Microb. Ecol. 6:349-355 (1980)
N 2 Fixation
DO~I:IL
CCOI.~
(C2H 2 Reduction) by Epiphylls on Coffee, Coffea arabica
Joann P. Roskoski InstitutoNacionalde lnvestigacionessobreRecursosBioticos,ApartadoPostal63, Xalapa,Veracruz,Mexico Abstract. Nitrogen fixation (C2H2 reduction) by epiphylls on coffee, Coffea arabica, grown in sites with different degrees of shade, was determined. Coffee leaves with nitrogen-fixing epiphylls were found in all sites in approximately equal numbers. Rates of C2H2 reduction were similar for all sites and throughout the year, averaging 3.21 nmoles C2H 2 reduced leaf with epiphylls-I d a y - I . Apparently, neither rates of activity nor abundance of leaves with nitrogen-fixing epiphylls is related to the degree of shade in a site. No correlation was found between percent epiphyll cover and the presence or magnitude of nitrogen-fixing activity. Calculated annual fixation by epiphylls on coffee was low, ranging from 0.7 g N2 ha -l year -I for the shadeless site to 1.4 g N2 ha -I year -~ for the site with lngafinicuil shade trees. These results suggest that epiphyll fixation is not an important source of nitrogen for the coffee ecosystem studied.
In 1965 Ruinen (11) observed that the total nitrogen content of petri dishes, containing nutrient media and coffee leaves with epiphyllic organisms, increased through time. The nitrogen increase was ascribed to nitrogen fixation by the leafepiphylls. Later, nitrogenfixing bacteria were isolated from leaves of I0 tree species in England (4-6), and from foliage of various tropical plants (12). Nitrogen fixation by leaf epiphylls has been suggested as an important source of nitrogen in some ecosystems (5, 12) but not in others (13). For example, Jones (5) calculated that up to 15 kg N2 h a - l year- 1 could be added to an English coniferous forest via epiphyll fixation, whereas Sucoff (13) found that the annual nitrogen input from epiphyll fixation in coniferous and deciduous forests of the United States was only between 0.2 and 0.06 kg ha -1 year -I . Maximum epiphyll development probably occurs in the tropics where leaf longevity, atmospheric moisture levels, and ambient temperatures favor abundant growth of a diverse epiphyll community (10). However, to date, no seasonal quantitative studies on epiphyllic fixation in tropical ecosystems have been reported. Previous evidence suggested that epiphyI1 fixation occurred on coffee leaves (11). Therefore, during an ecological study of a coffee agro-ecosystem, the question of epiphyll fixation was examined. The study focused on understory vegetation in order to concentrate on the strata where epiphylls were presumably most abundant (9). This 0095-3628180/0006-0349 $01.40 9 1980 Springer-VerlagNew Yorklne.
35I)
I . P . Roskoski
Table 1. Shade characteristics of the four study sites
Site
Shade species None
lnga vera H.B+ & K. Inga jinicuil SehL Musa sapientum L, Citrus sinensis Osb. tnga vera lnga jinicuil
Cover by shade speciesa (m 2 h a - 1) 0 8,237 12,792 40,374
aCover data from Jimenez (2).
paper presents the results of an investigation to quantify annual nitrogen fixation by epiphylls on coffee plants grown in sites with different degrees of shade.
Materials and Methods Site Description The study was conducted in a coffee plantation near the city o f Xalapa, Veracruz, Mexico, at 19~ ' latitude north, 96~ , longitude west, and an attitude of 1225 m above sea level. The climate is classified as semi-hot, humid, with warm summers and coot winters (2). Annual rr~an temperature is 19~ • 2~ and precipitation averages 1758 • 193 mm annually+ Four sites within the plantation, each differing primarily in the type and degree of shade present, were located, Table t presents the shade characteristics of the sites. Coffee cover and density were similar in all four areas.
