A CHRONOLOGY OF HURRICANE INDUCED CHANGES IN ’S LOWER MONTANE RAIN FOREST

Peter L. Weaver

SUMMARY

Ridges in the lower montane rain forests of Puerto Rico and generation of yagrumo hembra (Cecropia schreberiana Mig.) the Lesser Antilles from St. Kitts to Grenada are dominated by along with numerous small- to medium-sized species in forest tabonuco (Dacryodes excelsa Vahl), a long-lived tree adapted to openings. Stem density, species numbers, and the rate of biomass recurrent hurricanes. The oldest tabonuco trees in Puerto Rico accumulation are at a maximum 15yrs after the hurricane; about appear to survive 500 to 600yrs in forests that periodically (per- 50yrs later, most of the secondary species associated with past haps every 50 to 60yrs) lose nearly one-fifth of their biomass. forest disturbance have disappeared and the rate of biomass Post hurricane-recovery, characterized by greater rates of stem accumulation becomes asymptotic. ingrowth and mortality, showed an immediate and abundant re-

RESUMEN

Las cimas de las colinas en las selvas pluviales intermedias (Cecropia schreberiana Mig.) inmediata y abundante, junto con de Puerto Rico y las Antillas Menores, entre St. Kitts y Granada, varias especies de tamaño pequeño a mediano, en aperturas den- están dominadas por tabonuco (Dacryodes excelsa Vahl), un árbol tro del bosque. La densidad de los tallos, número de especies, y de larga vida adaptado a huracanes recurrentes. Los tabonucos tasa de acumulación de biomasa alcanzan un máximo cerca de más antiguos en Puerto Rico parecen sobrevivir unos 500 a 600 unos 15 años después del huracán; alrededor de 50 años después, años en bosques que periódicamente (tal vez cada 50 a 60 años) la mayoría de las especies secundarias asociadas con anteriores pierden casi la quinta parte de su biomasa. La recuperación pos- perturbaciones han desaparecid y la tasa de acumulación de bio- huracán, caracterizada por tasas más altas de reclutamiento y masa se hace asintótica. mortalidad, demostró una regeneración de yagrumo hembra

Introduction (Birdsey and Weaver, 1982; (Walker et al., 1991; 1996). stem density, species diversity, Franco et al., 1997). The Many of these observations, and species dominance on Puerto Rico’s lower mon- island’s best remaining stands however, were made on plots TR-1, the only long-term plot tane rain forest covered nearly of primary lower montane set up after the storm. summarized to date for all of one-quarter of the island at the forest are in the Luquillo The Forest Service estab- the major 20th century hurri- time of discovery by Europe- Mountains below 600m. The lished many of its permanent canes. ans. The largest trees, tabo- most impressive of these are plots in the Luquillo Moun- nuco (Dacryodes excelsa situated on ridges where tains during the mid-1940s Setting Vahl), ausubo (Manilkara tabonuco is the dominant tree (Crow, 1980; Weaver, 1998). bidentata A. DC. [Chev.]), (Wadsworth, 1953; Crow and Their long-term records show The lower montane rain granadillo (Buchenavia tetra- Grigal, 1979; Crow, 1980; changes in structure and com- forest (sensu Beard, 1949) or phylla [Aublet] R. Howard), Basnet, 1992). position after impacts by Hur- tabonuco forest of the and motillo (Sloanea berte- Hurricanes are a regular ricanes San Felipe of 1928, Luquillo Experimental Forest roana Choisey), grew to >2m phenomenon in the Caribbean San Nicolas of 1931, and San (LEF) in northeastern Puerto in dbh (diameter at breast Basin between June and Novem- Cipriano of 1932, mainly the Rico extends from the forest height, or 1.4m above the ber; most storms affecting Puerto latter, which traversed the border near 120m in elevation ground) and >30m in height Rico occur in July, August, mountains. The effects of to about 600m (Wadsworth, (Wadsworth, 1950). After and September (Salivia, 1972; Hurricanes Hugo of 1989, 1951). The 0.4ha permanent settlement, land clearing and Weaver, 1998). The impact of and Georges of 1998, both of plot (TR-1, or tabonuco ridge) agriculture removed most of hurricanes on Puerto Rico’s which impacted the moun- monitored in this study is Puerto Rico’s timber replacing forests has received much at- tains, are also evident. The situated on a northeast-facing it with the farmlands and sec- tention since the passage of purpose of this work is to ridge at 350m in elevation ondary forest evident today Hugo in September 1989 show the major changes in (Weaver, 1994), about 2km

KEY WORDS / Forest Recovery / Hurricanes / Lower Montane Rain / Received: 01/15/2002. Modified: 03/22/2002. Accepted: 04/04/2002

Peter L. Weaver. M.S., SUNY, search Forester, International dress: P.O. Box 25000. Rio Syracuse, New York. Ph.D., Institute of Tropical Forestry, Piedras, PR 00928-5000. Michigan State University. Re- Río Piedras, Puerto Rico. Ad- e-mail: [email protected]

