xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xlink http:www.w3.org1999xlink xmlns:xsi http:www.w3.org2001XMLSchema-instance
leader 00000nas 2200000Ka 4500
controlfield tag 008 000000c19749999pautr p s 0 0eng d
datafield ind1 8 ind2 024
subfield code a M39-00349
Gray, Miles, T.
Cambios foliares en Saurauia montana a lo largo de un gradiente altitudinal de la fertilidad y el suelo
Foliar changes in Saurauia montana along an altitudinal and soil fertility gradient
Despite high productivity, tropical forests are found on soils with low soil fertility. The creation of new biomass requires nutrients, so trees on these soils are more efficient at using and cycling nutrients. Tropical soils have been found to show soil fertility gradients in relation to annual precipitation, seasonality, age, and altitude. While tropical ecosystems and single temperate trees have been shown to respond to soil fertility gradients, similar studies of tropical trees are sparse. Here, I study a single species of tropical tree,
Saurauia montana, from 1530 to 1785 m, to find changes in leaf toughness, specific leaf area, and foliar phosphorous levels. I found a slight decrease in phosphorous levels with altitude along with slight decreases in calcium, magnesium, potassium, iron, and nitrogen. S. montana leaves differed altitudinally with increased toughness, decreased specific leaf area, and decreased leaflitter phosphorous levels with increasing altitude. Corresponding changes in decomposition rates of leaves along the gradient were not found. Leaflitter levels suggest that retranslocation occurs more at higher altitudes, but calculated retranslocation values showed no trend. With altitude, the leaf toughness increased, the specific leaf area decreased, and leaflitter phosphorous levels decreased, but none of these changes was shown to be directly
related to soil phosphorous levels. So, the foliar changes in S. montana may be related to another environmental factor that varies with altitude.
A pesar de su alta productividad, los bosques tropicales se encuentran en suelos de fertilidad baja. La creacin de biomasa nueva requiere nutrientes, as que los rboles en estos suelos son ms eficientes en la utilizacin y reciclamiento de nutrientes. Se ha demostrado que los suelos tropicales presentan gradientes de fertilidad que varan con la precipitacin anual, la estacin, la edad, y la elevacin. Mientras que los ecosistemas tropicales y los rboles individuales de bosques templados responden a declives de fertilidad del suelo, los estudios semejantes de rboles tropicales son escasos. Este estudio consider una sola especie de rbol tropical, Saurauia montana, de 1,530 a 1,785 m, para determinar los cambios en la dureza de las hojas, en rea especfica de las hojas y en los niveles de fsforo foliar. Se encontraron disminuciones leves en los niveles de fsforo en la capa de hojas descompuestas con la elevacin, as como disminuciones leves en las concentraciones de calcio, magnesio, potasio, hierro, y nitrgeno. Las hojas de S. montana variaron altitudinalmente con los incrementos en la dureza de las hojas, con los decrecimientos en el rea especfica de las hojas, y con la disminucin de las concentraciones de fsforo con el incremento en elevacin. No se encontraron cambios correspondientes en las tasas de descomposicin de hojas debidos al gradiente. Los niveles de hojas descompuestas sugieren que la retranslocacin sucede ms frecuentemente en altitudes ms altas; sin embargo, los valores de retranslocacin calculados no mostraron ninguna tendencia. La dureza de las hojas aument, el rea especfica de las hojas disminuy, y los niveles de fsforo de la capa de hojas descompuestas disminuy; pero ninguno de estos cambios estuvo relacionado directamente con los niveles de fsforo del suelo. De esta manera, los cambios foliares en S. montana pueden estar relacionados a otro factor ambiental que vara con la elevacin.
Text in English.
