Plant and Soil 142: 63-67, 1992. © 1992 Kluwer Academic Publishers. Printed in the Netherlands.
PLSO 9319
The occurrence of Frankia in tropical forest soils of Costa Rica MARK W. PASCHKE and JEFFREY O. DAWSON Department of Forestry, University of Illinois, Urbana, IL 61801, USA Received 20 August 1991. Revised December 1991
Key words: actinorhizal plants, Costa Rica, dinitrogen-fixation, Frankia, root nodules, symbiosis, tropical soils Abstract
Infective and effective Frankia were shown to occur in five diverse tropical forest soils of Costa Rica. Results of a plant infection assay indicated that Frankia is a common component of the soil biota in low and high elevation, primary and secondary forest soils. This is the first report of Frankia in lowland tropical rainforests of the Americas. These results suggest either a nonsymbiotic population of soil Frankia, the presence of unknown actinorhizal host species, or an ability of Frankia to be dispersed over long distances.
Introduction
Actinorhizal plants are rare in the neotropics with the exception of a few species of Myrica and Alnus which occur in Central and South American highlands. There are no known actinorhizal plants in neotropical lowlands where woody legumes seem to fill the roles that actinorhizal plants play in the nitrogen economies of cool temperate plant communities. Compared to the rhizobial symbionts of leguminous plants, relatively little is known regarding the biology of Frankia in soil. There are several reports of Frankia in northern hemisphere temperate soils lacking actinorhizal host plants (Dawson and Klemp, 1987; Huss-Danell and Frej, 1986; Rodriguez-Barrueco, 1968; Zitzer et al., 1991). Wollum et al. (1968) found that the population of infective Frankia in soils decreased proportionally with the amount of time that actinorhizal Ceanothus host plants had been eliminated from a site, indicating a slow decline of infective Frankia without the host plant. Frankia is usually abundant in soils beneath nodulated host plants (Arveby and HussDanell, 1988; Oremus, 1980; Smolander, 1990;
Van Dijk, 1979), probably due to the liberation of Frankia from decaying root nodules (Van Dijk, 1984). Many factors are believed to affect Frankia abundance in soils. The most commonly reported are pH (Griffiths and McCormick, 1984; Smolander, 1990; Smolander et al., 1988; Smolander and Sundman, 1987) and soil moisture (Dawson et al., 1989; Righetti et al., 1986). While these properties seem to influence Frankia abundance, Frankia can be present in soils that severely limit its infective capacity. Free living Frankia in soil may be saprophytic and, if so, the nature and quality of soil organic matter may regulate soil populations. Relationships between vegetation and Frankia abundance have been observed (Smolander, 1990) and Frankia characteristics in host plant (Diem et al., 1982; Lalonde et al., 1981; Vergnaud et ai., 1985) and nonhost plant (Smolander et al., 1990) rhizospheres indicate that it is most likely stimulated in the presence of some plant roots. If Frankia is a soil saprophyte or rhizosphere associate then we would expect it to occur in soils lacking nodulated host plants. The purpose of this study was to examine the surface soils of
64
Paschke and Dawson
diverse neotropical plant communities lacking known actinorhizal host species for the presence of Frankia. The presence of Frankia in neotropical soils beneath flora lacking symbiotic host plants would suggest that Frankia may be a common cosmopolitan soil or rhizosphere bacterium with an ability to compete with diverse soil biota. Alternatively, the presence of Frankia in some tropical plant communities may be indicative of yet unknown actinorhizal host species, or highly effective dispersal and, perhaps, dormancy capabilities.
