USVI Public Records

Document text

Historic, arcliived document Do not assume content reflects current scientific knowledge, policies, or practices. -A' : 9'. ^. §omt ffkmce - INSTITUTE OF TROPICAL FORESTRY RIO PIEDRAS, PUERTO RICO FOREST SERVICE - U. S. DEPARTMENT OF AGRICULTURE CEDRELA PROVENANCE TRIAL IN PUERTO RICO AND ST. CROIX: ESTABLISHMENT PHASE'^ , By Jacob L. Whitmore January 1978 RESUMEN Tres anos despues de plantaci6n, se evaluo un ensayo de procedencia de cedro espanol. Ninguno de los cinco sitios considerados probo ser apropiado para plantaciones de esta especie. De las siete pro- cedencias experimentadas, ninguna crecio satisfactoriamente en ninguno de los-sitios. Ninguna resulto resistente al barrenador de la caoba, H\^psip\;la grandella. La importancia relativa de los dos problemas mayores en el establecimiento de plantaciones de cedro espanol (sensibilidad al sitio y ataque del barre- nador) es discutida aqui. L SUMMARY oc. Three years after outplanting, a provenance trial of Spanish cedar was evaluated. Of the five sites considered, none proved appropriate for plantations of this species. Of the sevep' provenances tested, none grew satisfactorily on any of the sites. None proved resistant to the mahogahy shoot borer, H\;psi- pyla grandella. The relative importance of the two main problems of Spanish cedar plantation establish- ment (site sensitivity and borer attack) is discussed. ^ Retrieval Terms; Meliaceae, Spanish cedar, H\;psipyla grandella, tropical silviculturej/ • Spanish cedar {Cedrela spp.^), a tree of the mahogany family and native to most of the humid neotropics, produces a cabinet wood popular for both local use and export. In view of the commer- cial importance of and interest in this species, the Commonwealth Forestry Institute (CFI) provided several research stations with seedlots of cedar provenances (see Burley and Lamb 1971, Mel- chior and Quijada 1972, and Omoyiola 1972 for details) . The Institute of Tropical Forestry (ITF) in Puerto Rico received eight of these seedlots. which were sown in August and September 1%9 and outplanted in July and October 1970. The nur- sery phase of this study was described earlier (Whitmore, 1971). This report covers the esta- blishment phase (the 3 years immediately after the nursery phase). The establishment of cedar plantations has long been a problem. Most attempts have failed, apparently because of internal soil drainage and the species' site sensitivity. Concurrent tests 1 In cooperation with the University of Puerto Rico. 2 Revised from a paper presented at the I Congreso Brasileiro de Florestas Tropicais, Vicosa, October 1974. 3 This report deals with Cedrela odorata L. {sensu lata) Specific epithets of Cedrela are confusing, however, and use of geographic origin is recommended until the taxonomy of the genus is better understood. which supplement this provenance trial are being conducted to examine this regeneration problem. Other recent studies attempt to provide alterna- tive species with similar wood characteristics (Otarola et al. 1976, Sanchez et al. 1976). A second problem in cedar plantations is the attack of the shootborer, Hypsip\jla grandella Zeller. There are several recent studies on this plant-insect relationship (Grijpma 1973, Whit- more 1976 a & b). This trial served as a prelimi- nary attempt to find provenances that might help overcome either or both of these problems under growing conditions in Puerto Rico and St. Croix, Virgin Islands. ESTABLISHMENT PHASE PROCEDURES Seedlings were raised in two nurseries, one in Puerto Rico and one in St. Croix. Mean height, branching tendency and mortality varied among provenances at the beginning of the field phase (Table 1). Seven sites were planted, three