252 0378-1844/02/05/252-07 $ 3.00/0 MAY 2002, VOL. 27 Nº 5 RESUMO

Os cumes das selvas pluviais intermediárias de Porto Rico e schreberiana Mig.) imediata e abundante, junto com várias espéci- as Antilhas Menores, entre St. Kitts e Granada, estão dominadas es de tamanho pequeno a médio, em aberturas dentro do bosque. por tabonuco (Dacryodes excelsa Vahl), uma árvore de longa vida A densidade dos talhos, número de espécies, e taxa de acumula- adaptada a furacões recorrentes. Os tabonucos mais antigos em ção de biomassa alcançam um máximo de 15 anos aproximada- Porto Rico parecem sobreviver uns 500 a 600 anos em bosques mente depois do furacão; mais ou menos uns 50 anos depois, a que periodicamente (tal vez cada 50 a 60 anos) perdem quase a maioria das espécies secundárias associadas com anteriores per- quinta parte de sua biomassa. A recuperação pós-furacão, carac- turbações desapareceram e a taxa de acumulação de biomassa se terizada por taxas mais altas de recrutamento e mortalidade, de- faz assintótica. monstrou uma regeneração de yagrumo fêmea (Cecropia

distant from the LTER (long- TABLE I term ecological research) site SPECIES CHANGES ON THE TR-1 STAND FROM 1946 TO 2000 at Bisley (Scatena, 1989). This windward portion of the Parameter (units) Year LEF has also been classified 1946 1951 1956 1966 1977 1988 1994 2000 as subtropical rain forest

(Ewel and Whitmore, 1973) 1 in the Holdridge (1967) life Species (no.) 2 zone system. Rainfall aver- Original 46 44 42 40 36 36 32 28 ages about 3800mm/yr, mean Residual3 46 44 42 40 36 38 35 33 annual temperature is 22oC, New4 ––––4554 and the soils are acid clays Total 46 44 42 40 40 43 40 37 (Weaver, 1998). The dominant tree species are evergreen and Stems per species (no.) normally reach 30m in height 1 111112118107 6 and 0.6m in diameter 2 341234711 (Wadsworth, 1951). 3-4 8665911106 5-10 89897674 Methods 11-20 54535643 TR-1, established in 1946, 21-50 87786425 was remeasured in 1951, ≥51 33322232 1956, 1966, 1977, 1988, 1 Fifty-six tree species were recorded at least once during the 54 yrs of measurement on TR-1. TR-1 lost 13 residual spe- 1994, and 2000. In 1946, all cies by 2000 and gained 10 new species of which only 4 survived in 2000. trees rooted within TR-1 were 2 Original species are survivors of the 1946 population of stems. tallied sequentially but not 3 Residual species are those first recorded in 1946 and also found in the year indicated. Residuals may be lost temporarily mapped (Weaver, 1983, through mortality and then reappear as ingrowth. 1998). All trees ≥4.1cm in 4 New species are those that were not present in 1946 that were recorded in the year indicated. New species may be lost dbh were identified to spe- temporarily through mortality and then reappear as ingrowth. New species were not tallied from 1951 through 1966. cies, measured to the nearest 0.1cm in dbh, and perma- nently tagged with a nail 15cm below dbh to avert tion derived for the tabonuco determined to elucidate the Results swelling around the nail. forest, dominance of tabonuco on Most of the nail was left ex- ridge topography. Species posed to allow future growth. Y = 3.45 - 2.47 D + 0.533 The age of the largest Ingrowth, or recruitment into D2, r2 = 0.89 tabonucos was estimated by di- Fifty-six species were re- the smallest dbh class, was viding their dbh by the mean corded at least once during not tallied until 1977. In where D = dbh in cm and Y = annual increment for all surviv- the 54yrs of measurement on 1988, ingrowth was measured biomass in kg (Weaver and ing stems. Since measurements TR-1 (Table I; Appendix and tagged, and for the first Gillespie, 1992). Total abo- only involved stems ≥4.1cm in Table I). The total number of time, total tree heights were veground biomass for palma dbh, the age estimates assume species declined from 46 in recorded to the nearest 0.1m de sierra (Prestoea montana that seedlings and saplings 1946 to 40 in 1977, increas- with an optical range finder. [R. Grah.] Nicholls) was esti- grew at the mean rate until ing slightly to 43 in 1988. Species names follow local mated by they reached 4.1cm. Moreover, The decline continued after texts (Liogier, 1985-97; see the survival rates of four major the hurricanes to 40 in 1994, also, Little and Wadsworth, Y = 6.4 x -10.0, r2 = 0.96 canopy species, five common and 37 in 2000 (Figure 1). 1964; Little et al., 1974). understory species, and two The numbers of stems per Above-ground dry weight where x = height in m and Y major gap species (secondary species also declined in most woody biomass (hereafter, = total aboveground biomass species that regenerate in open- instances. biomass) was determined for in kg (Frangi and Lugo, ings) recorded in 1946, were Thirteen species recorded in dicotyledons using an equa- 1985). Biomass estimates were considered in detail. 1946 were absent from TR-1