Costa Rica--Puntarenas--Monteverde Zone
Morfologa de plantas
Costa Rica--Puntarenas--Zona de Monteverde
Tropical Ecology Spring 2005
Ecologa Tropical Primavera 2005
t Monteverde Institute : Tropical Ecology
1 Foliar changes in Saurauia montana along an altitudinal and soil fertility gradient Myles T. Gray Department of Earth and Atmospheric Science, Cornell University ABSTRACT Despite high productivity, tropical forests are found on soils with low soil fer tility. The creation of new biomass requires nutrients, so trees on these soils are more efficient at using and cycling nutrients. Tropical soils have been found to show soil fertility gradients in relation to annual precipitation, seasonality, age, and altitude. While tropical ecosystems and single temperate trees have been shown to respond to soil fertility gradients, similar studies of tropical trees are sparse. Here, I study a single species of tropical tree, Saurauia montana , from 1530 to 1785 m, t o find changes in leaf toughness, specific leaf area, and foliar phosphorous levels. I found a slight decrease in phosphorous levels with altitude along with slight decreases in calcium, magnesium, potassium, iron, and nitrogen. S. montana leaves differe d altitudinally with increased toughness, decreased specific leaf area, and decreased leaflitter phosphorous levels with increasing altitude. Corresponding changes in decomposition rates of leaves along the gradient were not found. Leaflitter levels sugg est that retranslocation occurs more at higher altitudes, but calculated retranslocation values showed no trend. With altitude, the leaf toughness increased, the specific leaf area decreased, and leaflitter phosphorous levels decreased, but none of these changes was shown to be directly related to soil phosphorous levels. So, the foliar changes in S. montana may be related to another environmental factor that varies with altitude. RESUMEN A pesar de su alta productividad, los bosques tropicales se en cuentran en suelos de fertilidad baja. La creaciÃ³n de biomasa nueva requiere nutrientes, asÃ que los Ã¡rboles en estos suelos son mÃ¡s eficientes en la utilizaciÃ³n y reciclamiento de nutrients. Se ha demostrado que los suelos tropicales presentan gradients d e fertilidad que varÃan con la precipitaciÃ³n anual, la estaciÃ³n, la edad, y la elevaciÃ³n. Mientras que los ecosistemas tropicales y los Ã¡rboles individuales de bosques templados responden a declives de fertilidad del suelo, los estudios semejantes de Ã¡rbol es tropicales son escasos. Este estudio considerÃ³ una sola especie de Ã¡rbol tropical, Saurauia montana , de 1,530 a 1,785 m, para determinar los cambios en la dureza de las hojas, en Ã¡rea especÃfica de las hojas y en los niveles de fÃ³sforo foliar. Se encon traron disminuciones leves en los niveles de fÃ³sforo en la capa de hojas descompuestas con la elevaciÃ³n, asÃ como disminuciones leves en las concentraciones de calcio, magnesio, potasio, hierro, y nitrÃ³geno. Las hojas de S. montana variaron altitudinalment e con los incrementos en la dureza de las hojas, con los decrecimientos en el Ã¡rea especÃfica de las hojas, y con la disminuciÃ³n de las concentraciones de fÃ³sforo con el incremento en elevaciÃ³n. No se encontraron cambios correspondientes en las tasas de de scomposiciÃ³n de hojas debidos al gradiente. Los niveles de hojas descompuestas sugieren que la retranslocaciÃ³n sucede mÃ¡s frecuentemente en altitudes mÃ¡s altas; sin embargo, los valores de retranslocaciÃ³n calculados no mostraron ninguna tendencia. La dure za de las hojas aumentÃ³, el Ã¡rea especÃfica de las hojas disminuyÃ³, y los niveles de fÃ³sforo de la capa de hojas descompuestas disminuyÃ³; pero ninguno de estos cambios estuvo relacionado directamente con los niveles de fÃ³sforo del suelo. De esta manera, lo s cambios foliares en S. montana pueden estar relacionados a otro factor ambiental que varÃa con la elevaciÃ³n.