Nitrogen Fixation Armfyses Once each month, from February I978 through February [979, twenty coffee leaves with visible epiphyIlie growth, two from each of 10 m ~ , were randomly collected from each site. Each leaf was it~m~..diatety placed in a separate 20 ml test tub~. Tubes were stoppered and incubated at 20~C tbr 24 hours under ambient laboratory, light levels. It was felt that these ie,cubation conditions approximated those experienced by epiphylls in the coffee plantation. After 24 hours a 0.2 cc gas sample was withdrawn from each tube and injected into a Curie model 9500 l i d gas chromatograph, equipped with two 1.8 m columns, packed with Poropak N+ Separations were run at 60~ using nitrogen as a carder gas. This procedure measured C2H 4 production by leaves with epiphylls in the absence of C2H2+ Typical C2H 4 production was zero. Following the above analysis, each test tube was opened to allow the re-establishment of ambient atmospheric conditions in the assay tube. Tubes were then restoppered, and C2H2, equal to 10% of the test tube volume, was introduced. A 0.20 cc gas sample, withdrawn after 24 hours, was analyzed for C2H4, as descrit~.,d above. Reagent controls were routinely ran. Peak heights for C2H 4 produced by leaves with epiphylls in the presence of C2H 2, corrected for the reagent controls, and C2H 4 production by the same leaves in the absence of C2H 2, were compared to peak heights of known C2H 4 concentrations. Moles C2H 4 produced were convened to moles N 2 fixed, employing the theoretical 3:1 ratio ( 1 )+ After C2H 2 reduction analysis, a clear plastic sheet, gridded into 25 ram 2 sections, was placed over each leaf and used to d~termine total leaf area and area occupied by epiphylls. The latter was defined as any area with
Nitrogen Fixation by Epiphylls
351
altered surface texture and/or coloration, excluding diseased areas. Thus epiphyll cover includes only visible epiphylls on leaf surfaces.
Test of the C2H 2 Reduction Method The C2H 2 reduction method, described above, assumed that C2H4 production during the fLrst 24-hour incubation period, without C2H2, represented natural C2H4 production during the second 24-hour incubation. To test the validity of this assumption, 30 coffee leaves with epiphylls were collected in January 1978, prior to the initiation of the annual study. These leaves were placed in test tubes, which were then stoppered. C2H 4 production was determined after 24 hours. Tubes were then opened, re-closed after a few minutes, and C2H 4 content measured after another 24 hours. This procedure was repeated for a third day, for a total of three l-day incubations. The entire test was repeated in March 1979. Results revealed no significant differences in background C2H 4 production among days 1,2, and 3.
Epiphyll Abundance To ascertain the number of leaves with epiphylls per site, a 0.2 ha plot in each site was subdivided into sixty-four 25 m 2 subplots. In May 1978, three subplots in each site, each containing four coffee trees, were randomly selected. Line transects were run out from the base of each tree at 900 intervals. Plastic screening, extended vertically from these lines, divided each tree into four equal sections. One section of each tree was randomly chosen, and the total number of leaves and leaves with visible epiphylls Were counted. Values for each tree section were averaged by site, multiplied by four, to yield average tree totals, and then multiplied by the average number of coffee trees ha- 1. To determine if the number of coffee leaves with epiphylls, obtained by the above method, was constant, this phase of the study was repeated in February 1980, but only two coffee trees per subplot were recounted.
Results and Discussion N i t r o g e n - f i x i n g epiphylls occurred in all sites. A l t h o u g h no attempt was m a d e to identify s p e c i f i c o r g a n i s m s responsible for m e a s u r e d activity, foliocolous lichens, p i g m e n t e d bacterial colonies, and b l u e - g r e e n algae w e r e periodically o b s e r v e d on assayed leaves. Rates o f C2H2 reduction ranged from 0 . 4 to 27.3 nmoles leaf with e p i p h y i l s - l d a y - I , w h i c h equals 0.1 to 9.1 n m o l e s N 2 fixed leaf with epiphylls - I day - I . T h e s e rates, a l t h o u g h s o m e w h a t higher than those reported by S u c o f f (13) for temperate vegetation, are l o w e r than those for other tropical plants (12). A n a l y s i s o f v a r i a n c e r e v e a l e d no significant differences between sites with respect to rate o f C 2 H 2 reduction. Apparently, absence, presence, or degree o f shade does not affect the m i c r o - e n v i r o n m e n t o f nitrogen-fixing epiphylls in a m a n n e r that causes rate differences. M e a n m o n t h l y rates, on the other hand, were statistically different (Table 2). H o w e v e r , if the single values for A u g u s t and S e p t e m b e r , which are considerably higher than rates for the other months, are r e m o v e d from the analysis, the remaining rates are similar. T h i s finding implies that the m i c r o - e n v i r o n m e n t a l factors controlling nitrogenfixing activity on leaf surfaces are remarkably constant throughout the year. In c o m p a r i s o n , nitrogen fixation rates in temperate e c o s y s t e m s are markedly seasonal (5). P o o l e d m o n t h l y data resulted in an average rate o f 3.21 nmoles C2H 2 reduced leaf with activity - l d a y - 1. T h e percent o f leaves with epiphylls that reduced C2H 2 did not differ significantly
352
J. P. Roskoski
Table 2. C2H 2 reduction rates for coffee leaves with epiphyUs
Month
Active samples
January February March April May July August September October
9a 26 10 3 7 1 1 1 2
November December
3 8
C2H 2 reduced l e a f - 1 day - I (nmoles) 1.36 3.04 4.33 1.28 3.21 1.73 16,40 27.30 0.79
___0.30 b + 0.72 + 2.03 -I- 0.19 + 1.12
+ 0.00
1.15 + 0.30 2.05 + 0.47
aTotal number of samples assayed in each month was 80, with the exception of data for February, which include data for 1978 and 1979, or 160 assays performed. bValues shown are the mean + S.E. of all active samples.