MAY 2002, VOL. 27 Nº 5 253 Figure 1. Numbers of species and number of stems per species for four different samples taken on TR-1 from 1946 to 2000. in 2000: guajón (Beilschmiedia W. Hamilton), hueso blanco thirds of the total in 1946. garciniifolia Pierre), 50%; pendula [Sw.] Hemsl.), (Chionanthus domingensis Their survival rates varied dur- palma de sierra (Prestoea maricao (Byrsonima spicata Lam.), camasey (Miconia pra- ing the 54yr period of mea- montana [R. Grah.] Nichols), [Cav.] HBK), cafeíllo (Casea- sina [Sw.] DC.), hoja menuda surement. The major canopy 24%; jusillo (Henriettea squa- ria sylvestris Sw.), sabinón (Myrcia fallax [A. Rich] dominants had an average sur- mulosa [Cogn.] Judd), 11%; (Croton poecilanthus Urban), DC.), nuez moscada (Ocotea vival of 57%, as follows: and muñeco (Cordia borin- helecho gigante (Cyathea moschata (Meissn.) Mez), but ausubo, 80%; tabonuco, 69%; quensis Urban), 8%. Survival arborea [L.] J.E. Smith), ma- only the last four species and motillo, 27% (Appendix of camasey (Tetrazygia urbanii jagua brava Daphnopsis were present in 2000 (Appen- Table I). Granadillo, also a Cogniaux), another common philippiana (Krug & Urban), dix Table I). Of the new spe- canopy dominant, had 80% understory species, was only haya blanca (Guatteria cari- cies, all but nuez moscada are survival, but with only 5 stems 2%. Lastly, survival of the baea Urban), azafrán (Hedy- small to medium in size. is not shown. Survival of the major secondary, or gap spe- osmum arborescens Sw.), The species shown in Figure major understory species aver- cies, averaged 10%, as fol- maricao verde (Laplacea por- 2 accounted for 472 stems aged 24%, as follows: cai- lows: 17% of the yagrumo toricensis [Krug & Urban] (i.e., 1171/ha), or about two- mitillo verde ( macho (Schleflerra morototoni Dyer), cieneguillo (Myr- [Aubl.] Maguire), and cia deflexa [Poir] DC.), none of the yagrumo hoja menuda (Myrcia hembra. splendens [Sw.] DC.), Whereas mortality ex- tabaiba (Sapium lauro- ceeded ingrowth for cerasus Desf.), and most species during the tabacón (Solanum rugo- measurement period, sum Dunal) (Appendix yagrumo hembra gained Table I). Twelve of 126 stems in 1994 of these species, totaling which 27 survived 22 stems, had ≤4 stems through 2000 (Figure 2; each in 1946; only ma- Appendix Table I). ricao verde with 20 Palma de sierra showed stems was well repre- a considerable flux in sented in 1946. stems; 46 trees were re- Ten new species en- corded as ingrowth and tered the plot after the 70 died, notably from 1946 measurement: ma- 1988 to 2000. Ausubo meyuelo (Ardisia obo- and motillo also regener- vata Desv. ex Hamilt), ated well, particularly af- ortegón (Coccoloba ter Hurricane Hugo (Fig- swartzii Meisner), ca- ure 2). feíllo (Ixora ferrea [Jacq.] Benth.), yaya Figure 2. Changes in the number of stems by tree species on TR-1. White: surviv- Structure (Oxandra laurifolia ing residual stems from 1946. Shaded: ingrowth of new stems. Species: DAEX: [Sw.] A. Rich), cachim- Dacryodes excelsa; MABI: Manilka bidentata; SLBE: Sloanea berteroana; MIGA: Stand density, at 1761 bo (Palicourea crocea Micropholis garciniifolia; PRMO: Prestoea montana; COBO: Cordia borinquensis; stems/ha, was greatest [Sw.] Roem. & Schul- HESQ: Henriettea squamulosa; TEUR Tetrazygia urbanii; SCMO: Schleflerra in 1946, about 14yrs af- tes), palo de cachimbo mororotoni; CESC: Cecropia schreberiana. Years: 1946, 1951, 1956, 1966, 1977, ter the passage of San (Palicourea croceoides 1988, 1994, and 2000. Cipriano (Table II).