in Puerto Rico and four in St. Croix. One site in Puerto Rico was destroyed by highway construction and one in St. Croix was abandoned after the second year due to poor growth, so this report deals with the remaining five sites (Figure 1). Because of pro- blems in the nurseries (reported in Whitmore, 1971), no site was planted with all eight pro- venances (Table 2). A randomized complete block design was used with four-tree plots and 16 replications at each site. The four-tree plot, laid out 1x4 with the axis parallel to the slope, was considered a suitable compromise between the highly efficient one-tree plot and the longer-term, CFI-recom- mended 100-tree plot. This decision was based on Line's (1976) definition of first-stage provenance research, on the argument of Wright and Free- land (1961) that plots with fewer trees are more efficient, and on the facts of high labor cost and unavailability of large tracts for research in Puerto Rico and St. Croix. Each site in Puerto Rico, planted with six provenances and a total of 384 trees at 3 m spac- ing, occupies 0.23 ha. Each site in St. Croix, with five provenances and 320 trees, occupies 0.19 ha. At Corozal and Guajataca cedars were planted under overstory vegetation. Corozal was a former coffee site with a dense shade of Inga laurina (Sw.) Willd., and Guajataca was a 35-year-old Swietenia mahagoni Jacq. plantation with a high, sparse canopy. The other three sites were not shaded. During the first year, the plantings were weeded as necessary to keep mortality low. On two sites. Bog of Allan and Estate Thomas, all competing vegetation was kept mowed year- round on 4 of the 16 replications in order to test the theory that weed competition limits cedar height growth. The remaining 12 replications on these sites, as well as the 16 on each of the other sites, were not weeded after the first year. Three years after outplanting, results were evaluated according to height, mortality, bran- ches, and attack by the shootborer. Analyses of variance and multiple range tests were used to evaluate the height data. RESULTS AND DISCUSSION Height growth (Table 3) In Puerto Rico the C source outgrew the others by a large margin, averaging 0.88 and 1.10 m in annual height growth. The G and E sources were next, ranging from 0.46 to 0.94 m, and the D, A and F sources followed with 0.24 to 0.64 m per year. At St. Croix sites, where C was not planted, G outgrew the others (0.59 to 0.93 m per year) except at Ham Bluff where it was equaled by D On these three sites, D and E averages ranged from 0.45 to 8.84 m; A and B grew least with 0.31 to 0.69 m annual height increment. There was a very low correlation between tree height during the first 12 months (nursery stage) and height 3 years after outplanting. One cannot expect the tallest trees in the nursery to necessarily be the tallest at the end of 3 years in the field. Cedar plantations often stagnate before the fifth year as the growth rate levels off, but this has not yet happened with any of these provenances. In each provenance, growth rate dropped slightly after the 18-24 month period, but has maintained a rather steady pace (Figure 2). Measuring the crop tree, the tallest in each 4- tree plot, can sometimes give a better estimate of the genetic potential of a provenance than height growth means because the effect of non-genetic defects and accidents is minimized. In this study, crop tree performance of provenances remained comparatively the same as plot means in height growth, but rates are substantially greater. The C source crop trees, for example, grew an average of 1.14 and 1.32 m annually (Guajataca and Coro- zal) compared with the 0.88 and 1.10 m averages for all C source trees, a difference of 20 to 