between sites (Table 3). This result is somewhat surprising since significant site differences were found with respect to the percent of leaves that possessed epiphylls (Table 3). From these data it appears that degree of shading or other site characteristics influence the abundance of leaves with epiphylls but not the fraction of those leaves that have nitrogen-fixing epiphylls. No correlation was found between the areal extent of the epiphyll community on any leaf and the presence or magnitude of nitrogen-fixing activity that could be expected for that same leaf. Leaves with 0.5 to 100% epiphyll cover exhibited Similar nitrogen-fixing activity. Nitrogen-f~xing epiphylls are apparently ubiquitous throughout the development of the epiphyll community. Average percent epiphyll cover on leaves was similar for all sites but varied throughout the year (Fig. 1A). In addition, the percent of leaves with epiphylls varied with time (Table 3). These annual changes largely reflect the phenology of coffee. Coffee leaves begin to fall in February. Total replacement occurs by May, shortly before flowering (3). Epiphyll cover per leaf and number of leaves with epiphyUs are lowest in May because the majority of coffee leaves present at this time are young and have not yet been colonized. Cover gradually increases as young leaves are colonized and epiphyll growth ensues. Maximum percent cover and percent leaves with epiphylls in February is due to the large number of mature leaves present at this time. As leaves begin to fall, percent cover decreases since new leaves form a larger and larger portion of the leaf population. In comparison to the pronounced seasonal pattern described above, the percent coffee leaves with nitrogen-fixing epiphylls shows less seasonal variation (Fig. 1B). Highest percent leaves with nitrogen-fixing epiphylls occurs in February coincident with maximum percent cover and percent leaves with epiphylls. However, lowest percent leaves with nitrogen-fixing epiphylls was found not in May but slightly later, when percent cover was increasing most rapidly (Fig. 1). During June, July, and August,
Nitrogen Fixation by Epiphylls
353
4.5-
"T b3 _1
Q:
hi ::P
30
o ..i .J )-lQ_ tlJ
15
A
20 O') W W .d X
jE
,J )-
I0 I,-
1,1
.-e
o
T ,u,,,,,=
i""""
i 2"
1978
! ~,ii, ~w~ v:
l
1979 TIME
Fig. I. A % Epiphyll cover leaf- I vs time. Eighty coffee leaves with epiphylls. 20 from each study site, were collected monthly and the percent epiphyll cover detemfined, Values shown are the mean :t: S.E~ of these 80 values. B % Epiphyll leaves that fix N 2. For each site the number of epiphylt leaves that reduced C2H 2 was divided by the number of leaves assayed. Values shown are the mean +__S.E. of data for all sites,
heavy summer rains, presumably containing nitrogenous compounds (8), may supply sufficient nitrogen to the phylloplane to reduce the need for nitrogen fixation. In addition, summer rains may dilute the concentration of carbohydrates on the leaf's surface (12), which may be the primary energy source for nitrogen-fixing activity. The increase in the percent nitrogen-fixing epiphylls at the end of the year may be a function
354
J.P. Roskoski
Table 3. Site values for percent leaves with epiphylis in May 1978 and February 1980, and percent leaves with epiphylls that reduced C2H 2 Leaves with epiphylls (%)a Site
May 1978
Febrhary 1980
Leaves with epiphylls that reduced C2H 2 (%)b
I 2 3 4
6.62 16.78 9.45 14.04
21.98 33.58 39.48 50.79
5.45 7.27 8.36 4.73
aCalculated as number of leaves with epiphylls/total number of coffee leaves. Analysis of variance revealed that values for May and February are significantly different at the O.005 level (F of 10.39, df = I. 64); and sites are significantly different at the 0.05 level (F of 3.15, df = 3.64). bCalculated as number of leaves with activity/total number of leaves assayed from February 1978 through February 1979. Analysis of variance revealed no significant differences with respect to site and percent active samples.