Materials and methods
Five study sites with diverse plant formations in Costa Rica were chosen. These sites are described in Table 1, and are classified according to the Holdridge life zone classification system (Holdridge et al., 1971) as mapped in Costa Rica by Tosi (1969). No actinorhizal host plants were found at any of the sampling sites. There are no known actinorhizal host species at or close to the lowland wet forest and dry forest sites. Actinorhizal Alnus acuminata H.B.K. occurs in pastures at lower elevations in the vicinity of the montane rain forest sampling site. Myrica phanerodonta Standl., which is presumed to be actinorhizal, is reported to occur at slightly higher elevations (about 300 m) than the montane rain forest sampiing site, although an exhaustive search of the
area yielded no specimens. Exotic Myrica cerifera L. is often found in road cuts in cloud forest regions of Costa Rica, and may have been present near the two premontane sites and the montane site although it was not observed in neighboring roadcuts that were examined. Exotic actinorhizal Casuarina equisetifolia L. is planted for windbreaks in highland areas of Costa Rica and nodulated individuals of this species were observed in pastures at lower elevations near the montane rain forest site. At each site a 5 km 2 area of continuous forest cover was chosen. Within this area, at five randomly selected locations, a 2-cm-wide × 10-cmdeep cylindrical soil core was removed with a steel soil probe. The soil probe was flame sterilized prior to each sample collection. Samples were placed immediately into sterile whirl-pak bags (NASCO, Inc.) and sealed. Samples were air-dried beginning the day of collection by opening the bags and sealing each in an individual paper bag for 3 to 4 days. Once airdried, the samples were resealed in the whirl-pak bags and stored at room temperature. In order to detect the presence of Frankia in the soil samples, a plant infection technique was employed. The soil cores were homogenized and a 20-cm 3 portion of each sample was diluted to 100 mL in sterile, deionized H20. The soil slurry was shaken for 1 h on a mechanical shaker and then allowed to settle for 60s. Twelve 5-mL aliquots were then removed and used as inocula directly on test plants. All soil sample handling
Table 1. Characteristics of the sites where soil samples were collected
Lifezone Tropical wet forest Tropical dry forest Premontane rain forest Montane rain forest Premontane wet forest, rain forest transition
Forest type primary primary secondary secondary primary
Location La Selva" 10°26'N, 83°59'W Palo Verde a 10°21 'N, 85°20'W Monteverde b I0°19'N, 84°47'W Cerro de la Muerte b 9°33'N, 83°43'W Las Cruces" 8°50'N, 82°53'W
Soil Suborder c
Collection date 1990
Aquept
June 22
Tropept
June 30
Andept
July 7
Tropept
July 24
Andept
August 2
a Biological reserves operated as research stations by the Organization for Tropical Studies (San Pedro, Costa Rica). b Reserves established by the Costa Rican government. c Soil taxonomy follows the USDA 7th Approximation and is based on Morera (1983).
Frankia in neotropical forest soils and inoculation was carried out under aseptic conditions in order to avoid the introduction of exogenous Frankia. One soil sample from the tropical wet forest site was lost, as a result, the data reflect only 4 soil cores for this site. The test plants were 4-month-old seedlings of Elaeagnus umbellata Thunb., and 5-month-old seedlings of Alnus glutinosa (L.) Gaertn. Seeds of the test plants were surface sterilized in 30% (v/v) H 2 0 2 and sown in 4-cm-wide x 14-cm-deep sterile plastic tubes. A 1:1:1 mixture of fine vermiculite, mixed sand, and fine gravel was used as the growth medium and this mixture was steam pasteurized twice prior to seeding. Plants were grown in a glasshouse isolation r o o m at 23-+ 5°C. Photoperiod was extended to 16 h using 1000-watt Sylvania LU1000 ( G T E Sylvania Inc.; Manchester, N H ) lamps. On alternate days, plants were watered with a complete 1/8-strength. nutrient solution (Huss-Danell, 1978). Two weeks prior to inoculation with soil suspension the nitrogen source was removed from the nutrient solution. The air supply to this isolation r o o m was filtered through Air Guard type dp 2-40 (Air G u a r d Industries Inc.; Louisville, K Y ) filters and the floor was cleaned every 48 h with 1% (v/v) NaCIO. For each species, six tubes containing plants were inoculated with 5 m L of the soil suspension. At the time of inoculation there were between 5 and 15 Alnus seedlings per tube, and 3 to 15 Elaeagnus. Although these tubes contained varying numbers of seedlings, the tubes with fewer seedlings contained larger individuals and the roots had fully exploited the soil in the containers resulting in a uniform and dense root mass in all the tubes at the time of inoculation. The inoculated tubes were arranged randomly on the glasshouse bench. Another 42 tubes of Alnus and 52 tubes of Elaeagnus seedlings were mixed in with the inoculated tubes and served as contamination detection plants to monitor exogenous Frankia and cross contamination. Fourteen weeks after inoculation, the number of nodules formed in each tube of test plants was determined. Formation of root nodules was interpreted as an indication of Frankia presence. Nodulated plants were larger and had greener leaves c o m p a r e d with non-nodulated plants. The effectiveness of the symbiosis was determined
65
prior to harvest by an acetylene-reduction assay on a subsample of tubes containing the larger plants with dark green leaves. One or two tubes containing nodulated plants representing each combination of host species and site were randomly selected. In addition, a random subsample of corresponding non-nodulated plants were assayed for acetylene reduction. Intact tubes of plants were incubated in a 10% atmosphere of acetylene for 1 h. Ethylene production was quantified using a Hewlett-Packard 5890 gas chromatograph equipped with a flame ionization detector (Hewlett-Packard Co. Avondale, Pennsylvania).
