Henry Doubleday Research Association, Ryton-on-Dunsmore, Coventry CV8 3LG, U.K.
Introduction
Prosopis juliflora has existed in Cape Verde for at least 50 years, and following its excellent performance in species trials, it formed the backbone of the national afforestation programme implemented by the post-independence government in 1975 to provide firewood, fodder, rural employment and to check soil erosion. Over 25 million trees have been planted, occupying a staggering 10% of the countrys surface area, and almost 70% of these are P. juliflora. However, surprisingly little work has been carried out on the selection and improvement of this, one of the most vital plant resources of the nation.
Cape Verde is an Atlantic island archipelago situated 600 km off the coast of Senegal, West Africa, in an oceanic extension to the Sahelian climatic belt. Thus despite a population of only 360,000, low rainfall and long droughts ensure that the people are still very much dependent on food aid. Irrigated valley bottoms and terraces, and the humid highlands, although small in area, account for the bulk of the agricultural productivity. Where soil conditions permit, rainfed maize and bean cropping predominates, with regular harvests in the upland sub-humid zone but only intermittent harvests in the semi-arid and arid interiors. Much of the arid coastal zone can support only poor pasture. The arid and semi-arid zones make up over 75% of the total land area with mean annual rainfall (m.a.r.) below 300 mm, compounded by a persistent, desiccating wind. It is here that much of the afforestation takes place.
Prosopis in the landscape
P. juliflora is by far the most common species, found in all zones and is the main source of firewood, preferred over all other species in terms of quality, even over Dichrostachys cinerea and Eucalyptus spp., the other popular fuelwood species. Resistance to drought, repeated cutting and browsing are vital characteristics of all firewood trees that are able to persist here. Its fast growth, ability to burn hot and burn as well when green are very much desired, but its thorniness is positively disliked. The government harvests plantations on a rotational basis and firewood is sold at the approximate equivalent of 11 Rupees (US$0.07) per kg. P. juliflora pods can provide an important fodder, and good yields are obtained with widely spaced trees, although production within plantations is very low. In urban areas, where sources of fodder are minimal, the pods of P. juliflora are in great demand. This species is very common in towns, along roads and also in rural areas, outside houses, as an ornamental and shade tree. Preliminary surveying would suggest the popularity of P. juliflora is income related, those that can afford bottled gas for cooking and do not have to raise livestock, quickly forget its value as a fuel and fodder tree. Comments concerning its monoculture, lack of aesthetic value and unconfirmed beliefs on the lowering of water tables come only from the more affluent. Rural farmers are invariably aware of its importance.
Natural regeneration from seed and as coppice regrowth is common, and within the rainfed fields it is often retained, pruned to a single stem and encouraged, primarily as a source of firewood, although it is always removed from irrigated land. This agrisilvicultural system, with P. juliflora randomly distributed at densities up to 100 stems per hectare (sph) warrants further study. Zizyphus mauritiana and Mangifera indica are often present but less common. On steeply sloping land where good rainfed harvests are common, P. juliflora can be found planted by the government in contour lines along terrace walls or in ditches, at densities around 150-200 sph, to help reduce soil erosion. The largest numbers are to be found in the poorest sites where agriculture is not possible, planted in microcatchments or along contour bunds at a regular 5 x 5 m spacing (400 sph) as a silvopastoral system with the undergrazing of goats and cattle. Four month old nursery-raised seedlings are planted out after good rains between August and October. No soil amelioration, fertilisation, irrigation or weeding are carried out.
Early species trials carried out after independence on Santiago, the largest island supporting half the population, showed P. juliflora to be rivalled only by Atriplex spp. in saline, arid coastal areas, Parkinsonia aculeata in other arid zones, and Eucalyptus spp. in the more humid regions. P. glandulosa often out-survived all other species, but was ignored due its slow growth. Only two provenances of P. juliflora were employed, even with the knowledge that the naturalised population had a very small genetic base. Widespread planting then followed. No further selection was carried out until an HDRA initiative in 1987 began testing various provenances for their adaptability to the harsh conditions. By 1991, this initiative had tested over 40 provenances, and an irrigated living germplasm collection was established, that presently contains 120 provenances of 11 confirmed species. Further field trials have been established and the evaluation of species, provenances and individual trees is being carried out. Techniques for preserving and multiplying such elite material via grafting and rooting of stem cuttings are also being developed, but are not detailed here.