T a b l e 4. N 2 fixed by coffee epiphylls ha - 1 y e a r - 1 for the 4 study sites Site
N 2 fixed ha -1 y e a r - l (g)
I 2 3 4
0.69 a 1.29 1.40 1.14
aValues were calculated as follows: g N 2 fixed h a - I y e a r - 1 = where: L P
F R T D
12 ~ i=I
"LPFRTD
= average number of coffee leaves tree - I s i t e - 1. = percent of leaves with epiphylls in each month determined by extrapolating between May 1978 and February 1980 percent epiphyll data from Table 3. = percent of leaves with epiphylls that reduced C2H 2 each month from Fig. lB, = average C2H 2 reduction rate, 3.21 nmoles l e a f - 1 d a y - I t3 x 28. = number of coffee trees ha - l ; 1600. = n u m b e r o f d a y s in month i.
of increased leakage of nutrients from the old leaves (5, 7), which stimulates nitrogenfixing activity. Nitrogen fLxed by coffee epiphylls in the four sites ranged from 0.7 g N2 ha- 1 year- l in the shadeless site, site 1, to 1.4 g N2 ha -1 year -1 in the site with Ingajinicuil shade trees, site 3 (Table 4). Interestingly, this leguminous tree is nodulated and fixes approximately 35 kg N 2 ha -1 year -1 (Roskoski, Plant and Soil, in press). However,
Nitrogen Fixation by Epiphylls
355
whether there is a relation between epiphyllic fLxation in this site and the presence of this particular shade tree is unknown. Since rates of C2H2 reduction and percent epiphyll leaves with activity were similar for all sites (see prior discussion), differences in the amounts of nitrogen fixed per site are due to differences in the percent leaves with epiphylls per site. The values presented above are considerably less than those reported by Jones (5) and Sucoff (13) for temperate forests. However, the latter were estimated maximizing all parameters. A more realistic estimate would probably approximate my values. The low values for coffee epiphyll fixation are somewhat surprising. Given the optimal conditions for epiphyll growth and development existent in a tropical coffee plantation, i.e., high humidity and warm temperatures, considerably higher annual fixation might be expected. However, Ruinen (12) postulated that maximum epiphyll fixation occurs when soil nitrogen levels are low, which promotes high carbohydrate exudation by the leaves. All four study sites are annually fertilized with about 86 kg nitrogen ha -l. Consequently, soil nitrogen levels are high, 50 ppm NH4 + and 0.5% total Kjeldahl nitrogen (Jimenez, personal communication). These levels of soil nitrogen may not favor carbohydrate exudation by the leaves, and consequently result in a leaf micro-environment less favorable for epiphyilic nitrogen fixation. Thus the amount of nitrogen fixed by coffee epiphylls is not an important source of nitrogen for the coffee ecosystem.
Acknowledgments. I thank Jorge Calderon and Guillermo Castilleja for their help with field and laboratory work. Financial support was provided by a Rockefeller Post-doctoral Fellowship in Environmental Affairs and by the Instituto National de Investigaciones sobre Recursos Bioticos.
References 1. Hardy, R. F. W., R. D. Holsten, E. K. Jackson, and R. C. Bums: The acetylene-ethylene assay for nitrogen fixation: laboratory and field evaluation. Plant Physiol. 43, 1185-1207 (1968) 2. Jimenez, E.: Ecological study of the coffee agro-ecosystem: I. Structure of a coffee plantation in Coatepec, Vet., Mexico. Biotica 4, 1-12 (1979) 3. Jimenez, E., and P. Martinez: Ecological studies of the coffee agro-ecosystem, lI. Organic matter production in different types of coffee plantations. Biotica 4, 109-126 ( 1979) 4. Jones, K.: Nitrogen fixation in the phyllosphere of Douglass fir, Pseudotsuga duglasii. Ann. Bot. 34, 239-244 (1970) 5. Jones, K.: Nitrogen fixing bacteria in the canopy of conifers in a temperate forest. In C. H. Dickinson and T. F. Preece (eds.): Microbiology of Aerial Plant Structures, pp. 41-66. Academic Press, New York. 6 . Jones, K., E. King, and M. Easflick: Nitrogen fixation by free-living bacteria in the soil and in the canopy of Douglass fir. Ann. Bot. 38,765-772 (1974) 7. Last, F. T., and F. C. Deighton: The non-parasitic micro-flora on the surfaces of living leaves. Trans. Br. Mycol. See. 48, 83-99 (1965) 8. McColl, J. G.: Properties of some natural waters in a tropical wet forest of Costa Rica. Bioscience 20, 1096-1100 (1970) 9. Reynolds, A. G.: Stratification of tropical epiphylls. Kalilasan Philippine J. Biol. 1, 7-10 (1972) 10. Ruinen, J.: The phyllosphere. 1. An ecologically neglected milieu. Plant Soil 15,81-109 (1961) 11. Ruinen, J.: The phyllosphere. II1. Nitrogen fixation in the phyllosphere. Plant Soil 22,375-394 (1965) 12. Ruinen, J.: Nitrogen fixation in the phyllosphere. InA. Quispel (ed.): The Biology of Nitrogen Fixation, pp. 86-121. North-Holland Publishing Co., Amsterdam (1974) 13. Sacoff, E.: Estimate of nitrogen fixation on leaf surfaces of forest species in Minnesota and Oregon. Can. J. Forest Res. 9, 474-477 (1979)
N 2 Fixation
DO~I:IL
CCOI.~
(C2H 2 Reduction) by Epiphylls on Coffee, Coffea arabica
Joann P. Roskoski InstitutoNacionalde lnvestigacionessobreRecursosBioticos,ApartadoPostal63, Xalapa,Veracruz,Mexico Abstract. Nitrogen fixation (C2H2 reduction) by epiphylls on coffee, Coffea arabica, grown in sites with different degrees of shade, was determined. Coffee leaves with nitrogen-fixing epiphylls were found in all sites in approximately equal numbers. Rates of C2H2 reduction were similar for all sites and throughout the year, averaging 3.21 nmoles C2H 2 reduced leaf with epiphylls-I d a y - I . Apparently, neither rates of activity nor abundance of leaves with nitrogen-fixing epiphylls is related to the degree of shade in a site. No correlation was found between percent epiphyll cover and the presence or magnitude of nitrogen-fixing activity. Calculated annual fixation by epiphylls on coffee was low, ranging from 0.7 g N2 ha -l year -I for the shadeless site to 1.4 g N2 ha -I year -~ for the site with lngafinicuil shade trees. These results suggest that epiphyll fixation is not an important source of nitrogen for the coffee ecosystem studied.
