Evolution 15(4): 407-417 (Dec., 1961)
Galápagos Tomatoes and Tortoises
Charles M. Rick and Robert I. Bowman
University of California, Davis, and San Francisco State College

One of the several unique features of tomatoes native to the Galápagos Islands—Lycopersicon esculentum var. minor (L. cheesmanii var. minor) and related forms—is seed dormancy. Less than one per cent of their seeds, which are notably smaller than those of other tomatoes, will germinate without treatment. Germination of up to ten percent can be obtained with var. minor if part of the seed coat is carefully excised (Rick, 1956). The present report deals with attempts to break the seed dormancy and, particularly, to find mechanisms in the native environment that might account for germination and establishment there of the Galápagos tomatoes.

Recently, improved germination of these forms was sought by trying methods that are familiar to horticulturists for the effectiveness in breaking seed dormancies of various plant species. We found that soaking seeds for various periods in stagnant or well-aerated water, subjection of dry or wet seeds to various temperatures or combinations of temperatures, and acid treatments were ineffective in raising emergence percentages. Following the discovery by Benedict and Robinson (1946) of the improved germination of guayule seeds effected by sodium hypochlorite treatments and successful applications of this method by Taylor (1949) to celery and by Laude (1951) to smilo grass, we tested its effectiveness on seeds of LA166, one of the most refractory of our Galápagos collections. Four lots were started: (1) control; (2) seed coats partly excised; (3) one-hour soaking in 2.6% sodium hypochlorite solution (half-strength household bleach): (4) soaking for two hours in the same strength solution. Seeds of treatments 3 and 4 were thoroughly rinsed after exposure to hypochlorite. Germination was tested on moist blotting paper kept in plastic sandwich boxes. Treatment 3 removed most of the seed coat partly exposing endosperm and embryo, while treatment 4 removed even more of the tissues. Most of the emergence occurred in the first eight days after sowing. After 15 days, the per cent germination in each lot was as follows: (1)—0, (2)—24, (3)—71, (4)—8. It is evident from these results and from the appearance of the treated seeds that damage had resulted from the two-hour treatment (4). It has been our recent experience with certain other Galápagos accessions that seeds which fail to respond to a one-hour treatment may be induced to germinate by repeated treatments of half-hour and one-hour duration. Seedlings have been observed literally to pop out of the seed coats within minutes after these repeated treatments as if they had been simply unable to break through the seed coats. The one-hour treatment has become a standard procedure in our work and has continued to effect marked improvement in germination of the Galápagos accessions and also collections of the tomato species L. peruvianum and L. chilense as well as of the nightshades, Solanum pennellii and S. lycopersicoides.

Although the hypochlorite treatment provides a neat solution to the problem of germinating Galápagos tomato seeds, it does not solve the problem of establishment in nature. Various forms of these tomatoes exist in a wide distribution over the archipelago. Though not very abundant in most populations, their distribution extends to all of the larger islands and to several localities on each as verified by our observations and by earlier collections documented in herbarium specimens. We were not able to find them in certain areas (for example, James and Hood Islands) from which they had previously been taken, but their absence can readily be explained by the plague of feral goats that infest these islands. In the course of these investigations it has been possible to test a number of mechanisms that might account for germination of the dormant seeds. Visits to the islands in 1952-53 and 1957 by RIB and in 1956 by CMR provided opportunity for making the following field observations. The great majority of the living plants are large, and it is evident from the extent of their growth that they are fairly long-lived perennials. Occasional small plants could be found in the crevices between the ubiquitous lava rocks, but their presence gave no clues as to the mode of their establishment.

The following two tomato accessions were selected for germination tests: LA166, L. pimpinellifolium, collected on Indefatigable Island 1/2 km from Academy Bay along the trail that leads from the coast to the farms in the highlands; LA317, L. esculentum var. minor, collected on Bartholomew Island about 200 m from the anchorage on the north shore. These collections, representing the two dominating biotypes in the archipelago, are described in detail by Rick (1956).