2 INTRODUCTION The majority of tropical forests are found on moderate to very infertile soils (Vitousek and Sanford 1986), yet they maintain hig her average productivity than forests at higher latitudes (Jordan 1985). To maintain high productivity on nutrient poor soils tropical forests use nutrients more efficiently in many ways, and these tend to occur in conjunction with one another (Vitousek a nd Sanford 1986, Jordan 1985). The various ways of using nutrients more efficiently can be broken into two general categories: nutrient use efficiency and nutrient cycling efficiency. Trees growing on nutrient poor soils can create a larger amount of bio mass with a given amount of nutrients (Vitousek 1984). Chapin (1980) called this nutrient use efficiency and asserted that trees with higher nutrient use efficiency should have lower nutrient concentrations. Tropical forests can also have greater nutrien t cycling efficiencies by limiting nutrient losses from the system (Vitousek 1984). This is especially important in the tropics where high rainfall leads to rapid rates of leaching (Richards 1996). More efficient cycling can occur in many ways. Reduced litterfall (which usually correlates to longer leaf life);(Jordan 1985) increases efficiency, and Vitousek (1982) found that those tropical sites with lower soil fertility produced less leaf litter. Roots near the surface of the soil also help to increase cycling efficiency by absorbing nutrients from leaf litter as soon as they are mineralized by decomposition (Jordan 1985). As soil fertility decreases, the percentage of roots in the superficial root mat increases (Vitousek and Sanford 1986, Jordan 1985). Tropical tress also tend to allocate a larger amount of production to the roots, and thus have greater root:shoot ratios to cycle nutrients more efficiently. With a larger relative amount of roots, trees have the ability to uptake a greater proportion o f available nutrients. Trees on nutrient poor soils also tend to reduce nutrient losses by producing tougher, more lignin rich sclerophyllous leaves with lower specific leaf areas (Richards 1996). Because litter quality has a profound effect upon decompo sition rates (Brady and Weil 1996, Kwabiah et al. 2001), lignin rich sclerophyllous litter decomposes more slowly (Jordan 1985), and reduces the rate at which nutrients are released from the leaflitter (Kwabiah et al. 2001). These slower rates of decompos ition limit the rate of nutrient mineralization, providing more time for roots to uptake all available nutrients before they are leached from the system. Thick sclerophyllous leaves also tend to be long lived and so help to reduce the total leaf litter (G rubb 1977). Greater amounts of retranslocation, that is the removal of nutrients from leaves to woody tissue prior to abscission, can also increase cycling efficiency (Richards 1996, Vitousek 1984). Vitousek and Sanford (1986) showed that forests on more low fertility soils retranslocate a larger percentage of nutrients from leaves prior to abscission. Although the majority of tropical forests have been found to have low soil fertility, there is a large amount of variation in response to a variety of fac tors (Vitousek and Sanford 1986, Richards 1996). The parent material of soils is often the most important factor regarding soil fertility. This is especially true for andisols, those soils derived from volcanic parent material, which tend to have higher than normal soil fertility (Brady and Weil 1996). Annual precipitation also affects soil fertility through leaching, with higher amounts of precipitation tending to decrease soil fertility (Brady and Weil 1996). The overall low fertility of tropical soil s can in large part be attributed to greater
3 rates of leaching caused by increased precipitation (Richards 1996). The seasonality of a site also has an effect on soil fertility. Tropical sites with year round rainfall experience year round leaching and t end to have lower soil fertility, while those tropical sites with more pronounced dry seasons tend to have higher soil fertility. Higher fertility in dry forests can be due to the high evaporation rates during the dry season, which can pull water and diss olved nutrients to the surface of the soil through the capillary action of water (Brady and Weil 1996). Altitude can also impact the soil fertility, with higher altitude sites tending to have higher soil fertility (Vitousek and Sanford 1986). The age of soils also has a profound impact on soils, with older soils tending to have lower soils fertility as a result of leaching over many years. This is of particular importance in the tropics, where many soils are very old and intensely weathered and so have l ow soil fertility (Brady and Weil 1996). The fertility of andisols is also dependant upon the age. Because phosphorous (P) is derived from the weathering of parent rock (Schlesinger 1991) young Andisols tend to have high P content. These young soils als o tend to be nitrogen (N) limited because N is ultimately derived from biological fixation (Schlesinger 1991), and young soils do not yet have large amounts of N fixing bacteria (Vitousek and Farrington 1997). As these Andisols progress in age they become highly weathered and lose most P, while accumulating N fixing bacteria and become P limited (Vitousek and Farrington 1997). The andisitic soils of Monteverde, Costa Rica, like most tropical montane soils, tend to have relatively high soil fertility in re lation to lowland sites (Nadkarni and Wheelright 2000). Although the soils of Monteverde have greater soil fertility than lowland sites, the soil fertility has been shown to decrease with increasing altitude (Chan 2002). However, Grubb (1977) found that total soil nutrient levels are not good indicators of available nutrients, and that montane soils often have a smaller percentage of available nutrients. One explanation is that decreased rates of transpiration in montane forests decrease the ability of plants to pull water and nutrients to the root hairs (Grubb 1977). Another possible explanation is that lower temperatures lead to decreased rates of mineralization of organic matter by microbes, and thus decreased rates of nutrient supply (Edwards 1977, Brady and Weil 1996). This is particularly important for montane sites which tend to have more organic matter in the soil and a greater accumulation of leaflitter (Edawards 1977). While ecosystem level changes in response to soil fertility have been wel l documented, relatively little has been documented about the responses of individual species. Demars and Boerner (1997) examined foliar nutrient concentrations of Lonicera maackii (Caprifoliaceae) along a soil P gradient in Ohio, U.S.A. They found incre asing leaf P concentrations and decreasing P retranslocation with increasing soil P. Knops and Reinhart (2000) studied several perennial grasses (Poaceae) along an artificially created soil N gradient. They found that the specific leaf area of the grasse s increased with increasing N fertilization. Similar studies of tropical plants are sparse. Saurauia montana (Actiniadaceae) is a tropical plant that occurs along a wide altitudinal band. With increasing altitude, S. montana may encounter less nutrient rich soil. How does S. montana react to changes in soil fertility with altitude? I this study I looked for changes in soil fertility with altitude and corresponding changes in S. montana foliar characteristics, including retranslocation, leaf toughness, specific leaf area, and decomposition rates.