In 1965 Ruinen (11) observed that the total nitrogen content of petri dishes, containing nutrient media and coffee leaves with epiphyllic organisms, increased through time. The nitrogen increase was ascribed to nitrogen fixation by the leafepiphylls. Later, nitrogenfixing bacteria were isolated from leaves of I0 tree species in England (4-6), and from foliage of various tropical plants (12). Nitrogen fixation by leaf epiphylls has been suggested as an important source of nitrogen in some ecosystems (5, 12) but not in others (13). For example, Jones (5) calculated that up to 15 kg N2 h a - l year- 1 could be added to an English coniferous forest via epiphyll fixation, whereas Sucoff (13) found that the annual nitrogen input from epiphyll fixation in coniferous and deciduous forests of the United States was only between 0.2 and 0.06 kg ha -1 year -I . Maximum epiphyll development probably occurs in the tropics where leaf longevity, atmospheric moisture levels, and ambient temperatures favor abundant growth of a diverse epiphyll community (10). However, to date, no seasonal quantitative studies on epiphyllic fixation in tropical ecosystems have been reported. Previous evidence suggested that epiphyI1 fixation occurred on coffee leaves (11). Therefore, during an ecological study of a coffee agro-ecosystem, the question of epiphyll fixation was examined. The study focused on understory vegetation in order to concentrate on the strata where epiphylls were presumably most abundant (9). This 0095-3628180/0006-0349 $01.40 9 1980 Springer-VerlagNew Yorklne.
35I)
I . P . Roskoski
Table 1. Shade characteristics of the four study sites
Site
Shade species None
lnga vera H.B+ & K. Inga jinicuil SehL Musa sapientum L, Citrus sinensis Osb. tnga vera lnga jinicuil
Cover by shade speciesa (m 2 h a - 1) 0 8,237 12,792 40,374
aCover data from Jimenez (2).
paper presents the results of an investigation to quantify annual nitrogen fixation by epiphylls on coffee plants grown in sites with different degrees of shade.
Materials and Methods Site Description The study was conducted in a coffee plantation near the city o f Xalapa, Veracruz, Mexico, at 19~ ' latitude north, 96~ , longitude west, and an attitude of 1225 m above sea level. The climate is classified as semi-hot, humid, with warm summers and coot winters (2). Annual rr~an temperature is 19~ • 2~ and precipitation averages 1758 • 193 mm annually+ Four sites within the plantation, each differing primarily in the type and degree of shade present, were located, Table t presents the shade characteristics of the sites. Coffee cover and density were similar in all four areas.
Nitrogen Fixation Armfyses Once each month, from February I978 through February [979, twenty coffee leaves with visible epiphyIlie growth, two from each of 10 m ~ , were randomly collected from each site. Each leaf was it~m~..diatety placed in a separate 20 ml test tub~. Tubes were stoppered and incubated at 20~C tbr 24 hours under ambient laboratory, light levels. It was felt that these ie,cubation conditions approximated those experienced by epiphylls in the coffee plantation. After 24 hours a 0.2 cc gas sample was withdrawn from each tube and injected into a Curie model 9500 l i d gas chromatograph, equipped with two 1.8 m columns, packed with Poropak N+ Separations were run at 60~ using nitrogen as a carder gas. This procedure measured C2H 4 production by leaves with epiphylls in the absence of C2H2+ Typical C2H 4 production was zero. Following the above analysis, each test tube was opened to allow the re-establishment of ambient atmospheric conditions in the assay tube. Tubes were then restoppered, and C2H2, equal to 10% of the test tube volume, was introduced. A 0.20 cc gas sample, withdrawn after 24 hours, was analyzed for C2H4, as descrit~.,d above. Reagent controls were routinely ran. Peak heights for C2H 4 produced by leaves with epiphylls in the presence of C2H 2, corrected for the reagent controls, and C2H 4 production by the same leaves in the absence of C2H 2, were compared to peak heights of known C2H 4 concentrations. Moles C2H 4 produced were convened to moles N 2 fixed, employing the theoretical 3:1 ratio ( 1 )+ After C2H 2 reduction analysis, a clear plastic sheet, gridded into 25 ram 2 sections, was placed over each leaf and used to d~termine total leaf area and area occupied by epiphylls. The latter was defined as any area with
Nitrogen Fixation by Epiphylls
351
altered surface texture and/or coloration, excluding diseased areas. Thus epiphyll cover includes only visible epiphylls on leaf surfaces.