Materials and methods
In 1992, 50 plants each of 24 provenances from five Prosopis species were planted, each provenance being represented by two replicates of 25 plants in a 5 x 5 lattice, randomly allocated within each of the two blocks. P. cineraria from India and P. alba, P. chilensis, P. flexuosa and P. nigra from Argentina, were planted at 4 x 10 m spacing on marginal agricultural land at an arid coastal site (m.a.r. 241 mm). The soil is a clay loam of moderate fertility and is slightly saline (1.7 mS/cm). Although the soil was moist at planting, only 30 mm rainfall fell after planting and none during the next 10 months. This drought was exacerbated by the strong, desiccating salt wind blowing persistently throughout the year. Goats freely grazed the site from 4 months after planting. In addition to survival, the length from the plant base to the tip of the longest living branch/stem was recorded, as this is considered to be the most useful growth measurement when dealing with shrubby, multi-stemmed species. In addition, form and thorniness were recorded, the form of young trees quantified using the ratio of actual vertical height to maximum branch length. Taken individually for each tree, a ratio of >0.9 would indicate an erect plant, <0.5 a prostrate tree, with >0.7 taken arbitrarily as an acceptable form. P. cineraria was always erect and was consistently thorny and was not scored for these parameters. In 1993, a further 40 provenances of 11 Prosopis species were raised in the nursery and planted out, 20 plants of each at 2 sites. Both sites are in the arid coastal zone (m.a.r. 191 mm and 229 mm) and similar to the site used in 1992, though unsuitable for agriculture, used only for rough pasture. Data for maximum branch length, height and thorn length were subjected to analysis of variance and the significance of differences between means tested with an LSD calculated using Tukeys t-test for all pairs comparisons. Survival data were subjected to Chi-square tests.
Results and discussion
Results from the 1992 trial after one year are presented in Table 1. The slow growth of all provenances of P. cineraria in comparison with the other species was noticeable in the nursery. Losses of P. cineraria in the field were high in early establishment, and the overall mean survival of the 12 provenances declined from 67% to 54% to 27% after 3, 6 and 12 months in the field, respectively. Neither growth nor survival differed significantly among the P. cineraria provenances, except between the best and worst performing provenances in terms of branch length. The highest survival recorded for a P. cineraria provenance after 12 months (36%) was considerably lower than that of the worst performing Argentinean provenance (64%).
Growth of P. cineraria after one year was also significantly lower than all the Argentinean provenances, excepting P. alba 328, but this was influenced by the browsing of many trees by goats, which may also have reduced overall survival. However, as most provenances suffered many mortalities in the first 3 months when a guard was present, and dieback was observed on most of the unbrowsed plants, environmental factors are likely to be more important than browsing. On the basis of the results of this trial to date, P. cineraria would appear to be of limited value in the arid zone of Cape Verde.
Of the Argentinean species, P. alba, P. chilensis, P. flexuosa and P. nigra, there is variation between both provenances and species at this stage, but some trends are apparent. None of these species exhibited evidence of browsing although some dieback was apparent. P. chilensis and P. nigra had the highest survival rates, with two provenances of each having greater than 85% survival after one year. Results for P. chilensis could have been higher but for farmers removing some individuals with very large thorns, this species being generally the thorniest. P. nigra 333 exhibited the greatest growth of all provenances, although this was not significantly greater than P. chilensis 339 or P. nigra 334. P. nigra 333 also had a moderate form ratio and small thorns, occasionally thornless. Nursery specimens of this provenance are to be multiplied for further trials. With P. flexuosa, provenance 336 which originates from a low rainfall, high altitude zone (m.a.r. 81 mm; 1820 m.a.s.l.), was shrubbier but had a higher survival rate than provenances 335 and 337 which are from an area of higher rainfall (m.a.r. 173 mm) and lower altitudes (1040 and 1150 m.a.s.l., respectively). Both P. nigra provenances and one provenance each of P. chilensis and P. flexuosa had significantly greater branch lengths than all three P. alba accessions tested. Low form ratios were compounded by the persistent wind which induced a leeward tilt on most trees and a spreading habit in some.
Table 1. Mean survival over 12 months, and mean growth and mean form at 12 months, of 24 Prosopis provenances under arid field conditions, ranked in order of mean maximum branch length. Means within parameters not followed by the same letter differ significantly at the p=0.05 level.