4 MATERIALS AND METHODS Study Species and Site Characteristics S. montana (Actinidiaceae) is a small tree commonly found along the forest edge in the Monteverde area. The study took place in lower montane w et between April 15 th and May 9 th 2005. Eight individuals of S. montana were located along the road to Cerro Amigos between 1530 and 1785 m and the altitude of each tree was found using an altimeter. Soil Fertility A soil sample was taken within two m of each individual of S. montana using a soil auger constructed from 4.5 cm diameter PVC pipe. The samples were taken to a depth of 20 cm but only the bottom 5 cm of the sample was collected. The samples were then placed in a dry room (a room wi th a dehumidifier) for three days. Once the soil was dry, relative P concentrations were found using the LaMotte Soil Test Kit along with a spectrophotometer. A modified LaMotte P test was used. Two grams of soil were weighed using an electronic scale a nd then placed into a plastic soil tube. The tube was then filled with Universal Extracting Solution to the 14 ml line. This solution was mixed for one minute and then poured through a filter to remove all the soil. The resulting solution was used in th e phosphorous test procedure (LaMotte, 2002). The blue liquid produced was poured into a spectrophotometer cuvette and topped off with distilled water. The cuvette was capped, shaken, and wiped, then inserted into the spectrophotometer and the transmitta nce was recorded. Before use each day the spectrophotometer was calibrated using a blank of distilled water. Because this test used a Universal Extracting Solution, the total soil P is tested, not the available P in the soil. Also, it was not possible t o create a solution of known P concentration to create a calibration curve; so all P levels that were found were relative P levels. Two additional soil samples of 250 g were taken, one from the site at 1530 m and one from the site at 1785. These two samp les were sent to the Laboratorio de Suelos at the Ministerio de Agricultura, Costa Rica. These samples were tested for a variety of plant essential nutrients and some other common soil nutrients. Leaf Morphology Specific leaf area was calculat ed, and the leaf penetrability was determined. Ten live leaves were collected from each individual of S. montana . The leaves were chosen two to three leaves down from tip of the leaf cluster to assure that all the leaves were of the same approximate age. These leaves were first used to find the leaf toughness using a leaf penetrometer, with leaf penetrability used as a measure of leaf toughness. Each leaf was inserted into the penetrometer and aligned as far as possible from any primary or secondary ve ins. The platform was then inserted and weighted with coins until the leaves
5 broke. The coins were then weighed without the last coin, and the value was recorded as the penetrability. To find specific leaf area, the same ten leaves were used. The area of each leaf was determined using an herbivory grid, which is a transparent piece of graph paper. The leaves were all labeled with masking tape, placed in open plastic containers, and separated by pieces of paper. The bins with the leaves were placed in a dry room for four days, after which time the dry weight of each leaf was taken using an electronic scale. The specific leaf area was then found as cm 2 /g by dividing the area of each leaf by the weight of each leaf. Leaf Decomposition Because rates of decomposition are related both to leaf properties and properties to site where decomposition occurs, a leaf decomposition test was carried out. Eighteen extra leaves were collected from only the upper most individual at 1785 m and the lower most indivi dual at 1530 m. All eighteen leaves from each individual were weighed using an electronic scale. The leaves were each placed in separate mesh litterbags and labeled. Nine bags from each individual tree were then placed at 1785 m, and nine bags were plac ed at 1530 m. This was done to find whether the rates of decomposition were related more to the site where decomposition occurred, or to the properties of the leaves themselves. The bags were collected 24 days after being placed at the two sites, and the final weight of each leaf was found using an electronic scale immediately after being collected. A ratio of decomposition was made by dividing the original leaf weight by the final weight. Retranslocation Three pre abscised dead leaves were collected from each individual using a slingshot. Each of these leaves was placed in a dry room for three days. Three leftover dried green leaves from the specific leaf area test were also used from each individual. After they were dried the relative P concentra tions of each live and dead leaf was measured using the LaMotte Soil Kit (LaMotte, 2002) along with the spectrophotometer. This was done by modifying the green tissue P test. The leaves were crushed and 0.4 g of each leaf was placed into a soil tube, whi ch was then filled to the 14 ml line with Universal Extracting Solution. This was shaken for one minute and then passed through a filter. The resultant solution was used in the LaMotte P test. These relative P concentrations were used to compare live an d dead leaf P levels to find the amount of retranslocation. The mean dead leaf transmittance was subtracted from the mean live leaf transmittance to calculate retranslocation. RESULTS Soil Samples Seven of the eight soil samples looked the same, hav ing a light to dark brown color. However, the sample from 1675 m was orange in color a felt very clayey. The measures