Test of the C2H 2 Reduction Method The C2H 2 reduction method, described above, assumed that C2H4 production during the fLrst 24-hour incubation period, without C2H2, represented natural C2H4 production during the second 24-hour incubation. To test the validity of this assumption, 30 coffee leaves with epiphylls were collected in January 1978, prior to the initiation of the annual study. These leaves were placed in test tubes, which were then stoppered. C2H 4 production was determined after 24 hours. Tubes were then opened, re-closed after a few minutes, and C2H 4 content measured after another 24 hours. This procedure was repeated for a third day, for a total of three l-day incubations. The entire test was repeated in March 1979. Results revealed no significant differences in background C2H 4 production among days 1,2, and 3.
Epiphyll Abundance To ascertain the number of leaves with epiphylls per site, a 0.2 ha plot in each site was subdivided into sixty-four 25 m 2 subplots. In May 1978, three subplots in each site, each containing four coffee trees, were randomly selected. Line transects were run out from the base of each tree at 900 intervals. Plastic screening, extended vertically from these lines, divided each tree into four equal sections. One section of each tree was randomly chosen, and the total number of leaves and leaves with visible epiphylls Were counted. Values for each tree section were averaged by site, multiplied by four, to yield average tree totals, and then multiplied by the average number of coffee trees ha- 1. To determine if the number of coffee leaves with epiphylls, obtained by the above method, was constant, this phase of the study was repeated in February 1980, but only two coffee trees per subplot were recounted.
Results and Discussion N i t r o g e n - f i x i n g epiphylls occurred in all sites. A l t h o u g h no attempt was m a d e to identify s p e c i f i c o r g a n i s m s responsible for m e a s u r e d activity, foliocolous lichens, p i g m e n t e d bacterial colonies, and b l u e - g r e e n algae w e r e periodically o b s e r v e d on assayed leaves. Rates o f C2H2 reduction ranged from 0 . 4 to 27.3 nmoles leaf with e p i p h y i l s - l d a y - I , w h i c h equals 0.1 to 9.1 n m o l e s N 2 fixed leaf with epiphylls - I day - I . T h e s e rates, a l t h o u g h s o m e w h a t higher than those reported by S u c o f f (13) for temperate vegetation, are l o w e r than those for other tropical plants (12). A n a l y s i s o f v a r i a n c e r e v e a l e d no significant differences between sites with respect to rate o f C 2 H 2 reduction. Apparently, absence, presence, or degree o f shade does not affect the m i c r o - e n v i r o n m e n t o f nitrogen-fixing epiphylls in a m a n n e r that causes rate differences. M e a n m o n t h l y rates, on the other hand, were statistically different (Table 2). H o w e v e r , if the single values for A u g u s t and S e p t e m b e r , which are considerably higher than rates for the other months, are r e m o v e d from the analysis, the remaining rates are similar. T h i s finding implies that the m i c r o - e n v i r o n m e n t a l factors controlling nitrogenfixing activity on leaf surfaces are remarkably constant throughout the year. In c o m p a r i s o n , nitrogen fixation rates in temperate e c o s y s t e m s are markedly seasonal (5). P o o l e d m o n t h l y data resulted in an average rate o f 3.21 nmoles C2H 2 reduced leaf with activity - l d a y - 1. T h e percent o f leaves with epiphylls that reduced C2H 2 did not differ significantly
352
J. P. Roskoski
Table 2. C2H 2 reduction rates for coffee leaves with epiphyUs
Month
Active samples
January February March April May July August September October
9a 26 10 3 7 1 1 1 2
November December
3 8
C2H 2 reduced l e a f - 1 day - I (nmoles) 1.36 3.04 4.33 1.28 3.21 1.73 16,40 27.30 0.79
___0.30 b + 0.72 + 2.03 -I- 0.19 + 1.12
+ 0.00
1.15 + 0.30 2.05 + 0.47
aTotal number of samples assayed in each month was 80, with the exception of data for February, which include data for 1978 and 1979, or 160 assays performed. bValues shown are the mean + S.E. of all active samples.
