Volume 63 1954 > Volume 63, No. 1 > Portulaca - a specialty in the diet of the Gilbertese in Phoenix Islands, Central Pacific, by I. G. Turbott, p 77-86
                                                                                             Previous | Next   

- 77

IN the Phoenix Islands on Manra Island, Orona and Nikumaroro, natives from the Gilbert Islands have built their homes under a Government sponsored settlement scheme. Mr H. E. Maude, O.B.E., in an article ‘The Colonization of the Phoenix Islands’ which appeared in Vol. 61, Nos. 1 and 2 of the Journal of the Polynesian Society with typical thoroughness described the Settlement Scheme which he had pioneered.

During 1951-1952, I lived some 6 months on the 3 isolated outer Phoenix Islands which are part of an Administrative District of the Gilbert and Ellice Islands Colony. As a result of a detailed survey into the diets of their Gilbertese inhabitants, particularly on Manra and Orona Islands, and with the help of the Department of Scientific and Industrial Research in Wellington, N.Z., I now submit this article on the particular value of portulaca in the diet of the Phoenix Islands settlers followed by a brief description of the methods by which it is prepared before eating.

At Manra Island the land-locked lagoon is too salty for fish to live in. Reef fish, except at the places furthest from the settled area, are for the most part poisonous. Deep sea fish are hard to obtain owing to dangers from the fast currents which run almost continuously through the Phoenix Group. And even more important, the lagoon has apparently sunk considerably from its old level leaving many acres which subsequently were either planted with coconut or on which vegetation grew naturally. It is possible also that the salt from the lagoon has seeped into the coral atoll itself so that the normal Gilbertese diet plants, pandanus, ‘babai’ (a large coarse (Cyrtosperma Alocasia) bread fruit and small berried trees die or produce no fruit. Only the - 78 coconut grows adequately and this even in a more scraggly form. It was obvious therefore that on Manra and to a lesser extent Orona and Nikumaroro (on the latter Islands the lagoons are not land locked, fish is abundant, and diet trees are growing slowly), the settlers from the Gilberts had found some new major food to augment their meagre diet. In a previous article ‘Diets, Gilbert and Ellice Islands Colony’, published in Vol. 58, No. 1, March 1949 at p. 41 mention was made of the use of Portulaca or ‘te boi’ (Gilbertese translation) in the Phoenix more as an emergency food than as at present an essential part of the diet. Later figures for amounts of food eaten on Manra will show the extent to which it is now used.

CLASSIFICATION: The ‘te boi’ or Portulaca grows profusely on all 3 Phoenix Islands mentioned, also at Canton Island. It was from Canton Island that I collected fresh and dried samples of the plant and personally carried them to New Zealand by aeroplane. There arrangements had been made for immediate onward transmission to the Department of Scientific and Industrial Research in Wellington, N.Z. Thus within a few hours the specimens had travelled many thousands of miles. The Dominion Laboratory of the Department of Scientific and Industrial Research acted at once and I am deeply grateful to them for their interest and for the complete analysis which they carried out. At the same time samples were sent to the Botany Division of the Department of Scientific and Industrial Research in Wellington.

The plant was at first referred to as Portulaca—Canton Island. Its description fits apparently on first examination, that of Portulaca oleracea as given on p. 151 of ‘The Flora of the Malay Peninsula’, London, Vol. 1, by Ridley, H. N. (1922). Mr Maude in the article previously mentioned on the Colonization of the Phoenix Islands, refers to the use of this plant as follows: ‘a fleshy green pig weed or boi (Sesuvium portulacastrum) which was found to be an invaluable emergency food during the early days of settlement’. I suggest that Mr Maude received a wrong identification. Sesuvium portulacastrum has stems rooting at the nodes, petaloid sepals which are greenish outside, reddish within and furnished with a small spur on the outer surface below the apex. Its seeds are smooth.

- 79

The Portulaca—Canton Island on the other hand, grows to a height of 6 to 12 inches rather than purely a fleshy prostrate creeping plant, and has a yellow flower rather than the rose pink of Sesuvium portulacastrum as described at p. 866 of Ridley's ‘The Flora of the Malay Peninsula’.

After further detailed study the Botany Division of the Department of Scientific and Industrial Research in Wellington, N.Z., stated it was almost certain that the specimens I had collected would be classified as Portulaca lutea. Portulaca lutea has stems not rooting at the nodes and fleshy roots. There are 2 sepals, much smaller than petals and green in colour. The petals are yellow, more or less notched at the top with the margin forming a small apiculus. The seeds are stellately rugulose. This description fitted that of the Canton specimens which in turn were similar to the Portulaca on the other Phoenix Islands. It is further interesting to note that in a recent publication on the plants of Bikini (Marshall Is.) that Portulaca lutea had been identified as the main species of Portulaca on Bikini and other northern Marshall Islands, a group some 400 miles to the north of the Gilberts. As in the Phoenix Islands, the Portulaca in the Marshalls is found in ‘somewhat considerable patches in sterile open ground’ and is described as ‘stout, with erect fleshy stems’.