6 of soil P generally decreased with altitude from 1530 to 1785 m (Fig. 1). Transmittance values ranged from 33.5 to 67.5%. The soil P l evels for the site at 1675 m were consistently the lowest, with transmittance values between 33.5 and 37.2%. P measurements from MAG did not differ in the samples from 1530 and 1785 m, with both having readings of 2 mg/l. The data also showed that Ca, Mg , K, Zn, Fe, and N were lower for the sample from 1785 m (Table 1). Leaf Morphology Leaves became progressively tougher as altitude increased (Fig. 2). The toughness ranged from 99.3 to 219.3 g. However, toughness did not respond to soil fertility ( Fig. 4). Specific leaf area decreased with altitude (Fig. 3), but did not respond to soil fertility (Fig. 5). Specific leaf area ranged from 86.22 to 201.03 cm 2 /g, the changes were due to changing weights, and not due to changing areas. Leaf Decompos ition All leaves lost weight. The ratio of original weight to final weight varied between 1.625 and 3.5. The ratio of decomposition did not differ between the leaves from 1785 m and the leaves from 1530 m at either decomposition site (Fig. 6 and 7). Th e ratios of decomposition for all of the leaves decomposed at 1785 m were significantly smaller than the ratios for leaves decomposed at 1530 m (Fig. 8). Retranslocation Relative live leaf P levels did not change with altitude (Fig. 9). Live leaf P transmittance values ranged from 2.8 to 33.6%. Relative leaflitter levels decreased with altitude (Fig. 10), but did not vary with soil P levels (Fig. 11). Leaflitter P transmittance values ranged from 32 to 65.4 %. Calculated retranslocation values did not differ with altitude (Fig. 12), or with soil P levels (Fig. 13). DISCUSSION S. montana responds to altitude by increasing its toughness and decreasing its specific leaf area. The P levels in litterfall decreased with altitude but P levels in live l eaves did not differ with altitude. While soil P decreases with altitude, the difference is very small and there is no direct relationship with foliar characteristics. Further, while retranslocation may occur, as suggested by leaflitter P values, no evid ence was found by comparing live leaf P levels and litterfall P levels. Decomposition rates were not shown to be related to leaf characteristics, but were shown to be related to the site where decomposition occurred. Increasing leaf toughness and speci fic leaf area show that as altitude increases, leaves of S. montana become more sclerophyllous. The tree is reacting to some environmental factor that changes with altitude. The fact that the decomposition test showed no difference was not surprising, as it seems that the design was flawed. Leaf decomposition is a long process, and the rate of decomposition is linked to water
7 availability (Brady and Weil 1996). Because there was very little rain during the study, I would expect decomposition rates at bo th sites to have been negligible, and so no difference would have been found between the leaf types. Because the leaves were placed at the sites when they were green, all the leaves lost weight over the study period, mainly due to drying. However, the am ount of weight lost between the leaves at the different decomposition sites was found to be significant, with the leaves that decomposed at 1530 m losing significantly more weight than those that decomposed at 1785 m. As higher sites generally tend to hav e more mist, and thus higher moisture, it seems logical that the leaves at the lower drier site would have dried more than the leaves at the higher wetter site, and thus lost more weight. Soil P declines from 1530 to 1785 m, but only slightly over this range. Other factors of soil fertility were shown to be lower at the site at 1785 m by the MAG. These and the soil P differences were probably due to increased precipitation at the higher site, leading to increased leaching and lower nutrient levels. Th e site at 1675 m was a clear outlier from the line of best fit for soil P levels, and was also visually distinct with high clay content. Due to the very small size of clay particles they have very high surface areas, and so have they tend to have very hig h CEC (Cation Exchange Capacity). However, the high exchange capacity of most tropical soils with high clay contents tend to be dominated by hydrogen and aluminum ions, leaving few exchange sites for essential cations (Brady and Weil 1996). So even thoug h soils with high clay contents have the ability to have high fertility, they rarely do. However, this site could be an exception to the rule. The amount of P in leaflitter declined with altitude, indicating that retranslocation of P may be variable wi th altitude. However, calculated retranslocation values