Species |
HDRA No. |
Survival % |
Max. Branch Length (cm) |
Form Ratio |
Max. Thorn Length (cm) |
||
3 mts |
6 mts |
12 mts |
|||||
P. nigra |
333 |
94 |
92 |
92 |
94 a |
0.64 abcde |
4 ef |
P. chilensis |
339 |
94 |
94 |
90 |
81 ab |
0.69 abc |
8 cde |
P. nigra |
334 |
96 |
96 |
86 |
80 abc |
0.52 e |
16 ab |
P. flexuosa |
336 |
92 |
86 |
80 |
78 bcd |
0.56 de |
3 f |
P. chilensis |
332 |
90 |
88 |
64 |
66 bcde |
0.64 abcde |
17 a |
P. chilensis |
331 |
96 |
94 |
74 |
66 cde |
0.66 abcd |
12 bc |
P. flexuosa |
335 |
84 |
82 |
66 |
65 cde |
0.75 ab |
5 def |
P. flexuosa |
337 |
90 |
86 |
66 |
64 de |
0.76 ab |
6 def |
P. alba |
329 |
98 |
98 |
84 |
62 e |
0.60 bcde |
9 cd |
P. chilensis |
338 |
100 |
98 |
88 |
61 e |
0.70 abc |
15 ab |
P. alba |
330 |
98 |
98 |
74 |
54 ef |
0.64 abcde |
6 def |
P. alba |
328 |
98 |
98 |
84 |
51 efg |
0.61 bcde |
5 def |
P. cineraria |
313/20 |
82 |
60 |
20 |
40 fgh |
- |
- |
P. cineraria |
313/17 |
42 |
38 |
26 |
26 ghi |
- |
- |
P. cineraria |
313/27 |
76 |
66 |
28 |
34 ghi |
- |
- |
P. cineraria |
313/2 |
68 |
68 |
32 |
34 ghi |
- |
- |
P. cineraria |
313/9 |
56 |
48 |
30 |
33 ghi |
- |
- |
P. cineraria |
313/1 |
48 |
36 |
20 |
30 hi |
- |
- |
P. cineraria |
313/22 |
90 |
72 |
36 |
29 hi |
- |
- |
P. cineraria |
313/25 |
76 |
62 |
22 |
28 hi |
- |
- |
P. cineraria |
313/24 |
58 |
46 |
28 |
28 hi |
- |
- |
P. cineraria |
313/14 |
78 |
64 |
20 |
23 hi |
- |
- |
P. cineraria |
313/4 |
48 |
28 |
22 |
22 hi |
- |
- |
P. cineraria |
313/12 |
76 |
64 |
34 |
22 i |
- |
- |
Plants heights in the nursery, of the provenances in the 1993 trial are given in Table 2. The Peruvian Prosopis provenances attained maximum height. These provenances are also reported to have performed well in Haiti and at CAZRI, Jodhpur, India (Tewari et al., this volume). The superior provenances amongst these are from Trujillo (8oS, 79oW), and are probably P. juliflora, and being mainly thornless, possibly var. inermis. The local provenance of P. juliflora from Cape Verde (no. 737) and another from Senegal (no. 738), also grew well in the nursery. Of the other provenances, those of P. chilensis, P. glandulosa and P. velutina also performed well. P. alba and P. caldenia, from more temperate native ranges, grew more slowly, and may be less adapted to the higher temperatures encountered here. P. articulata showed only moderate growth, but being salt tolerant and from coastal ranges at a similar latitude, may yet exhibit good establishment.
Table 2. Provenance details and nursery performance at 3 months, of 40 Prosopis provenances. Means not followed by the same letter differ significantly at the p=0.05 level; masl - metres above sea level; mar - mean annual rainfall.