ANALYSIS AND FOOD VALUE: The last analysis of Portulaca oleracea available in New Zealand was dated as far back as 1887 and 1897.

The following is the analysis and brief report made on the stems by the Dominion Laboratory of the Department of Scientific and Industrial Research, Wellington, New Zealand, on the samples which I carried from Canton Island. The analysis is dated 11th November, 1952.

Portulaca—Canton Island (1) Fresh, (2) dried. Analyses:—

  (1) Per cent. (2) Per cent.
Water 86.5 16.1
Protein (N × 6.25) 1.03 5.2
Ash 1.9 11.0
Fat (ether-soluble extract) 0.2 1.0
Crude Fibre 1.2 9.2
Ascorbic acid, mg/1009 10.9 3.3
- 80

‘The plant affords a reasonable supply of ascorbic acid and this is probably its chief value as a food where other greens are lacking. The ascorbic acid content is approximately that of new potatoes and many other vegetables with the exception of the Brassicas.

‘Drying of the material causes loss of most of the ascorbic acid.’

For immediate reference and to illustrate the obvious value of the ascorbic acid or Vitamin C element of Portulaca I would quote here an extract from an article on vitamins by A. HN in Encyclopoedia Britannica 14th Edition, 1929.

Vitamin C (the antiscorbutic Vitamin)—The cure for scurvy was long recognized to be fresh food but quantitative experiments have shown that the vitamin occurs in very varying proportions in different food materials. Its richest sources are green leaves, especially of the cabbage, the juice of citrus fruits (lemon, orange, and grape fruit), the tomato and certain roots such as the swede turnip. On the other hand meat and milk, the potato and many vegetables and fruits usually only contain the vitamin in comparatively small amount. It is absent from seeds but is produced on germination (1912, Furst). Different animals vary greatly in their requirements of this vitamin. Thus guinea pigs (250-300 g.) need 100-150 c.c. of milk or 1.5 c.c. of orange juice or 1.5 grams of fresh cabbage per day, whilst monkeys, ten times their weight, require exactly the same ration; rats on the other hand require extremely little and can exist for long periods without it.

‘Of all the known vitamins, the antiscorbutic is the most readily inactivated by oxidation. This process is comparatively slow at air temperature but becomes very rapid when the temperature is raised. As a result of this a large proportion of the antiscorbutic potency of food materials is lost when they are cooked or dried. On the process of canning there is less exposure to air and some canned articles, e.g. tomatoes, are still potent. Inactivation by oxidation is greatest in alkaline and least in acid solutions. On observance of air, materials containing the vitamin may be heated to a comparatively high temperature without serious loss, especially in acid solution. On storage after this treatment, - 81 however, the potency disappears more rapidly than from the untreated materials.’

The Actual Use of Portulaca in the Phoenix Islands

During 1950-51 the Phoenix Islands experienced one of the droughts which has proved to be a hazard of the settlement scheme. Small quantities only of European store food were available. The following is the average amounts of native food consumed on Manra Island:—

In 1950:
(a) Adults:  
Coconut toddy (fresh) 2½ pints per day
Fish 2½ lbs. per week
Te boi (Portulaca) 1½ lbs. per week.
(b) Children:  
Coconut toddy (fresh) 1 pint per day
Fish ½—1 lb. per week.
Te boi (Portulaca) ¼ to ½ lb. per week
In 1951:
(a) Adults:  
Coconut toddy (fresh) 3 pints per day
Fish 2½ lbs. per week
Coconuts 10 nuts per week
Te boi (Portulaca) 2—2½ lbs. per week
(b) Children:  
Coconut toddy (fresh) 1½ pints per day
Fish 2½ lbs. per week
Coconuts 2—5 per week
Te boi (Portulaca) ½—¾ lbs. per week

In 1950 the rainfall was 6.05 inches; in 1951 30.96 inches.

In spite of the apparently very meagre diet, when all the Manra and Orona people were medically examined by a Native Assistant Medical Practitioner in 1951 he found them to be remarkably free from disease. Portulaca is generally not given to sick people as it is considered too strong a food. Babies likewise do not eat it, but it is given to young children in reduced quantities as soon as they start to walk.

When the Gilbertese first came to the Phoenix because of their restricted diet they began eating Portulaca in quantity. They did not like the taste at first. Many felt like - 82 vomiting after their first substantial meal of it but after a few days they report that the taste ‘grew on them’ until they liked it and indeed craved for it. There really is no particular taste to Portulaca. In the food mixtures described later, it is the toddy etc. mixed with it which is tasted. There is no smell.

If too much ‘te boi’ is introduced to the diet the stomach swells and there is muscular wasting. But eaten as above ‘te boi’ is a soothing food and from reports of Assistant Medical Practitioners who have lived long periods in the Phoenix, it prevents constipation and indigestion. It does produce much borborygmi.

In the Gilberts ‘te boi’ grows in lesser quantity and is used as an emergency drought food only.

