Sorva (FAO, NWFP 6)

From PlantUse English
Jump to: navigation, search
Coppen, Gums, resins and latexes of plant origin, 1995
Gutta percha

  • Extract from : NWFP 6. Coppen J.J.W., 1995. Gums, resins and latexes of plant origin. FAO, Rome. 142 p. (Non-Wood Forest Products, 6). on line





Sorva is collected as a milky white latex from the trunks of certain Amazonian Couma species.

Sorva is traded in the form of large balls or blocks, produced by the tapper from the latex by a process of boiling and coagulation. Further processing may take place either before or after export, although final processing is undertaken by the ultimate consumer industries.

Processed sorva has traditionally been used as a natural base for chewing gum. It competes in this application with other natural masticatory gums such as chicle and with synthetic gums.



Sorva has now largely been replaced by synthetic gums in chewing gum manufacture and this has led to a drastic decline in demand. The United States used to be the major importer of sorva but the main international market now is the Far East, with some limited interest in Europe.

Brazil is currently the only supplier of sorva to the market and the decline in international demand is indicated by the downturn in Brazil's exports from 3 500 tonnes in 1978 to an estimated 500 tonnes in 1992 (see Table 34).

In 1978, sorva exports were valued at almost US$ 10 million (Table 34) and for Amazonas State it was the most important export product (and marginally more valuable than Brazil nuts). The most recent value of sorva exports from Brazil is probably of the order of US$ 1.5-2 million.

Brazil has a large chewing gum industry but this, too, is based principally on synthetic gums and there is, therefore, effectively no domestic market for sorva.

Supply sources

Brazil is the sole world exporter of sorva. National statistics (Table 34) show a five-fold reduction in Brazilian sorva production over a 12-year period, from something of the order of 5 500 tonnes in 1978 to just over 700 tonnes in 1990. Peak production occurred in 1976 (just over 6 000 tonnes), having risen steadily from about 1 500 tonnes in 1960.

Amazonas State has been the centre of Brazilian production, accounting for 90% or more of the total throughout the period 1978-89. Roraima has been the only other source of sorva of any significance and in recent years has accounted for all of the balance.


Quality and prices

There are no international specifications for sorva and no formal grading system appears to exist for material which is exported.

Recent price data are not available but some FOB export values are included in Table 34. In 1988, the average value of exported sorva was US$ 3 315/tonne.


Botanical/common names

Family Apocynaceae:

Description and distribution

Couma macrocarpa is a tree up to 30 m high. It is the most widely distributed of the species - found in Peru, Venezuela, Ecuador, Colombia and the Brazilian Amazon, particularly Amazonas - and is the main source of sorva. The smaller Couma utilis (Amazonas and the upper Orinoco basin, Venezuela) and Couma guianensis (eastern Amazonia and the Guianas) are other species that yield a sorva-type latex. Some hold the view that Couma guianensis is a synonym of Couma utilis.


Sorva processing and export in Brazil is based in Manaus, a major town on the Amazon. The raw material is collected from intermediary traders based in medium-sized river towns, who in turn commission collection by groups of sorveiros; these make trips of up to three months to find and extract the gum. The trader finances both the expedition and the extractors' families whilst they are away.

All parts of the tree exude a white latex when cut. Until relatively recently the usual practice for collecting the latex was destructive: a single, oblique channel was cut in the trunk to a height of about 1 m and after draining the liquid into a cup the tree was felled. A series of further cuts was then made the entire length of the tree, sometimes including the branches, to obtain additional quantities of latex. This form of sorva harvesting, undertaken by commissioned groups of sorveiros, is not "extractivism" in the sense applied to the non-destructive collection of rubber or Brazil nuts. Instead, the destructive felling of the natural resource parallels that of Aniba rosaeodora, which is felled for rosewood oil production.

During a recent study of selected non-wood forest products in Brazil (COPPEN et al., 1994), traders stated that Couma is now tapped, although much less frequently than rubber; three times a year was claimed by one trader. It was reported that the recommended tapping interval is between 6 months and 2 years in Rondônia.


On returning from the forest, the collector strains the latex to remove any forest debris and then boils it over a fire for a short time to induce coagulation. Alternatively, the latex is first mixed with water, salt added and the mixture left to stand overnight; this initial coagulum is then boiled in water.

The crude sorva is removed from the water and formed into large balls or blocks. These are usually brown on the outside where they are exposed to the air. The sorva is then wrapped in banana or palm leaves and placed in baskets for transportation and sale to the trader or further processor.