254 MAY 2002, VOL. 27 Nº 5 TABLE II Dbh CLASS DISTRIBUTIONS FOR TR-1 BETWEEN 1946 AND 20001

Parameter Year (units) 1946 1951 1956 1966 1977 1988 1994 2000

1946 stems/ha 1761 1578 1420 1267 1045 877 613 450 Percentage of 1946 stems in each class by year Dbh (cm) 4.1-9.9 58.4 50.3 45.4 38.5 32.0 27.5 20.2 16.4 10.0-19.9 28.0 32.6 35.3 36.7 38.4 38.3 40.1 38.1 20.0-29.9 6.3 8.8 9.9 2.5 14.8 15.7 16.3 17.5 530.0-39.9 2.2 2.5 3.0 4.9 5.4 7.2 7.9 7.9 ≥40.0 5.1 5.8 6.4 7.4 9.4 11.3 15.5 20.1 total stems/ha 1761 – – – 1190 1136 1193 916 Figure 3. Changes in stem numbers on TR- Percentage of total stems in each class by year 1. Complete stand measurements are pre- Dbh (cm) sented for 1946, 1977, 1988, 1994, and 4.1-9.9 58.4 —— —— —— 37.9 38.8 52.5 45.3 2000. Ingrowth was not available for 1951, 1956, and 1966. Ingrowth in 1977 represents 10.0-19.9 28.0 —— —— —— 36.1 34.3 26.6 31.8 cumulative data from the first measurement 20.0-29.9 6.3 —— —— —— 12.9 12.4 8.7 8.9 in 1946. White, number of residual stems 30.0-39.9 2.2 —— —— —— 4.8 5.6 4.1 3.8 from 1946; black, mortality since last mea- ≥40.0 5.1 —— —— —— 8.3 8.9 8.1 10.2 surement; light gray, most recent ingrowth; dark gray, surviving ingrowth previous to 1Per hectare values rounded to nearest unit. last measurement.

Thereafter, the residual stems t·ha-1yr-1 between 1951 and Age of tabonuco ing Hurricane Hugo. It may (those surviving from 1946) 1977, becoming asymptotic have been about 600yrs old. steadily declined to 450/ha in between 1977 and 1988. Dur- The mean annual dbh in- 2000, a loss of 75% of the ing the last 12yrs, after Hugo crement for the 59 tabonuco Discussion trees originally tallied. Total and Georges, biomass change trees that survived the 54yr stems (including ingrowth) de- was negative, averaging -4.3 period of measurement was Species and structure clined continuously to 1136/ha t·ha-1yr-1. Tabonuco comprised 0.19 ±0.02cm/yr. Although in 1988 (Figure 3). After Hur- from 51 to 55% of the biom- growth by class varied, no The most notable phenom- ricane Hugo, total stems in- ass from 1946 through 1994, trends were apparent. Mean ena of TR-1’s history are the creased slightly in 1994, and increasing to 59% in 2000. dbh growth rates by class, in persistent dominance of declined again in 2000 after The distribution of stems cm (and sample size), were tabonuco, both in numbers and Georges. by dbh class in 1946, 14 biomass, in the pre- and post- Fluxes in stem numbers, years after San Cipriano, 4.1-9.9: 0.12 ± 0.02 (20) hurricane stand, and the sud- both ingrowth and mortality, shows nearly 60% in the 10.2-19.9: 0.26 ± 0.03 (11) den and abundant regeneration were greatest immediately after smallest class and only 5% 20.0-29.9: 0.16 ± 0.07 (6) of yagrumo hembra a few the hurricanes. Ingrowth totaled in the largest class (Table II). 30.0-39.9: 0.14 ± 0.03 (6) years after Hugo (Figure 2). 111 stems (6ha-1yr-1) and 276 The proportion of stems 40.0-49.9: 0.22 ± 0.05 (7) Past inventories of 4ha of stems (57ha-1yr-1) before and present in 1946 subsequently 50.0-59.9: 0.30 ± 0.04 (6) lower montane rain forest on after Hugo, respectively (Ap- declined in the smallest ≥60.0: 0.17 ± 0.02 (5) several sites showed that pendix Table I). Comparable class, and increased in the tabonuco comprised 10% of values for mortality were 363 remaining classes through Of the 27 tabonuco trees the stems ≥10cm in dbh stems (21 ha-1yr-1) and 365 2000. For total stems which that died during measurement, (Wadsworth 1951). Ausubo, stems (75ha-1yr-1). include ingrowth, there was only 2 grew at 0.19cm/yr or granadillo, and motillo in the Stand biomass, about 235t/ more variability. In general, faster. The remaining trees same inventories accounted for ha in 1946, gradually in- there was a proportional in- averaged only 0.06cm/yr another 6% of the stems. On creased to about 290t/ha in crease in the largest dbh while they were alive. ridge topography in mature 1988, and then declined after class after San Cipriano. In Assuming the mean growth lower montane rain forest, both hurricanes to 240t/ha, vir- contrast, the smallest class rate, the largest surviving tabonuco attains large size and tually the same value as in initially declined after 1946, tabonuco on TR-1, with a dbh may account for 50 to 60% of 1946. The biomass accumula- then fluctuated after Hugo, of 94.5cm, should be about the total basal area (Wads- tion rate after San Cipriano reflecting the ingrowth and 500yrs old. The largest previ- worth, 1953; Weaver, 1983). decreased from 3.6 t·ha-1yr-1 competition for survival of ously measured tree, 111cm The decline in the total between 1946 and 1951, to 1.3 yagrumo hembra. in dbh, was blown over dur- number of species on TR-1