30%. Even so, initial height growth is not satisfac- tory. Considering the problem of weed growth and the threat to slow-growing trees by the form- ruining shootborer, the performance of 1.1 to 1.3 m per year shown by the best trees of the best provenance is disappointing. Lamb (1968), for example, described a Cedrela odorata L. planta- tion in Nigeria which, 29 months after outplant- ing, was 6.1 to 7.6 m in height with excellent form. Furthermore, natural cedar regeneration on abandoned slash-burn agricultural land in the Uraba region of Colombia produced trees 30-40 cm dbh at age 13 years, when they were har- vested. Form was good and wood quality ex- cellent. The effects of shootborer attack were overcome by rapid growth (Arturo Romero, Sil- viculturist, Medellin, Colombia: Personal Com- munication). Also, a cedar plantation in the San Carlos region of Costa Rica had, at age 14 years, trees up to 53 cm dbh, straight, tall stems of high quality, harvestable wood, in spite of early shootborer attack. These examples demonstrate the potential of Cedrela on optimum sites and the goal to which we must aspire in cedar plantation research. Though Corozal, Ham Bluff, and Estate Thomas proved better for cedar height growth than Guajataca and Bog of Allan, performance was basically poor for all sites, and no conclusions can be drawn from this limited study to compare these sites to optimum growing conditions. Weed Competition, Mortality, and Branching Tendency ( Table 4) In St. Croix, the Bog of Allan mowed vs. un- mowed plots showed mean heights of 1.4 and 2.1 m respectively, after three years of growth: the trees with more weeds and less maintenance grew faster in height. The Estate Thomas site showed no difference: 2.3 m for mowed and 2.3 for un- mowed. Another mowed-unmowed trial at Estate Thomas had means of 1.2 and 1.6 m, again indi- cating that weed competition can induce height growth. These results do not support the theory that elimination of weedy competition would pro- duce taller cedars. Adaptability of a provenance can be mea- sured by survival and height growth. In this trial, survival and height growth were directly cor- related: those provenances that grew fastest generally had least mortality. Of the fastest grow- ing provenances, mortality was less than 20%, which is quite acceptable. Overall mortality of trees planted on sites in the subtropical moist forest life zone (Table 2) was 25%, but only 18% on sites in the subtropical dry forest life zone. This tends to support the belief that cedar resists drought well but cannot tolerate excessive moisture. At the end of the nursery phase, the C pro- venance seedlings had 9 times more branches (2.7 per seedling) than any other provenance. The ten- tative conclusion was that, if this branching ten- dency was indicative of later form, the C prove- nance would prove unacceptable. Three years later, we now see that the C source is the most promising in height growth and has excellent form. The tendency for early branching was ap- parently an expression of vigor. The trees on the three St. Croix sites grew in the open and tended to branch more than trees in Puerto Rico, although, the trees in Puerto Rico branched more during the nursery phase (Table 1). The nursery in Puerto Rico was unshaded and the one in St. Croix was heavily shaded. There- fore, the number of branches three years after outplanting is probably related to whether the seedlings are shaded or not. It is also apparently related to height growth: faster growing sources tend to have more branches. Shootborer attack (Table 4) To obtain accurate information on shoot- borer attack, trees must be inspected once every two months and preferably more often. In this study, recent attacks