Results
Frankia presence was detected at all five of the sites (Fig. 1). Soil from the premontane rain forest did not form nodules on any of the Alnus 40
T] II
Alnus glutinosa
[ ] Trol~ClJwet for,ill m.,o,,~..1 ....
30, [] PJ)%~1"n/or ....
E
20,
-~
10.
J~ t--
o
EL
e
2n i,l]l
r-
2341234512345123451234
120 .
o "6
100
80 6O 4O 2O 0
12
__,fil I
U
45123451
e
t
4
Soil Samples
Fig. 1. Mean number of nodules produced per tube of Elaeagnus or Alnus test plants R)r each soil sample collected at the five study sites. Each site is represented by different shading of the thick bars. Thin bars represent the standard error of the mean (n = 6). Numbers below the x-axis refer to the soil core sample numbers at each site.
66
P a s c h k e and D a w s o n
seedlings in the experiment. However, a few nodules were detected in one of six tubes of Elaeagnus from one sample at this site (Fig. 1). Levels of nodulation were greater at the other four sites for both host species. In general, nodules were more numerous and also smaller on the Elaeagnus than on the Alnus test plants, similar to findings in Illinois soils (Zitzer and Dawson, 1992). Most soil samples that produced nodules on test plants did so on both host species. No nodules were found in any of the 94 tubes of uninoculated plants. All tubes of nodulated plants tested were effective at reducing acetylene to ethylene. Rates of ethylene production were between 0.57 and 2 . 4 5 / z m o l t u b e - ~ h -1 for Elaeagnus test plants and 2.68 to 7.15 ~ m o l tube -~ h -1 for Alnus test plants. Nonnodulated plants produced between 0.07 and 0 . 1 1 / z m o l of ethylene per tube per hour. The effectiveness of the symbiosis was also apparent from the dark green foliage and large size (about 20 cm in height for both species) of nodulated plants compared to small (about 5 cm in height for both species), yellowish non-nodulated plants.
Discussion These results indicate the presence of infective and effective Frankia in some neotropical forest soils. Perhaps the most notable of these results is the detection of Frankia in primary tropical wet and dry forests. Neither of these plant assemblages are known to contain actinorhizal species, and there are no known host species in the vicinity of these two sites. This is the first report of F r a n k i a presence in lowland tropical forests of the Americas. The presence of Frankia in the highland premontane rain forest and the wet forest transition zones is also an important finding since there were no known actinorhizal species at the sampling sites. The detection of Frankia in these soils suggests that Frankia may be a c o m m o n soil bacterium, even in neotropical regions lacking host plants. Alternatively, these results may indicate the presence of unknown actinorhizal host species, which is plausible given the paucity of root
studies coupled with the immense floral diversity of these Costa Rican forests (Hartshorn, 1983). Long distance dispersal of Frankia spores is another possible reason for the presence of Frankia in these soils. We cannot rule out any of these possibilities, and it is apparent that our understanding of the ecology of this important plant symbiont in the neotropics is rudimentary.
Acknowledgements This work was conducted under the auspices of the Organization for Tropical Studies in Costa Rica. The assistance of Dr Richard F Fisher with site selection and sample collection is greatly appreciated.