MAY 2002, VOL. 27 Nº 5 255 from a high 14 years after fined to recuperation of 72yr. Ratios of biomass to or about one every 50 to hurricane San Cipriano in branches and foliage. production (Odum, 1971), or 60yrs (Scatena and Larsen, 1946 to a low in 2000 imme- Next is the building stage possibly the annual proportion 1991). Given that each storm diately after Hurricane lasting about 35yrs. It in- of canopy openings to total causes a loss of foliage, Georges reflects two main volves competition among area under study (Lawton and breakage of numerous bran- trends (Figure 1). The first, both the survivors of the Putz, 1988) could also be ches, and occasionally the from 1946 to 1988, is the growth stage and new in- used to approximate turn-over. loss of tree tops, the mean gradual loss of several species growth, as all stems increase However, as is evident from growth rate for the survivors that regenerated in openings in size. Mortality, at first this study, an estimate of seems to be a plausible value caused by San Cipriano as rapid, decreases over time as turn-over may vary with the to estimate age. well as the loss of uncommon the total number of stems de- starting point used to charac- Tabonuco growth rates species (Appendix Table I). cline. Biomass accumulation, terize stand conditions as well spanning 18yrs (1957-1975) These losses were tempered initially rapid, also decreases as climatic events, regardless are available for LEF sites by the occasional ingrowth of over time. Species richness of the criterion used to deter- with different disturbance re- new species as the stand ma- peaks early in the building mine it. Although turn-over gimes (Crow and Weaver, tured (Crow, 1980; Weaver, stage 15 to 20yrs after the on portions of plots similar to 1977). Tabonuco’s average 1989). The second, from 1989 storm (Table I), when stem TR-1 may occur several times dbh increment at Sabana 8, to 2000, is the sudden loss of numbers are greatest, and in one-half of a millennium, Río Grande 3, and Sabana 4, some uncommon species due both secondary and primary other parts will remain cov- was 0.74, 0.46, and 0.25cm/ to the impact of Hurricane species are interspersed. ered for much or all of that yr, respectively. The first two Hugo. The immediate in- Gradually, short-lived second- period by long-lived tabonuco sites had been disturbed. growth and rapid decline of ary species succumb, and trees and other canopy domi- Sabana 8, in 1947, was cov- yagrumo hembra is associated long-lived, shade-tolerant spe- nants. ered with heavy brush, sap- with its life history strategy cies dominate the stand. lings, and pole-size trees. Río (Brokaw, 1998). Yagrumo In the last, or mature stage, Age of tabonuco Grande 3, in 1934, had pas- hembra is a typical pioneer about 50yrs after the storm, ture at lower elevations and (gap) species with frequent the stand reaches its greatest The largest living tabonuco cutover forest at mid-eleva- and abundant flowering and standing biomass. Short-lived tree on TR-1 appears to have tions. The dbh growth on seed production, wide seed secondary species have regenerated about the time both of these secondary sites dispersal, and a seed bank in largely disappeared, their Columbus discovered Puerto was more rapid than at the soil whose germination is numbers being maintained at Rico, and the largest tree re- Sabana 4, which, like TR-1, tied to canopy opening. After low levels through gap dy- corded on the plot, about remained undistubed by hu- canopy disturbance, seeds in namics. Stem ingrowth and 100yrs earlier. Most arith- man activities. the seed bank germinate. Ini- mortality remain low and metic estimates of age for The arithmetic approach, tial height growth is rapid, consistent compared to earlier tropical trees without annual used to estimate the age of but competition for growing stages. rings rely on the assumption other major species in the space results in high mortal- Some differences will likely that the particular tree under LEF, indicated that the oldest ity. be seen during the post-Hugo consideration is growing at trees of two other species Similar species changes recovery pattern, however. the average rate for all trees most likely survived at least 5 were evident after San TR-1’s peak in stem density measured in the sample, a centuries. Palo colorado Cipriano in the lower mon- observed 15 to 20yrs after condition that is questionable (Cyrilla racemiflora L.) at tane rain forest at El Verde, San Cipriano, a phenomenon since most survivors should 100cm in dbh, and laurel also in the LEF (Crow, 1980): also noted on the El Verde grow at a more rapid pace. sabino (Magnolia splendens tabonuco and ausubo declined plot (Crow, 1980), will most When procedures for estimat- Urban) at 65cm in dbh, were by about 20% from 1946 to likely be delayed due to the ing tree ages were first devised estimated to be about 650 and 1976 whereas yagrumo hem- impact of Georges during the (Foggie, 1945; Osmaston, 500yrs old, respectively bra virtually disappeared from post-Hugo growth stage. The 1956), many tree growth re- (Weaver, 1986, 1987). the stand. Moreover, between combined destructive effect of cords were of short duration 1943 and 1976, 12 new spe- both hurricanes, without fur- and field samples included Implications cies were recruited into the ther disturbance on TR-1, some trees that would suc- stand while 19 were lost. could result in stem densities cumb within a few years. TR-1 is but one small, very Long-term recovery, as wit- around 2012 similar to those The mean growth estimate exposed ridge in the middle nessed on TR-1, has different of 1946. Moreover, the com- for tabonuco trees on TR-1, of a montane rain forest. Al- stages. The first is the growth paratively high biomass accu- however, is based on long- though hurricane damage var- stage lasting about 10 or mulation rate similar to that term survivors. For the largest ied consideraby by site (as- 15yrs. It is characterized by witnessed 15 to 20yrs after trees, that covers about 10% pect, topography, slope, and rapid ingrowth of many San Cipriano will most likely of their life cycle. Moreover, exposure) within the LEF stems, particularly secondary occur at this time. the oldest trees have suffered (Boose et al., 1994), the TR-1 species (Crow, 1980; Weaver, The loss of three-quarters at least five hurricanes since stand appears representative 1998). Biomass accumulation, of the original stems during the mid-1700s (Weaver, 1989; of ridge topography in the however, is slow because the 54yrs, especially understory this paper). Assuming that the forest’s northeastern sector. new stems are small, most and gap species, provides record before 1750 contained The LEF’s largest tabonuco being light in weight (i.e., some insights to tree turn- a similar number of storms, trees grow on ridges suggest- low specific gravity). Much of over on the TR-1 stand. A the largest trees should have ing that root anchorage the initial growth on residual simple extrapolation based on experienced about 10 hurri- among rocks or possibly root stems, in turn, is largely con- tree survival would be about canes during the last 500yrs, grafts that occur among clus-