were recorded once a year at the time of height measurement, a highly unsatis- factory evaluation of borer damage. Nevertheless, the 1973 borer data from St. Croix sites indicate a weak, but direct relation between height growth and borer attack: the borer tends to attack trees that grow faster. Borer attack influenced branching: when the borer kills the terminal shoot several side shoots grow and later persist as branches. Borer data from the Puerto Rico sites are confounded with the factor of shade from overstory, thus are not included here. In general, these shaded trees were attacked less than those growing in the open. No provenance proved exceptionally more resistant to attack except when related to height growth. Gara, et al (1973) observed that the borer selects the tallest trees that have fresh, green growth. This observation is supported by the pre- sent study. To the contrary, however, Lamb (1968) found that borer attack failed on the most vigorous trees due to a heavy resin flow which entraps or otherwise discourages the insect. The apparent anomaly suggests that (1) studies of this feature in stands which are not vigorous (i.e. Gara et al. and the present study) may be mis- leading, (2) cedar growing under optimum condi- tions will not be bothered by borer attack, from either a biological or commercial point of view, and (3) the problem must be met silviculturally by identifying and providing optimum growing con- ditions. Interaction It should be noted that the above mentioned items, while listed separately, are definitely inter- related. Mortality, branching tendency and shoot- borer attack are all correlated with height growth. The interaction between branching and borer attack is the more obvious: the borer kills the ter- minal shoot thus causing several side shoots to grow which later persist as branches. Also, there is a strong site/provenance interaction for each of the parameters. CONCLUSIONS 1. Height growth of the most promising pro- venance, on the most promising site, was at best disappointing. C (Turrialba, Costa Rica) is by far the fastest provenance in height growth on the sites in Puerto Rico. G (Belize) was the fastest in height growth on St. Croix. D (Campeche, Me- xico) and E (Cuba) were intermediate on both islands. A (Guanacaste, Costa Rica), B (Jamaica) and F (Argentine) grew poorly. Weeding after the first year provided no stimulus to height growth. 2. Mortality of the faster-growing prove- nances was low and could be lowered even more by further site adaptability research. Cedar re- sists drought well: mortality was lower on dry sites than on moist sites. 3. Branching and borer attack were both more common on the open sites than on the shaded sites. No provenance stood out as borer- resistant. 4. Provenance research, as with borer re- search, is of limited value if none of the trees grow at maximum potential. Until we can define and provide optimum site and growing conditions for cedar, borer and provenance studies will only tell us what happens to a tree growing under an un- known handicap. ACKNOWLEDGEMENT The Commonwealth Forestry Institute pro- vided the seed, Mr. E.L. Bough of the U.S. Virgin Islands, Department of Agriculture, assisted in both the nursery and field phases, Mr. Mariano Antoni of the University of Puerto Rico Agricul- ture Experiment Station guided in the processing of data, and Dr. Robert Gara provided inspiration and ideas. LITERATURE QTED BURLEY. J. and LAMB. A.F.A. 1971. Status of the C.F.I. International Provenance Trial of Cedrela odorata (Including C. mexicana and C. tubiflora) June 1971. Commonw. For. Rev. 50(3): 234-237. EWEL. J.J. and WHITMORE, J.L 1973. The ecological life zones of Puerto Rico and the U.S. Virgin Islands. USDA For. Serv. Res. Pap. ITF-18. 