between sites (Table 3). This result is somewhat surprising since significant site differences were found with respect to the percent of leaves that possessed epiphylls (Table 3). From these data it appears that degree of shading or other site characteristics influence the abundance of leaves with epiphylls but not the fraction of those leaves that have nitrogen-fixing epiphylls. No correlation was found between the areal extent of the epiphyll community on any leaf and the presence or magnitude of nitrogen-fixing activity that could be expected for that same leaf. Leaves with 0.5 to 100% epiphyll cover exhibited Similar nitrogen-fixing activity. Nitrogen-f~xing epiphylls are apparently ubiquitous throughout the development of the epiphyll community. Average percent epiphyll cover on leaves was similar for all sites but varied throughout the year (Fig. 1A). In addition, the percent of leaves with epiphylls varied with time (Table 3). These annual changes largely reflect the phenology of coffee. Coffee leaves begin to fall in February. Total replacement occurs by May, shortly before flowering (3). Epiphyll cover per leaf and number of leaves with epiphyUs are lowest in May because the majority of coffee leaves present at this time are young and have not yet been colonized. Cover gradually increases as young leaves are colonized and epiphyll growth ensues. Maximum percent cover and percent leaves with epiphylls in February is due to the large number of mature leaves present at this time. As leaves begin to fall, percent cover decreases since new leaves form a larger and larger portion of the leaf population. In comparison to the pronounced seasonal pattern described above, the percent coffee leaves with nitrogen-fixing epiphylls shows less seasonal variation (Fig. 1B). Highest percent leaves with nitrogen-fixing epiphylls occurs in February coincident with maximum percent cover and percent leaves with epiphylls. However, lowest percent leaves with nitrogen-fixing epiphylls was found not in May but slightly later, when percent cover was increasing most rapidly (Fig. 1). During June, July, and August,
Nitrogen Fixation by Epiphylls
353
4.5-
"T b3 _1
Q:
hi ::P
30
o ..i .J )-lQ_ tlJ
15
A
20 O') W W .d X
jE
,J )-
I0 I,-
1,1
.-e
o
T ,u,,,,,=
i""""
i 2"
1978
! ~,ii, ~w~ v:
l
1979 TIME
Fig. I. A % Epiphyll cover leaf- I vs time. Eighty coffee leaves with epiphylls. 20 from each study site, were collected monthly and the percent epiphyll cover detemfined, Values shown are the mean :t: S.E~ of these 80 values. B % Epiphyll leaves that fix N 2. For each site the number of epiphylt leaves that reduced C2H 2 was divided by the number of leaves assayed. Values shown are the mean +__S.E. of data for all sites,
heavy summer rains, presumably containing nitrogenous compounds (8), may supply sufficient nitrogen to the phylloplane to reduce the need for nitrogen fixation. In addition, summer rains may dilute the concentration of carbohydrates on the leaf's surface (12), which may be the primary energy source for nitrogen-fixing activity. The increase in the percent nitrogen-fixing epiphylls at the end of the year may be a function
354
J.P. Roskoski
Table 3. Site values for percent leaves with epiphylis in May 1978 and February 1980, and percent leaves with epiphylls that reduced C2H 2 Leaves with epiphylls (%)a Site
May 1978
Febrhary 1980
Leaves with epiphylls that reduced C2H 2 (%)b
I 2 3 4
6.62 16.78 9.45 14.04
21.98 33.58 39.48 50.79
5.45 7.27 8.36 4.73
aCalculated as number of leaves with epiphylls/total number of coffee leaves. Analysis of variance revealed that values for May and February are significantly different at the O.005 level (F of 10.39, df = I. 64); and sites are significantly different at the 0.05 level (F of 3.15, df = 3.64). bCalculated as number of leaves with activity/total number of leaves assayed from February 1978 through February 1979. Analysis of variance revealed no significant differences with respect to site and percent active samples.