did not correlate with altitude, mainly because green leaf P levels were not shown to correlate with altitude. This could have been due to variable leaf sampling in relation to the amount of sunlight hitting leaves. Leaves in direct sunlight have higher nutrient concentrations (Vitousek and Sanford 1986). Because the leaves of S. montana are clustered, the amount of light hitting each leaf could be highly variable and was not controlled for. Given that no trend was found for the green leaf P levels, I believe that retranslocation may still have been more pronounced at higher altitudes due to the fact that leaflitter P levels decreased with altitude. Controlling for light levels and repeating the te st could yield more conclusive evidence. The data did not show any relation between soil P levels and leaf characteristics. This could have been due to the fact that total soil nutrient levels are not always good indicators of available nutrients (Vitous ek and Sanford 1986). Available nutrient levels are hard to measure, and many different indices have been developed to approximate them (Chapin et al. 1986). In particular this could be an effect of the site at 1675 m, which had exceptionally high relati ve soil P levels. If the clay soil at this site held tightly to its nutrients, the available phosphorous levels at this site would have been substantially lower than the total P that was found using the soil test kit. The other samples had lower clay con tents, and so larger particle sizes, and so may not have held nutrients as tightly. The fact that leaf characteristics did not respond to soil P levels could have also been due to the fact that the soils of Monteverde might not be P limited. It has been found that tropical montane soils are commonly N limited (Tanner et al. 1990, Vitousek and Farrington 1997) due to low rates of N mineralization (Vitousek and Sanford 1986), and
8 can have high levels of P (Edwards and Grubb 1982) unlike most tropical lowlan d forests. If N were the limiting nutrient of the soils of Monteverde, I would not expect these foliar morphological characteristics to vary with P levels, but instead with N levels. However, given that the leaflitter P levels were shown to change with al titude, it is unlikely that the system is N limited, otherwise P would not be conserved. The soil tests done by the Laboratorio de Suelos showed only a .02% difference in the N content of the soil at 1530 m compared to that at 1785 m. Thus it seems unlik ely that the observed changes in foliar characteristics could be due to such minute variation. A third possible explanation has to due with the causes of sclerophyllous leaves. Plants produce sclerophyllous leaves in response to herbivore pressure, water availability, and also nutrient availability (Vitousek and Sanford 1986). Water availability probably does not cause leaves of Saurauia montana to be sclerophyllous, as the water availability generally increases with increasing altitude. Likewise, I woul d not expect there to be any great deal more herbivory at higher altitudes in Monteverde, so herbivory is not likely to be a driving force. The increasingly sclerophyllous could be a response to resist wind damage at higher sites which tend to be windier. Plants respond to changes along narrow altitudinal gradients, as shown here. That a single species of tree responds from 1530 to 1785 m suggests that other plant responses to soil fertility and altitudinal gradients may occur on small spatial scales. Because soil Weil 1996), corresponding changes in foliar characteristics of plants on small spatial scales could have drastic ecological effects. ACKNOWLEDGEME NTS I would like to thank Alan Masters for guiding me through my project and making me laugh. Thanks to Bill Haber and Willow Zuchowski for helping me to find a suitable tree species for my study. Thanks to Matt and Ollie for helping with all of the lit tle questions and for keeping me entertained during breaks and late at night. Thank you to Javier Mendez for translating my abstract, and to Jesse and Phil for editing my paper. Thank you to all of the wonderful students on this trip for making this all the better. And finally thank you to my parents for helping me to do the things I love. LITERATURE CITED Brady, N. C. and Weil. R. R. 1996. The Nature and Properties of Soils. Prentice Hall, Upper Saddle River, New Jersey. Chan, J. W. 2002. A Compari son of Canopy and Forest Floor Soils Along an Altitudinal Gradient in Monteverde, Costa Rica. EAP Costa Rica Tropical Ecology Program, Spring 2002. Chapin, F. S. 1980. The Mineral Nutrition of Wild Plants. Annual Review Ecology and Systematics 11:233 260. Chapin, F. S., Vitousek, P. M., Cleve, K. V. 1985. The Nature of Nutrient Limitation in Plant Communities. The American Naturalist 127: 48 58. Demars, B. G. and Boerner, R. E. J. 1997. Foliar Nutrient Dynamics and Resorption in Naturalized Lonicera maackii (Caprifoliaceae) Populations in Ohio, USA. American Journal of Botany 84: 112 117. Edwards, P. J. 1977. Studies of Mineral Cycling in a Montane Rain Forest in New Guinea: II. The Production and Disappearance of Litter. The Journal of Ecology 65: 971 992.