Species |
HDRA No. |
Texas No. |
Provenance details |
Height(cm) |
P. sp. Peru 1 |
394 |
0431 |
Trujillo, Peru, masl 40 m |
76 a |
P. sp. Peru 1 |
396 |
0433 |
Trujillo, Peru, masl 40 m |
69 ab |
P. sp. Peru 1 |
397 |
0434 |
Trujillo, Peru, masl 40 m |
69 ab |
P. sp. Peru 1 |
395 |
0432 |
Trujillo, Peru, masl 40 m |
65 bc |
P. sp. Peru 1 |
398 |
0435 |
Trujillo, Peru, masl 40 m |
65 bc |
P. sp. Peru 2 |
382 |
0418 |
Trujillo, Peru, masl 60 m |
64 bcde |
P. juliflora |
737 |
- |
Sao Jorge, Cape Verde |
64 bcde |
P. juliflora |
738 |
- |
Richard Toll, Senegal |
64 bcde |
P. sp. Peru 3 |
422 |
0549 |
Sullana, Peru |
61 bcdef |
P. sp. Peru 2 |
381 |
0417 |
Trujillo, Peru, masl 60 m |
58 cdefg |
P. glandulosa |
374 |
0933 |
Riverside Row, California, P53R10, parent 0001 |
57 cdefg |
P. glandulosa |
463 |
0392 |
Imperial County, California |
56 defg |
P. sp. Peru 3 |
423 |
0550 |
Sullana, Peru |
55 efg |
P. chilensis |
163 |
- |
Setropa Seed Suppliers |
55 fg |
P. chilensis |
338 |
- |
Talampaya, Argentina, masl 1670 m, mar 150 mm |
53 fgh |
P. velutina |
355 |
0454 |
Riverside Row, California, P47R6, parent 0020 |
53 fgh |
P. glandulosa |
459 |
0385 |
Whitehavens, California |
53 fghi |
P. velutina |
378 |
0943 |
Riverside Row, California, P47R4, parent 0031 |
53 ghij |
P. glandulosa |
369 |
0475 |
Riverside Row, California, P52R11, parent 0001 |
50 ghijk |
P. alba |
441 |
0166 |
Thermal, California |
50 ghijk |
P. velutina |
361 |
0464 |
Riverside Row, California, P49R4, parent 0020 |
45 hijkl |
P. alba |
362 |
0465 |
Riverside Row, California, P49R11, parent 0032 |
44 ijkl |
P. nigra |
517 |
0744 |
Univ. of California, Riverside, M3V23T1, parent 0001 |
44 jkl |
P. juliflora |
739 |
- |
Doli, Burkina Faso |
44 kl |
P. juliflora |
685 |
- |
Comayagua, Honduras, masl 630 m |
43 klm |
P. articulata |
349 |
0593 |
Kingsville, Texas, Parent BK13V4, grandparent 0016 |
43 klm |
P. caldenia |
652 |
- |
Santa Luis, Argentina, masl 500 m, mar 350 mm |
41 klmn |
P. caldenia |
651 |
- |
Santa Luis, Argentina, masl 500 m, mar 350 mm |
40 lmno |
P. cineraria |
680 |
- |
Suweig, Oman, masl 1000 m, mar 200 mm |
39 lmno |
P. alba |
350 |
0591 |
Kingsville, Texas, Parent BK14V2, grandparent 0039 |
39 lmno |
P. tamarugo |
683 |
- |
Pampa del Tamarugo, Chile, masl 1200-1500 mm |
39 lmno |
P. alba |
555 |
0900 |
Univ. of California, Riverside, M1V21T10, parent 0137 |
38 lmnop |
P. alba |
513 |
0751 |
Univ. of California, Riverside, M1V21T1, parent 0032 |
37 lmnop |
P. cineraria |
678 |
- |
Sur, Oman, masl 14 m, mar 95 mm |
37 lmnop |
P. sp. |
500 |
0520 |
Univ. of California, Riverside, M3V26T10, parent 0080 |
34 mnopq |
P. tamarugo |
561 |
1116 |
Chile |
33 nopq |
P. pubescens |
505 |
0627 |
Brawley Biomass Plot, California, BK8V27, parent 0245 |
31 opq |
P. caldenia |
650 |
- |
Santa Luis, Argentina, masl 600 m, mar 600 mm |
29 pq |
P. cineraria |
677 |
- |
Bilad Hassan, Oman, masl <50 m, mar 80 mm |
28 q |
P. cineraria |
676 |
- |
Nizwa, Oman, masl 1000 m, mar 250 mm |
27 q |
Again, P. cineraria (section Prosopis) grew slowly, as did P. pubescens and P. tamarugo (section Strombocarpa). P. africana (section Anonychium) germinated but failed to survive 3 months in the nursery. All but these latter 4 species are from the section Algarobia, which contains virtually all of the species of economic importance to Africa and the Americas, and almost all are from the series Chilenses within this section. Thus these findings, allied with earlier experiences, suggest that any further work in this region should place emphasis on this section. Priority should also be given to detailed seed provenance collection from the Trujillo region of Peru for further testing. Field performance will be monitored over several years to assess whether these preliminary differences persist and whether such early growth measurements can be used as an indicator of species and provenance performance.
Acknowledgements
The research was funded as project R4733 in the Forestry Research Programme component of the Renewable Natural Resources Research Strategy of the U.K.s Overseas Development Administration.