256 MAY 2002, VOL. 27 Nº 5 tered tree groups help them APPENDIX TABLE I survive storms (Wadsworth, NUMBER OF STEMS ON TR-1 FROM 1946 TO 2000 1953; Basnet, 1992; Basnet et al., 1993). In addition, ridges Number of stems/acre1 are better drained and more Stand Ingrowth Mortality Res.2 Stand stable, with fewer landslides 1946 46-88 88-00 46-88 88-00 1946 2000 and treefall gaps, than other topographic positions (Scatena Species and Lugo, 1995). The rela- Alchornea latifolia Sw. 11005422 tively high 54yr survival of Alchorneopsis portoricensis Urban 3010133 other major dominants on Andira inermis (W. Wr.) DC. 3001022 TR-1, ausubo and granadillo, Antirhea obtusifolia Urban 1100012 also lends some support to Beilschmiedia pendula (Sw.) Hemsl. 1000100 this hypothesis. Buchenavia tetraphylla (Aublet) R. Howard 5001044 Byrsonima spicata (Cav.) HBK 2001100 Tabonuco is a dominant Byrsonima wadsworthii Little 3000211 tree species in the lower mon- Casearia sylvestris Sw. 2002000 tane rain forests of Puerto Cassipourea guianensis Aubl. 1230402 Rico and the Lesser Antilles Cecropia schreberiana Mig. 19 0 131 18 105 0 27 from St. Kitts to Grenada Cordia borinquensis Urban 49 3 4 40 6 4 10 (Beard, 1949; Little and Croton poecilanthus Urban 1000100 Wadsworth, 1964; Lugo and Cyathea arborea (L.) J.E. Smith 4004000 Cyrilla racemiflora L. 6002044 Wadsworth, 1990). Hurricane Dacryodes excelsa Vahl 86 2 1 14 13 59 62 damage in Dominica after Daphnopsis philippiana Krug & Urban 2202200 David of 1979 was consider- Eugenia borinquensis Britt. 3000122 able; tabonuco associations, Garcinia portoricensis (Urban) Alain 10 12 8 3 14 3 13 however, suffered the least Guarea glabra Vahl. 1100012 damage whereas palm brakes Guatteria caribaea Urban 1000100 were most affected (Lugo et Hedyosmum arborescens Sw. 3003000 Henriettea squamulosa (Cogn.) Judd. 37 6 3 24 12 4 10 al., 1982). The combination Hirtella rugosa Pers. 25 9 2 6 17 8 13 of longevity and the capacity Homalium racemosum Jacq. 5000332 to survive numerous storms Inga laurina (Sw.) Willd. 11 005422 during its life cycle are what Laplacea portoricensis (Krug & Urban) Dyer 20 0 0 17 3 0 0 account for the abundance of Magnolia splendens Urban 10 006222 tabonuco on ridges in the Manilkara bidentata (A. DC.) Chev 10 8 17 0 10 8 25 Caribbean hurricane belt. In- Margaritaria nobilis L.f. 1000011 Matayba domingensis (DC.) Radlk. 8001344 terestingly, tabonuco is also Meliosma herbertii Rolfe 13 526658 very abundant in the montane Micropholis garciniifolia Pierre 48 1 2 12 11 24 28 rain forests of Grand Etang, Myrcia deflexa (Poir) DC. 1001000 Grenada, an area described as Myrcia splendens (Sw.) DC. 3003000 physiognomically mature but Ocotea leucoxylon (Sw.) Mez 1011001 floristically impoverished Ocotea spathulata Mez 42 2 1 27 16 2 2 (Beard, 1949). Grenada lies Prestoea montana (R. Grah.) Nichols. 78 22 24 25 45 19 54 Psychotria berteriana DC. 24 1 17 24 4 0 14 south of the Caribbean hurri- Sapium laurocerasus Desf. 1001000 cane belt and has suffered Schleflerra morototoni (Aubl.) Maguire 30 0 6 15 11 5 10 only one recorded hurricane, Sloanea berteroana Choisy 34 11 20 23 15 6 27 Janet of 1955. Solanum rugosum Dunal 1001000 Tabebuia heterophylla (DC.) Britt. 9237304 ACKNOWLEDGMENTS Tetragastris balsamifera (Sw.) Kuntze 3000122 Tetrazygia urbanii Cogniaux 81 2 18 56 24 2 21 Juan Ramírez, Alberto Subtotals 713 92 264 357 346 183 366 Rodríguez, and Iván Vicens assisted in data collection. New species from 1977-2000 Two anonymous reviewers of- Ardisia obovata Desv. ex Hamilt. 0100100 fered very helpful comments. Chionanthus domingensis Lam. 0030003 This work was done in coop- Coccoloba swartzii Meisner 0010100 eration with the University of Ixora ferrea (Jacq.) Benth. 0502300 Puerto Rico at Río Piedras. Miconia prasina (Sw.) DC. 0010001 Myrcia fallax (A. Rich) DC. 0110101 REFERENCES Ocotea moschata (Meissn.) Mez 0403001 Oxandra laurifolia (Sw.) A. Rich.0101000 Palicourea crocea (Sw.) Roem. & Schultes 0020200 Basnet K (1992) Effect of topogra- Palicourea croceoides W. Hamilton 0740110 0 phy on the pattern of trees in tabonuco (Dacryodes excelsa) Totals 713 111 276 363 365 183 372 dominated rain forest of Puerto Rico. Biotropica 24: 31-42. 1Multiply by 2.47 to convert to hectare basis. Basnet K, Scatena FN, Likens GE, 2 Lugo AE (1993) Ecological Res. = residuals, or survivors from 1946.