References Arveby A S and Huss-Danell K 1988 Presence and dispersal of infective Frankia in peat and meadow soils in Sweden. Biol. Fertil. Soils 6, 39-44. Dawson J O and Klemp M T 1987 Variation in the capacity of black alder to nodulate in central Illinois soils. In Sixth Central Hardwood Forest Conference. Eds. R L Hay, F W Woods and H De Selm. pp 255-260. Dept. For., Wildlife and Fisheries, University of Tennessee, Knoxville. Dawson J O, Kowalski D G and Dart P J 1989 Variation with soil depth, topographic position and host species in the capacity of soils from an Australian locale to nodulate Casuarina and Allocasuarina seedlings. Plant and Soil 118, 1-13. Diem H G, Gauthier D and Dommergues Y R 1982 Extranodular growth of Frankia of Casuarina equisetifolia. FEMS Microbiology Letters 15, 181-184. Griffiths A P and McCormickL H 1984 Effects of soil acidity on nodulation of Alnus glutinosa and viability of Frankia. Plant and Soil 79, 429-434. Hartshorn G S 1983 Plants In Costa Rican natural history. Ed. D H Janzen. pp 118-350. University Chicago Press, Chicago. Holdridge L R, Grenke W C, Hatheway W H, Liang T and Tosi J A Jr. 1971 Forest Environments in Tropical Life Zones: A Pilot Study. Pergamon Press, New York. 747 p. Huss-Danell K 1978 Nitrogenase activity measurements in intact plants of Alnus incana. Physiol. Plant. 43, 372-376. Huss-Danell K and Frej A 1986 Distribution of Frankia in soils from forest and afforestation sites in northern Sweden. Plant and Soil 90, 407-418. Lalonde M, Calvert H E and Pine S 1981 Isolation and use of Frankia strains in actinorhizae formation. In Current Perspectives in Nitrogen Fixation. Eds. A H Gibson and W E Newton. pp 295-299. Elsevier/North Holland Biomedical Press, New York.
F r a n k i a in neotropical ]brest soils Morera A V 1983 Soils In Costa Rican Natural History. Ed. D H Janzen. pp 63-65. University Chicago Press, Chicago. Oremus P A I 1980 Occurrence and infective potential of the endophyte of Hippophae rhamnoides L. ssp. rhamnoides in coastal sand-dune areas. Plant and Soil 56, 123-139. Righetti T L, Chard C H and Backhaus B A 1986 Soil and environmental factors related to nodulation in Cowania and Purshia. Plant and Soil 91, 14%160. Rodriguez-Barrueco C 1968 The occurrence of the rootnodule endophyte of Alnus glutinosa and Myrica gale in soils. J. gen. Microbiol. 52, 189-194. Smolander A 1990 Frankia populations in soils under different tree species-with special emphasis on soils under Betula pendula. Plant and Soil 121, 1-10. Smolander A and Sundman V 1987 Frankia in acid soils of forests devoid of actinorhizal plants. Physiol. Plant. 70, 297-3(13. Smolander A, Ronkko R, Nurmiaho-Lassila E-L and Haahtcla K 1990 Growth of Frankia in the rhizosphere of Betula pendula, a non-host tree species. Can. J. Bot. 36, 649-656. Smolander A, Van Dijk C and Sundman V 1988 Survival of Frankia strains introduced into soils. Plant and Soil 106. 65-72. Tosi J A Jr 1969 Mapa Ecologico, Republica de Costa Rica:
67
Segun la Classificacion de Zonas de Vida del Mundo de L.R. Holdridge. Centro Cientifico Tropical, San Jose, Costa Rica. Van Dijk C 1979 Endophyte distribution in the soil. In Symbiotic Nitrogen Fixation in the Management of Temperate Forests, Eds. J C Gordon, C T Wheeler and D A Perry. pp 84-94. Oregon State University Press, Corvallis, OR. Van Dijk C 1984 Ecological Aspects of Spore Formation in the Frankia-Alnus Symbiosis. Ph.D. Thesis. University of Leiden. Vergnaud L, Chaboud A, Prin Y and Rougier M 1985 Preinfection events in the establishment of Alnus-Frankia symbiosis: Development of a spot inoculation technique. Plant and Soil 87, 67-78. Wollum A G, Youngberg C T and Chichester F W 1968 Relation of previous timber stand age to nodulation of Ceanothus velutinus. Forest Sci. 14, 114-118. Zitzer S F and Dawson J O 1992 Soil properties and actinorhizal vegetation influence nodulation of Alnus glutinosa and Elaeagnus angustifolia by Frankia. Plant and Soil 140, 197-204. Zitzer S F, Archer S R and Boutton T W 1991 Infective capacity of symbiotic nitrogen-fixing bacteria in southern Texas savannas. Bull (Suppl.) Ecol. Soc. Am. 72, 299.