T a b l e 4. N 2 fixed by coffee epiphylls ha - 1 y e a r - 1 for the 4 study sites Site
N 2 fixed ha -1 y e a r - l (g)
I 2 3 4
0.69 a 1.29 1.40 1.14
aValues were calculated as follows: g N 2 fixed h a - I y e a r - 1 = where: L P
F R T D
12 ~ i=I
"LPFRTD
= average number of coffee leaves tree - I s i t e - 1. = percent of leaves with epiphylls in each month determined by extrapolating between May 1978 and February 1980 percent epiphyll data from Table 3. = percent of leaves with epiphylls that reduced C2H 2 each month from Fig. lB, = average C2H 2 reduction rate, 3.21 nmoles l e a f - 1 d a y - I t3 x 28. = number of coffee trees ha - l ; 1600. = n u m b e r o f d a y s in month i.
of increased leakage of nutrients from the old leaves (5, 7), which stimulates nitrogenfixing activity. Nitrogen fLxed by coffee epiphylls in the four sites ranged from 0.7 g N2 ha- 1 year- l in the shadeless site, site 1, to 1.4 g N2 ha -1 year -1 in the site with Ingajinicuil shade trees, site 3 (Table 4). Interestingly, this leguminous tree is nodulated and fixes approximately 35 kg N 2 ha -1 year -1 (Roskoski, Plant and Soil, in press). However,
Nitrogen Fixation by Epiphylls
355
whether there is a relation between epiphyllic fLxation in this site and the presence of this particular shade tree is unknown. Since rates of C2H2 reduction and percent epiphyll leaves with activity were similar for all sites (see prior discussion), differences in the amounts of nitrogen fixed per site are due to differences in the percent leaves with epiphylls per site. The values presented above are considerably less than those reported by Jones (5) and Sucoff (13) for temperate forests. However, the latter were estimated maximizing all parameters. A more realistic estimate would probably approximate my values. The low values for coffee epiphyll fixation are somewhat surprising. Given the optimal conditions for epiphyll growth and development existent in a tropical coffee plantation, i.e., high humidity and warm temperatures, considerably higher annual fixation might be expected. However, Ruinen (12) postulated that maximum epiphyll fixation occurs when soil nitrogen levels are low, which promotes high carbohydrate exudation by the leaves. All four study sites are annually fertilized with about 86 kg nitrogen ha -l. Consequently, soil nitrogen levels are high, 50 ppm NH4 + and 0.5% total Kjeldahl nitrogen (Jimenez, personal communication). These levels of soil nitrogen may not favor carbohydrate exudation by the leaves, and consequently result in a leaf micro-environment less favorable for epiphyilic nitrogen fixation. Thus the amount of nitrogen fixed by coffee epiphylls is not an important source of nitrogen for the coffee ecosystem.
Acknowledgments. I thank Jorge Calderon and Guillermo Castilleja for their help with field and laboratory work. Financial support was provided by a Rockefeller Post-doctoral Fellowship in Environmental Affairs and by the Instituto National de Investigaciones sobre Recursos Bioticos.
References 1. Hardy, R. F. W., R. D. Holsten, E. K. Jackson, and R. C. Bums: The acetylene-ethylene assay for nitrogen fixation: laboratory and field evaluation. Plant Physiol. 43, 1185-1207 (1968) 2. Jimenez, E.: Ecological study of the coffee agro-ecosystem: I. Structure of a coffee plantation in Coatepec, Vet., Mexico. Biotica 4, 1-12 (1979) 3. Jimenez, E., and P. Martinez: Ecological studies of the coffee agro-ecosystem, lI. Organic matter production in different types of coffee plantations. Biotica 4, 109-126 ( 1979) 4. Jones, K.: Nitrogen fixation in the phyllosphere of Douglass fir, Pseudotsuga duglasii. Ann. Bot. 34, 239-244 (1970) 5. Jones, K.: Nitrogen fixing bacteria in the canopy of conifers in a temperate forest. In C. H. Dickinson and T. F. Preece (eds.): Microbiology of Aerial Plant Structures, pp. 41-66. Academic Press, New York. 6 . Jones, K., E. King, and M. Easflick: Nitrogen fixation by free-living bacteria in the soil and in the canopy of Douglass fir. Ann. Bot. 38,765-772 (1974) 7. Last, F. T., and F. C. Deighton: The non-parasitic micro-flora on the surfaces of living leaves. Trans. Br. Mycol. See. 48, 83-99 (1965) 8. McColl, J. G.: Properties of some natural waters in a tropical wet forest of Costa Rica. Bioscience 20, 1096-1100 (1970) 9. Reynolds, A. G.: Stratification of tropical epiphylls. Kalilasan Philippine J. Biol. 1, 7-10 (1972) 10. Ruinen, J.: The phyllosphere. 1. An ecologically neglected milieu. Plant Soil 15,81-109 (1961) 11. Ruinen, J.: The phyllosphere. II1. Nitrogen fixation in the phyllosphere. Plant Soil 22,375-394 (1965) 12. Ruinen, J.: Nitrogen fixation in the phyllosphere. InA. Quispel (ed.): The Biology of Nitrogen Fixation, pp. 86-121. North-Holland Publishing Co., Amsterdam (1974) 13. Sacoff, E.: Estimate of nitrogen fixation on leaf surfaces of forest species in Minnesota and Oregon. Can. J. Forest Res. 9, 474-477 (1979)