The ‘te boi’ or Portulaca grows profusely on the stony back part of the Phoenix Islands where the coconut tree has not yet been planted. The land is mostly covered with scrub with bare patches between where the Portulaca grows to a height of one or possibly two feet. The Gilbertese break off only the outer stems of the plant thus allowing the plant to keep alive. As is seen from the figures for the amount of food eaten, in 1950, the reason for less Portulaca being eaten was that it was harder to obtain. In 1951 with more rains the plant grew quicker and supplies were increased. Drought did not cause the Portulaca to die out—only that its growth was not so profuse.


The Gilbertese go by canoe across the lagoon or by foot around the island to the rougher undeveloped back country to collect their supplies of Portulaca. Usually the whole family goes with woven baskets and collects sufficient for the rest of the week. The women do the actual collecting—the males sailing the canoe or reef fishing during the collection. On return to the village much of the Portulaca is prepared immediately; some is left for tomorrow's meal and some dried for future use.

- 83

The methods of preparation are:—

A. For eating stem (the part generally eaten)

Method (a) First the stem is broken into 6 in.—8 in. pieces, dropped into hot or boiling water and cooked until the fleshy stem is soft. The skin is then peeled off and the remainder mixed with fresh toddy, molasses made from coconut juice, plus a little grated coconut if available. The mixture is either a semi-thick paste or watery paste. Eaten in this form, it is known to the Gilbertese as ‘te katoka’.

Method (b) Scrape the skin off the stem while it is raw, then boil in water until soft, grind it and mix with the same ingredients as above. From experience, the Gilbertese find this the easiest method of preparation.

Method (c) Scrape the skin off the stem while it is raw, then cut into very small pieces, wash thoroughly with sea-water. Boil until soft, add coconut milk and a small amount of flour and/or molasses if available. While adding the flour and coconut milk, the mixture must be thoroughly stirred until the flour is properly cooked. The mixture can be prepared and eaten just as well without the addition of flour but to the Gilbertese palate the addition of flour is an advantage. The initial thorough washing prevents the final product from being sticky and is thus easier to eat.

Method (d) Scrape the skin off the stem while raw, cut into small pieces and boil slowly until quite soft. The soupy result is then drunk just as a soup, i.e. liquid and body are eaten.

Method (e) The ‘te boi’ is first soaked, then boiled until soft. The cooked stems are then ground thoroughly, mixed with coconut milk and grated coconut. This mixture is put aside. Molasses is then boiled by itself until it becomes thick. The ‘te boi’ mixture plus a little flour if available, is then poured into the concentrated molasses while still on the fire, and this is stirred until it becomes almost solid. The resultant browning mixture is then cooled down, rolled into pieces about the size of a cricket ball. This Portulaca preparation is very popular on Manra Island with the Gilbertese adult. But it is so sweet and hard it is not given to - 84 the young children as reports indicate that it causes their stomachs to swell resulting in diarrhoea.

B. Methods of Preparation of Dry Portulaca

The skin is peeled off the stem either before or after boiling. The stems are boiled until soft then spread in the sun until absolutely dry. The dried product will keep for many years if stored in a dry place—generally in woven baskets suspended from the rafters of eating houses.

C. Methods of Preparation of Leaves of ‘te boi’

The fleshy xerophitic leaves are picked, boiled and the sticky resultant mixture squeezed through several changes of fresh water in order to remove all the salty taste. Then it is pounded and mixed wtih grated coconut, fresh toddy or molasses and eaten as a paste.


The methods of preparation are the same as on Manra with these variations, all of which are also used to a lesser degree at Manra.

  • (a) Scrape off the skin of the stem, boil it in fresh water, or dry bake it in a Gilbertese oven until cooked. It is then pounded with wooden native mallet, mixed with grated coconut and toddy to a thick or thin watery paste according to taste.
  • (b) To prepare a dry ‘te boi’ or Portulaca
  • (i) Boil stems until soft, grind thoroughly and mix with molasses until a semi-solid mixture results. Boil again stirring vigorously while on the fire. If the mixture becomes too solid add more molasses; after several boilings remove from fire and cool. When luke warm spread out to a thickness of one-sixth of an inch to one-quarter inch on leaves. Leave until absolutely dry. In this state it can be stored for many months. The Gilbertese on Orona call this mixture ‘te tuae’. In the Gilberts ‘te tuae’ is a thick pandanus fruit base. The ‘te tuae’ with a Portulaca base is identical in appearance, and it is no doubt the memories of eating ‘te tuae’ in the Gilberts which has resulted in the
- 85
  • Phoenix Island Settlement naming the nearest equivalent food by the same name.
  • (ii) Another dried Portulaca is made by scraping off the skin from the stem, boil only until cooked then spread out in the sun for 2-3 days until almost brittle and quite dry. This may be stored in this state for many months.

To the Dominion laboratory and the Botany Division of the Department of Scientific and Industrial Research, Wellington, New Zealand.

- 86 Page is blank