MAY 2002, VOL. 27 Nº 5 257 consequences of root grafting est. The Empire Forestry Jour- pact of Hurricane David on the Biotropica special issue. in tabonuco (Dacryodes excel- nal 24: 176-191. forests of Dominica. Canadian Biotropica 23. 521 pp. sa) trees in the Luquillo Expe- Franco PA, Weaver PL, Eggen- J. Forest Res. 13: 201-211. Walker LR, Silver WL, Willig MR, rimental Forest, Puerto Rico. McIntosh S (1997) Forest re- Odum EP (1971) Fundamentals of Zimmerman JK (Eds) (1996) Biotropica 25: 28-35. sources of Puerto Rico, 1990. Ecology. W.B. Saunders. Phila- Biotropica special issue. Bio- Beard JS (1949) Natural vegetation Resource Bulletin SRS-22. delphia, PA. 574 pp. tropica 28. 614 pp. of the Windward and Leeward USDA Forest Service, Osmaston MA (1956) Determina- Weaver PL (1983) Tree growth and Islands. Oxford Forestry Mem- Asheville, NC. 45 pp. tion of age girth and similar stand changes in the subtro- oirs 21: 1-192. Frangi JL, Lugo AE (1985) Eco- relationships in tropical fores- pical life zones of the Luquillo Birdsey RA, Weaver PL (1982) The system dynamics of a subtropi- try. Empire Forestry Review Mountains of Puerto Rico. Re- forest resources of Puerto cal floodplain forest. Ecologi- 35: 193-197. search Paper SO-190. U.S. De- partment of Agriculture, Forest Rico. USDA Forest Service cal Monographs 55: 351-369. Salivia LA (1972) Historia de los Resource Bulletin SO-85. New Holdridge LR (1967) Life zone Service, Southern Forest Ex- temporales de Puerto Rico y periment Station, New Orleans, Orleans, LA. 59 pp. ecology. Tropical Science Cen- las Antillas, 1492 a 1970. Edi- ter. San José, Costa Rica. 206 LA. 24 pp. Boose ER, Forester DR, Fluet M torial Edil. Universidad de Weaver PL (1986) Growth and age pp. Puerto Rico, Puerto Rico. 385 (1994) Hurricane impacts to of Cyrilla racemiflora L. in tropical and temperate forest Lawton RO, Putz FE (1988) Natu- pp. montane forests of Puerto landscapes. Ecological Mono- ral disturbance and gap-phase Scatena FN (1989) An introduction Rico. Interciencia 11: 221-228. graphs 64: 369-400. regeneration in a wind-exposed to the physiography and his- Weaver PL (1987) Ecological obser- Brokaw NVL (1998) Cecropia tropical cloud forest. Ecology tory of the Bisley Experimental 69: 764-777. vations on Magnolia splendens schreberiana in the Luquillo Watersheds in the Luquillo Urban in the Luquillo Moun- Mountains of Puerto Rico. The Liogier HA (1985-97) Descriptive Mountains of Puerto Rico. tains of Puerto Rico. Caribbean Botanical Review 64: 91-120. flora of Puerto Rico and adja- GTR-SO-72. U.S. Department J. Science 23: 340-351. cent islands-Spermatophyta. of Agriculture, Forest Service, Weaver PL (1989) Forest changes Crow TR (1980) A rainforest chro- Vols. I-V. Editorial de la Southern Forest Experiment after hurricanes in Puerto nicle: a 30-year record of Universidad de Puerto Rico. Station, New Orleans, La. 22 Rico’s Luquillo Mountains. In- change in