72 p. Institute of Tropical Forestry, Rio Piedras, Puerto Rico. GARA. k.I.. ALLAN. G.G.. WILKINS. R.M.. and WHITMORE. J.L. 1973. Flight and host selection behaviour of the mahogany shootborer. Ht;psipi)la grandeUa Zeller (Lepid., Phycitidae). Angew. Entomol. 72: 259-266. GRIJPMA. P. (ed). 1973. Studies on the shootborer Hi^psipt^la grandella (Zeller). Volume I, Turrialba, Costa Rica. IICA Misc. Publ. No. 101 91 p. LAMB. A.F.A. 1968. Fast growing timber trees of the lowland tropics No. 2: Cedrela odorata. Commonwealth Forest Institute. Oxford, England. 46 p. LINES, R. 1967. Standardization of methods for provenance research and testing. lUFRO XIV Congress, Munich 3: 672-718. MELdHIOR, G. y QUIJADA, M. 1972. Sobre el comportamiento de unas procedencias de Cedrela odorata comparadas con una de C. angustifolia nativa y plantada como "stumps", en condiciones de vivero. Ins. For. Latinoamer. Invest, y Capacit. Bol. 41/42: 57-62. OMOYIOLA. B.O. 1972. Initial observations on a Cedrela provenance trial in Nigeria. Niger. Dep. of For. Res.. Res. Pap. (Forest Series) No. 2. 8p. OTAROLA, A., WHITMORE, J.L., and SALAZAR, R. 1976. Analisis de 12 plantaciones de Toona ciliata var. australis in Turrialba, Costa Rica. Turrialba 26 (1): 80-85. RIVERA. L.H. et al. 1970. Soil survey. Virgin Islands of the United States. U.S. Soil Conserv. Serv., Wash., D.C 78 p. -|- maps. ROBERTS. R.C. 1942. Soil survey of Puerto Rico. USDA Bureau of Plant Ind.. and Univ. of P.R. Agric. Exp. Sta. Series 1936. No. 8. Wash., D.C. 503 -|- maps. SANCHEZ, J.C.. HOLSTEN, E.H.. and WHITMORE, J.L. 1976. Comportamiento de 5 especies de Meliaceae en Florencia Sur, Turrialba. \n Whitmore, J.L. (ed.) 1976. Studies on the shootborer Hi^psipvla graridella (Zeller). Volume III, Turrialba, Costa Rica. CATIE Misc. Publ. No. 1 (IICA Misc. Publ. No 101). 116 WHITMORE, J.L. 1971. Cedrela provenance trial in Puerto Rico and St. Croix: nursery phase assessment. Turrialba 21 (3): 343-349. WHITMORE. J.L (ed.) 1976a. Studies on the shootborer Hi;psipiila grandella (Zeller). Volume II. Turrialba, Costa Rica, IICA Misc. Publ. No. 101. 139 p. WHITMORE, J.L. (ed.) 1976b. Studies on the shootborer Hvpsipvia grar^della (Zeller). Volume III. Turrialba, Costa Rica. CATIE Misc. Publ. No. 1 (IICA Misc. Publ. No. 101). 116 p. WRIGHT, J.W. and FREELAND, JR.. F.D. 1961. Tamano de las parcelas de ensayo en investigaciones de genetica forcstal. Caribbean Forester 22 (3-4): 79.83. Translated by J Marrero from: Plot size in forest genetic research. Pap. Mich. Acad. Sci. Arts and Lett. 44: 177-182. (1958 meeting). T3 an rm Q Ol u q; O ^ 0) JC 3 Q- on c E CT) l/l c 1— <D -4—' •D 1/1 O) n o 1/1 1/1 (E pre CD Q. X Q) 'o CD LO 0) X an ca 'o i/i u ry. Tim ctT E q;u CT) E c 1— 3 na <\: fO <u D AO. !6n T3 Q. < O c x: o '(D CT C c or E u cu u CT TO o (_> ^ 1- r^ > U r O ^ 4) r- ra a £ c CM 3 L o 1-1 0) 03 U) e n) 3 <U C B 0) 4J 4-1 rH M-l to O U t: O c B a) V c 4-1 M 0 c •rH u o 0) c rt 0) iJ S-i U) (U x: 3M •H H 01 U-l J2 <D T3 cn c 01 n) H in lia td 1< C H 0) O > U o o ua 4-1 C/3 hs 4-1 co M-l 6 O coH -a J-l c a nj •H O o CD a 0) •tH O 0) ^ COH >> 4-1 •H o o in o o <r .H CD ^ o o o '— CN in CO CO CO CO 4-1 J-l < w O u Q <; Q W O CQ o C/] 60 00 •H o X •H T3 •H H -C 0) cd in (Vl O O o (>j Csl r-^ O OJ 1-1 CN c CO o CN U 4-1 O >-l P w O < U o W < O 05 pa O CU 4-1 D CO "~ P-i OJ -D XI o 0) ,a o T3 d •K ct) nJ c« Xi O K 4-1 in in CN \o <t CO o C3^ o ^ 9" M u in -3- <r m CN o C3^ H c^ OJ CM Oi CO CO |>1 Csl Csl CU 33 < W o u a < CQ o o W cw 01 4^o o O 0)o CO •H O PJ CO •H 4-1 CO CO O d 4-1 cn O 4-1 O o d o •H cn •H •H 0) o CO 0 X 4-1 60 o 01 CO u 4-1 < 0) C 4J co' CO m QJ cn CO T3 ^ x: a 0) o t-( CO o d 01 0) CO OJ rH 0) o tsl tsl •H M & CO •H H •l-l 14-1 t4 E cn rH tH C D CO H 0) 0) o W H u (J P3 CN Osl d d d d d d d d o 0 o o o o o o X X X X X X o o o O o o o o CN sD CO CO CN CN 00 O 00 so C7S 00 oo rsi CN r\l CN Csj CM CM CN ! 