see Villalba-Macias 1987) and now being held in the São Paulo Zoo had been described to us as having been taken from the known nest in early February 1987 by trappers sent out by the dealer in Pernambuco. The dealer then sold the young for approximately US $ 18,000 to a dealer in Curitiba, who in turn sold them to a dealer in Asunción. In Asunción these birds apparently had been offered for sale for US $40,000 to an aviculturist from the Federal Republic of Germany.

Few other successes appear to mark the attempted recovery ofCyanopsitta spixii. We are about to lose a species and a genus from the neotropical fauna and we seem not to be able to decide what to do about it. At this point, the only future for the species is through captive breeding, but we are facing one major problem which is the fact that most of the birds held in captivity are held by private aviculturists. About half of the birds we have been able to locate in captivity are held outside Brazil, in countries such as F.R. Germany, France, Philippines, Portugal, Singapore, Spain and the United Kingdom. ICBP and the Parrot Specialists Group have put great efforts into organizing these keepers of Cyanopsitta spixii in a "breeding consortium", but without much luck. In August last year, these people were approached with an invitation to attend a meeting at Loro Parque, Tenerife, to draw up guidelines for an ex situ breeding operation involving their birds. Except for Loro Parque, which is in possession of two Cyanopsitta spixii, none of the owners presently known outside Brazil accepted the invitation.

In addition ICBP and WWF in conjunction with the Instituto Brasileiro de Desenvolvimento Florestal (IBDF) sponsored a field project, of which the primary goal has been to investigate the status of the species in the wild. In spite of the project and the visibility it has given to the species in and outside Brazil, we still have only little biological information on the species; worse, we have lost all of the offspring from the only known nest from the last three breeding seasons to local bird traders.

With these experiences in mind, it seems imperative that the situation of this species be re-evaluated and a formal "Recovery Plan " with all the necessary institutional commitments be drawn up, as has been done with some success with the Puerto Rican Amazon (Amazona vittata) and the Black-footed Ferret (Mustela nigripes). Two initiatives that seem crucial in this respect are:
i. Establishment of an in situ breeding operation in Brazil. The advantages behind such an operation are the following: at least half of the birds held in captivity are presently in Brazil; of these 15 , five are held under government control in the São Paulo Zoo. The legal and logistical problems involved in establishing a breeding consortium would evidently be easily resolved in Brazil than elsewhere.
ii. Change the scope of the field project. At present, there is no information whatsoever indicating that a large undetected population ofCyanopsitta spixii survives in Brazil; most probably there is only the known wild population consisting of three to four individuals.

Even if a second population were to be found in the wild, we would still need much more information about the species' biology, which we might obtain from the known population if a field-ornithologist were posted there. Such information will obviously be crucial to any efforts to reintroduce captive-bred individuals to the wild. We should begin collecting it immediately. It should be emphasized that currently, the most basic information on the species' breeding biology and feeding strategy is virtually unknown. Considering the limited funds available for Cyanopsitta spixii field work, it seems wise to change the scope of the field project from surveying for another population to investigating, and consequently protecting, the known population.

Obviously, had we unlimited funds available we would not be forced to make such a choice: given the circumstances, however, we cannot afford to waste any time or money.