9 Edwards, P. J. and Grubb, P. J. 1982. Studies of Mineral Cycling in a Montane Rain Forest in New Guinea: IV. Soil Characteristics and the Division of Mineral Elements between the Vegetation and Soil. The Journal of Ecology 70: 649 666. Grubb, P. J. 1977. C ontrol of Forest Grwoth and Distribution on Wet Tropical Mountains: With Special Reference to Mineral Nutrition. Annual Review of Ecology and Systematics 8: 83 107. Haber, W. A., Zuchowski, W, and Bello, E. 2000. An Introduction to Cloud Forest Trees Mont everde, Costa Rica. Mountain Gem Publications, Monteverde de Puntarenas, Costa Rica. Jordan, C. F. 1985. Nutrient Cycling in Tropical Forest Ecosystems. John Wiley and Sons, New York. Knops, J. and Reinhart, K. 2000. Specific Leaf Area along a Nitrogen Fe rtilization Gradient. American Midland Naturalist 144: 265 272. Kwabiah, A. B., Stoskopf, N. C., Voroney, R. P., and Palm, C. A. Nitrogen and Phosphorous Release from Decomposing Leaves under Sub Humid Tropical Conditions. Biotropica 33: 229 240. LaMotte. 2002. LaMotte Company. Chestertown, Maryland. Nadkarni, N. M. and Wheelright, N. T. 2000. Monteverde Ecology and Conservation of a Tropical Cloud Forest. Oxford University Press, New York. Richards, P. W. 1996. The Tropical Rainforest. Cambridge University Press, Cambridge. Schlesinger, W. H. 1991. Biogeochemistry An Analysis of Global Change. Academic Press, San Diego, California. Tanner, E. V. J., Kapos, V., Freskos, S., Healey, J. R., Theobald, A. M. 1990. Nitrogen and Phosphorous Fertilization of Jamaic an Montane Forest Trees. Journal of Tropical Ecology 6: 231 238. Vitousek, P. M. 1982. Nutrient Cycling and Nutrient Use Efficiency. The American Naturalist 119: 553 572 Vitousek, P. M. 1984. Litterfall, Nutrient Cycling, and Nutrient Limitation in Tropica l Forests. Ecology 65:285 298. Vitousek, P. M. and Farrington, H. 1997. Nutrient Limitation and Soil Development: Experimental Test of a Biogeochemical Theory. Biogeochemistry 37: 63 75. Vitousek, P. M. and Sanford, R. L. 1986. Nutrient Cycling in Moist Tr opical Forest. Annual Review of Ecology and Systematics 17: 137 167 TABLES TABLE 1. Soil analysis for two soil samples sent to the Laboratorio de Suelos at the University of Costa Rica. The samples were from 1530 m and 1785 m. Al, Ca, Mg, and K are rep orted as cmol/l. P, Zn, Mn, Cu, and Fe are reported as mg/l. N was reported as the percentage of the total volume. Site pH Al Ca Mg K P Zn Mn Cu Fe N 1530 5.9 .3 3.4 .6 .10 2 .5 1 4 27 .16 1785 5.6 .3 2.0 .4 .03 2 .4 1 5 19 .14
10 FIGURES FIGURE 1. Soil phosphorous levels decreased with altitude. Soil samples were taken near each individual of Saurauia montana . Relative soil phosphorous levels are reported as transmittance as measured by a spectrophotometer. Higher values of transmittance correspond to lower soil phosphorous concentrations, so soil phosphorous values decrease with increasing altitude. Linear regression showed a significant negative correlation between relative soil phosphorous levels and al titude (R 2 = .188, p = .0344, n = 24).