structure and com- San Juan, Puerto Rico. 2337 pp. terciencia 14: 181-192. position at El Verde, Puerto pp. Rico. Biotropica 12: 42-55. Scatena FN, Larsen MC (1991) Weaver PL (1994) The Baño de Oro Little EL Jr, Wadsworth FH (1964) Physical aspects of Hurricane Crow TR, Grigal DF (1979) A nu- Natural Area: Luquillo Moun- Common trees of Puerto Rico Hugo in Puerto Rico. Biotro- merical analysis of arborescent tains, Puerto Rico. GTR-S0- and the Virgin islands. Agricul- pica 23: 317-323. 111. U.S. Department of Agri- communities in the rain forest tural Handbook No. 249. of the Luquillo Mountains, Scatena FN, Lugo AE (1995) Geo- culture, Forest Service, South- USDA Forest Service, Wash- morphology, disturbance, and ern Forest Experiment Station. Puerto Rico. Vegetatio 40: 135- ington, D.C. 548 pp. 146. the soil and vegetation of two New Orleans, LA. 56 pp. Little EL Jr, Woodbury RO, subtropical wet steepland wa- Weaver PL (1998) Hurricane ef- Crow TR, Weaver PL (1977) Tree Wadsworth FH (1974) Trees of tersheds of Puerto Rico. Geo- fects and long-term recovery in growth in a moist tropical for- Puerto Rico and the Virgin is- morphology 13: 199-213. subtropical rain forest. In est of Puerto Rico. Forest Ser- nd lands, 2 . Vol. Agricultural Dallmeier F, Comisky JA (Eds.) vice Research Paper ITF-22. Wadsworth FH (1950) Notes on the Handbook 449. USDA Forest Forest biodiversity in North, Río Piedras, Puerto Rico: U.S. Service, Washington, D.C. climax forests of Puerto Rico Department of Agriculture, and their destruction and con- Central and 1024 pp. and the Caribbean: Research Forest Service, Institute of servation prior to 1900. Carib- Lugo AE, Wadsworth FH (1990) and monitoring. Man and the Tropical Forestry. 17 pp. bean Forester 11: 38-47. Dacryodes excelsa Vahl, tabo- Biosphere Series. Vol. 22. Ewel JJ, Whitmore JL (1973) The nuco, Burseraceae, Bursera Wadsworth FH (1951) Forest man- UNESCO and the Parthenon ecological life zones of Puerto family. p. 284-287. In Burns agement in the Luquillo Publishing Group. Carnforth, Rico and the U.S. Virgin Is- RM, Honkala BH (Eds.) Silvics Mountains, I. The setting. Car- Lancashire, UK: (Chapter 17). lands. USDA Forest Service of North America, Vol. 2, hard- ibbean Forester 12: 93-114. pp. 337-352. Research Paper ITF-18. Río woods. Agriculture Handbook Wadsworth FH (1953) New obser- Weaver PL, Gillespie AJR (1992) Piedras, PR. 72 pp. 654. U.S. Department of Agri- vations on tree growth in Tree biomass equations for the Foggie A (1945) On girth incre- culture, Forest Service. Wash- tabonuco forest. Caribbean forests of the Luquillo Moun- ment and age of trees with ington, DC. 877 pp. Forester 14: 106-111. tains, Puerto Rico. Common- special reference to manage- Lugo AE, Applefield M, Pool DJ, Walker LR, Brokaw NVL, Lodge wealth Forestry Review 71: 35- ment in irregular tropical for- McDonald RB (1982) The im- DJ, Waide RB (Eds) (1991) 39.

258 MAY 2002, VOL. 27 Nº 5