00 r- oo so so so so so so SD 1 o O C7^ .— r—i >i -o rH 4J 3 o 0) 4-1 -) o CO j_i c CO !0 t3N Q r-l. CO CU I-H u •o I-H O (U •H 4J ^ 4-1 a C •H cd <U •H (3 00 Px O •H 4J -a 0) CO O 4J <u cd 6 CO 0) •o Q) 3 3 O O "O * C (U ^— c * CO c C c CO M •r^ (U 60 0) •H U 0) 6 CO 01 •H a O OJ o. to 'H o a. W 4J o O (-1 4J N iH iH O 4J •a CO XI to OJ OJ ?^ CO o c U-l iH c CO Q) to 3 iH •H •H •H ^ c >H O •H CO rH O CO o o U OJ 6 O ?^ CO 4-1 CO to to 13 4-1 a >^ CO O t>D O rH to fi. C OJ -O -H rH ^ O <H 0) (U 0 e to O I-l u c CO .H J-i -H IJ " CO d 0) o o iH (1) to iH to iH O 3 14H N U 3 to 0 rH X 4J CO iH iH O c; ' 01 CO B ^ CO i-H O. < p- >4-l 0) 13 CO O O CO C (1) •a •rA !-i .-1 c o CO x: O 0) G OJ i-l CO O CO M 01 to •H T3 CO -a i-H H Q CO CO CO •s o iH •rl iH P< CO O 14-1 * c 0) CU 4-) 4-1 4-1 4-1 •H M-i c CO to to XI CO •H o H •H •H 3 M hJ N o O O CO e t3 g i-H 1 (U CO •o 3 rH 3 C .H S-i u c CO Id •H CO 14-1 CN o in 4-1 C a\ LTl CX3 <?^ r-i c •H ON CT^ o\ 1— (U CO CO CO r-H 1— ( f— c 0) U o s N CO OJ 0) 01 u T3 14H c 3 ^— s CM O O o o •H CO JJ E in o CTi CN rH G •H CN 1— 4-1 4-1 > tH to o <s 0) 1-4 )-l a. o o HH 4J CO O 4-1 •H (U •o CO (J <u a •H ^1 4J O 3 CD Q e T3 O 01 (U CO 4-1 u pq H 4J to c •< U CO •Ha a o »-i 01 en 4-1 <u (d X3 4J CO e 3 •H I-H o CO w o cd iH x: o u < <4-l II <u •H rH CO 3 4-i d to 4-) 'H M-J (U 1— 4-1 CN •H N CO O O 4J to CO o o •r-) to •H 0) 4-1 S-l CO U 60 4-1 e O rH S-i o 3 o to CO S ,i3 OJ u w CO 3 4-> * H a. CO 0) (U 3o 03 c CO (U B (U COo 60 C 0) >s 0) 0) • 4J O u d OOV •rj Pm (U <u M u 4-1 (U <4-l a. u-i O -H (-1 T3 o -a ^ C w ca c n) 4J a 60 <+-l •H -H 0) C ^ M •H »-i CO O IH 4-) O CD C c c4 (U (1) H 01 B 01 o O 60 !-l -H 4= aC C >o >-l o. >> 43 43 a C n)C 0) >o a XI 60 * (Uo c C >o a. 42 O 60 I-l T-l 4= ci3 43 43 o n) o o n) X X cd X o o XI CO X CM CM vO CM <r CM i-H I-H o 00 00 in in tn 00 m m m o C3^ m 00 CM o m CM ON <r o CM CM (M CM CM CM CO rsi CM CM CO CM CM CM X X o ca 43 o o ta X X ca cfl ca X X ON -J- o 00 00 in m m 00 o vD ON oo O o oo u-i (NI in 00 <f CM in CO in CO CM CM CM 1— CSl CM CM CM CM CM r-H cn to cn ca c ca c ca c ca c a ca a ca a ca a ca o o o iH o U-I X ca U-I ca X U-l ca m ca IJ-l H < a 14H o U-l o <: U-l <: CJ 3 ca 3 I-H ca 3 ca 3 ca 0) ca 14-1 4-1 .H ca uh 4-1 ca 0) r-i 4-1 ca 4-1 « *J N o ca pa 4-1 N o ca N 4-1 n ca o N u « o ca a ca o r-) 0 ca O •I—) o ca •r-> o ca U u 60 ca 4-J u 60 ca u 4-1 S ca 60 u 4J a 60 ca ca cn o O 3 ca o O D o cn ca 3 O o cn ca O 3 w u o w u O W o m o w CJ ca X X o O ca o> On vD o o ON in CM cn -3- m CM CM U O W O < ca X X .-H 00 in cjN ON CO CM <N CJ w o Q <: ca X X X u ca X X a o in CO CO CT> in in I-H 00 CO CJN ON CM CTi CO CO CM 1— 1— CO CM CM CM CM o O <; P3 Q W pq < ca ca X X o 00 vO CJN CO oo ON ^ in I-H CM CSl CM CM CM p o w <; PQ ca X X u -a T3 ca X X o u o ca X X X o ca X CJ -a ca ca X X o X CJ a\ (N 00 CM CO r- o 00 1— vO o -a- 00 CM 00 CJN CO CJN (M o ON CM 00 <r CT\ in in ro o -3- CO CM ON CM CM o 00 in in o 00 CO CM (N CM 1— .— I— .— 1— CM CM CM CM CM CM CM C_) w Q <; CJ w n < Ph CJ Q w <; pa CJ o w m < p CJ w < pa (03O 4-1 O 60 O (Uu n) 4-1 CO M-l M-l3 1-1 PQ § oo cn cuo C ca c 111 >ou (X o Table 4. Ce.dAQZa data three years after outplanting, according to island and provenance. Numbers represent means of two sites in Puerto Rico and three in St. Croix. Height (cm) Mortality (%) Branches (#/live tree) Borer attack (#/live tree) Puerto Rico C 292 8 0.8 G 205 16 1.1 E 198 12 0.8 No D 141 30 0.7 data A 115 50 0.5 F 90 58 0.1 St. Croix G 235 13 3. 0.77 D 207 12 2.9 0.73 E 192 12 1.9 0.74 A 175 16 2.3 0.70 B 159 30 2.2 0.66