11 FIGURE 2. Leaves of Saurauia montana became tougher with altitude. Linear regression showed a significant positive relationship (R 2 = .122, p = .0015, n = 80). FIGURE 3. Specific leaf area of Saurauia montana declined with altitude. Leaves became heavier, not smaller. Linear regression showed significant negative correlation (R 2 = .175, p = .0001, n = 79).
12 FIGURE 4. No relat ion between soil P levels and toughness of Saurauia montana leaves . Relative soil phosphorous levels are reported as transmittance as found by a spectrophotometer. Higher values of transmittance correspond to lower soil phosphorous levels. Linear regr ession showed no significant correlation (R 2 = .033, p = .1087, n = 80). FIGURE 5. No relation between specific leaf area of Saurauia montana and soil P levels. Relative soil phosphorous levels are reported as transmittance as found by a spectrophotometer. Higher values of transmittance correspond to lower soil phosphorous levels. Linear regression showed no significant correlation (R 2 = .003, p = .6345, n = 79).
13 FIGURE 6. Mean ratios of original weight to final weight of leaves of Saurauia montana decomposed at 1530 m. Altitude of tree individual of origin is shown in the legend. Error bars are 1 SE from the mean. Leaves from 1530 m had a mean ratio of 2.66 (SE = .153). Leaves from 1785 m had a mean ratio of 2.57 (SE = .087). Unpaired t test showed no significant difference between ratios for leaves at 1530 m compared to ratios for leaves from 1785 m (t value = .537, DF = 16, p = .5983). FIGURE 7. Mean ratios of ori ginal weight to final weight of leaves of Saurauia montana decomposed at 1785 m. Altitude of tree individual of origin is shown in the legend. Error bars are 1 SE from the mean. Leaves from 1530 m had a mean ratio of 1.97 (SE = .087). Leaves from 1785 had a mean ratio of 1.99 (SE = .065). Unpaired t test showed no significant difference between ratios for leaves from 1530 m compared to ratios of leaves from 1785 m (t value = .188, DF = 16, p = .8529).
14 FIGURE 8. Mean r atios of original weight to final weight of all leaves of Saurauia montana decomposed at 1530 m compared to all leaves decomposed at 1785 m. Altitude of decomposition is shown in the legend. Error bars are 1 SE from the mean. Leaves decomposed at 1530 m had a mean ratio of 2.61 (SE = .086). Leaves decomposed at 1785 m had a mean ratio of 1.98 (SE = .055). Unpaired t test showed the ratio of leaves from 1530 to be significantly higher than leaves at 1785 (t value = 6.173, DF = 34, p < .0001). FIGURE 9. No relation between green leaf phosphorous levels of Saurauia montana and altitude. Relative live leaf phosphorous levels are reported as transmittance as found by a spectrophotometer. Higher values of transmittance corresp ond to lower live leaf phosphorous levels. A linear regression showed no significant correlation (R 2 = .032, p = .4024, n = 24).
15 FIGURE 10. Leaflitter phosphorous levels of Saurauia montana leaves decreased with altitude. Relative leaflitter phosphorous levels are reported as transmittance as found with a spectrophotometer. Higher transmittance values correspond to lower leaflitter phosphorous levels. Linear regression showed a significant negative correlation between l eaflitter phosphorous levels and altitude (R 2 = .217, p = .0334, n = 21). FIGURE 11. No relation between litterfall phosphorous levels of Saurauia montana leaves and relative soil phosphorous levels. Relative leaflitter and soil phosphorous levels are reported as transmittance as found with a spectrophotometer. Higher values of transmittance correspond to lower phosphorous levels. Linear regression no significant correlation between relative litterfall phosphorous levels an d relative soil phosphorous levels (R 2 = .046, p = .3522, n = 21).
16 FIGURE 12. No reltion between calculated retranslocation of phosphorous from leaves of Saurauia montana and altitude. Relative phosphorous levels were found a s transmittance with a spectrophotometer, so higher transmittance readings correspond to lower phosphorous levels. Retranslocation was found by subtracting the average live leaf transmittance from the average dead leaf transmittance. A linear regression found no significant correlation (R 2 = .200, p = .3150, n = 7). FIGURE 13. No relation between calculated retranslocation of phosphorous from leaves of Saurauia montana and relative soil phosphorous levels. Relative phosphoro us levels were found as transmittance with a spectrophotometer, so higher transmittance values correspond to lower phosphorous levels. Retranslocation was found by subtracting the average live leaf transmittance from the average dead leaf transmittance. A linear regression showed no significant correlation (R 2 = .117, p = .4522, n = 7).