Yields of latex obtained by destructive means have been stated to be up to 6 litres from an adult tree of Couma macrocarpa. Average production per man-day has been reported to be 15 litres of latex from three trees. More recently, it has been claimed that a Couma macrocarpa tree of 50 cm diameter could yield up to 20 litres of latex after felling.

In Rondônia, an average of 2.5 litres of latex per tree can be obtained from sorvinha [Couma utilis] and the sorveiro can tap up to 14 trees in a day. For "sorva mole" [Couma macrocarpa], where the trees are more widely scattered, four trees per day are said to be tapped, producing an average of 20 kg per tree. LESCURE (1990) has stated that 1.5-2.0 litres of latex are obtained on average from one Couma utilis tree by tapping and that a skilled man can collect 50-60 kg per day; the frequency of tapping is not stated.


Further processing (in Manaus) entails digestion of the crude sorva in water, blending of the various lots to produce a more consistent product, and drying to a moisture content of less than 1%; a little wax is added as a preservative. Although plans were made in the early 1970s to produce refined sorva, resin and trans polyisoprene ("gutta") at Manaus from a feedstock of sorva, balata and maçaranduba, the plant for doing this is not believed to have been built.


Couma utilis is amenable to cultivation for fruit production and it is grown for this purpose in parts of Amazonas. Some research has been undertaken on this aspect and one Brazilian institution (FCAP, Faculdade de Ciencias Agrarias do Para, in Belém) has a 16-year old trial plot at one of its field stations. However, no research appears to have been carried out on tapping and its effects on fruit production.


The earlier study cited (COPPEN et al., 1994) concluded that although the recent trend was one of marked decline in world markets for sorva, it seemed unlikely that this would continue. Trade sources in Brazil were cautiously optimistic that energetic marketing would lead to a small increase in demand, perhaps to around 800 tonnes/year (but not to historical levels of 3 000 tonnes and more).


However, any foreseeable increase in the volume of demand for sorva could be met by harvesting of the wild resource. A network of collectors, intermediate market buyers and exporters is in place and there appears to be no immediate pressure on the wild tree resource. Research and development work on sorva production by cultivation in an agroforestry context, either solely for the latex or in combination with fruit production, is therefore difficult to justify. Moreover, the indications are that the economics of production would not be attractive or competitive with the present form of extractivism.

Research needs

There appears to have been little or no recent systematic research carried out into ways of improving sorva production on a sustainable basis. In the early 1960s, WILLIAMS (1962) carried out tapping trials on Couma macrocarpa over a 5-month period. He concluded on the basis of his own work and that of others that it was not economic to recover sorva from this species in a manner analogous to rubber tapping. Furthermore, unlike the rubber tree, bark renewal is very slow and even after several years the channels have not healed sufficiently to allow re-tapping of the original surface.

However, if future demand for sorva is judged to warrant the investment in research, technological improvements to the present tapping methods are desirable. Such improvements should be possible by drawing on more recent experience with other latex and resin-yielding species (such as rubber and pine trees), and should be aimed at improving yields and reducing the risk of permanently damaging the trees.


  • COPPEN, J.J.W., GORDON, A. and GREEN, C.L. (1994) The developmental potential of selected Amazonian non-wood forest products: an appraisal of opportunities and constraints. Paper presented at the FAO Expert Consultation Meeting on Non-Wood Forest Products, Santiago, Chile, 4-8 July.
  • LESCURE, J.P. (1995) [Extractivism in Amazonia. Viability and Development] (in French). Final Project Report. ORSTOM/INPA/Aarhus University.
  • LESCURE, J.P. and CASTRO, A. (1990) [Extractivism in central Amazonia. An outline of economic and botanical aspects] (in French). Paper presented at UNESCO-IUFRO-FAO Workshop "L'Aménagement et la Conservation de l'Ecosystème Forestier Tropical Humide", Cayenne, 10-19 May.
  • WILLIAMS, L. (1962) Laticiferous plants of economic importance. III. Couma species. Economic Botany, 16, 251-263.


Table 34. Sorva: production and exports from Brazil, 1978 and 1986-92 (tonnes; US$/tonne)
1978 1986 1987 1988 1989 1990 1991 1992
Production 5555 3002 1524 1059 1106 736 na na
Exports 3481 1570 991 931 na na na 500
FOB unit value 2760 3255 3175 3315 na na na na

Source: National statistics (taken from COPPEN et al., 1994, and LESCURE, 1995)