Comparison of Nitrogen Content in Tree Litterfall in Three Dry Dipterocarp Forests Under Different Fire Regime in Northeast Thailand

TROPICS Vol. 17 (3) Issued May 30, 2008

Comparison of nitrogen content in tree litterfall in three dry dipterocarp forests under different fire regime in northeast Thailand

1,＊ 2 3 1 4 Tetsuya TODA , Hiroshi TAKEDA , Naoko TOKUCHI , Seiichi OHTA , Chongrak WACHARINRAT and 4 San KAITPRANEET

1 Laboratory of Tropical Forest Resources and Environments, Division of Forest Science, Graduate School of Agriculture, Kyoto University, Kyoto 606−8502, Japan 2 Laboratory of Forest Ecology, Division of Environmental Science and Technology, Graduate School of Agriculture, Kyoto University, Kyoto 606−8502, Japan 3 Field Science Education and Research Center, Kyoto University, Kyoto 606−8502, Japan 4 Department of Silviculture, Faculty of Forestry, Kasetsart University, Bangkok 10900, Thailand ＊ Corresponding author: Tel: +81−75−753−6361, Fax: +81−75−753−6372, E-mail: [email protected]

ABSTRACT Tree litterfall was measured and of forest. In addition, transformations of N are variable nitrogen (N) return by tree litterfall was estimated and dynamic in forest ecosystems. Consequently, many in three plots, F0, F10, and F35, with different studies on nutrient cycling place special emphasis on N fire histories (protected from fire for 0, 10, and 35 dynamics. years, respectively) in dry dipterocarp forests (DDF) Fire is a major disturbance and causes substantial in northeast Thailand. Annual litterfall was 3.92, loss of N from terrestrial ecosystems, especially through 7.13, and 8.79 Mg ha−1, for F0, F10, and F35, volatilization (Gillon and Rapp, 1989). The frequency respectively. Leaf litter was the main component of fire affects the extent of N loss from the ecosystem in all the plots, ranging from 67.4 % to 77.9 %, (Marafa and Chau, 1999). In Thailand, dry deciduous peaking in the dry season. Other components dipterocarp forest (or dry dipterocarp forest, DDF) is of the litterfall had no clear seasonality. The N a major forest type and occupies over 30 % of the total concentration of the tree litterfall increased in the forest area (Royal Forest Department, 2000). The DDF rainy season and decreased in the dry season in all in northeast Thailand is a biotic forest that has been three plots. The N return due to the tree litterfall maintained by human disturbances such as forest fires. was estimated to be 33.4, 75.2, and 123.8 kg ha−1, Generally speaking, forest fires in the dry season are for F0, F10, and F35, respectively. Fire protection almost inevitable because of the extremely dry conditions increased the N return by tree litterfall, as well as that occur for a couple of months a year (around January tree litter production in DDF. to February). Local people burn forest floors of DDF to improve the yield of non-timber forest products such as Key words: Dry dipterocarp forest, Forest fire, mushrooms, tree vegetables, etc (Akaakara, 2002). In Litterfall, Nitrogen dynamics, Thailand DDF, forest fires influence the forest ecosystems through the removal of understory vegetation and the litter layer. Frequent fire often led a species-rich forest into the INTRODUCTION simple combination of fire-resistant trees and perennial In forest ecosystems, litterfall is an important flux of grass (Tsutsumi, 1989). Fires play a role in maintaining N carbon (C) and mineral nutrients into the soil decomposer limitation for plant growth in such ecosystems (Vitousek system where nutrients are mineralized. The amount of and Howarth, 1991). mineral nutrients returned by litter is a good indicator Since 1970s, Thai government has started practicing of nutrient cycling (Proctor, 1983; Spain, 1984). Thus, the fire protection in DDF for multiple uses on a measurement of litterfall and estimation of nutrient return sustained yield basis (Komkris, 1971). Recently, some via litterfall have been carried out, not only for temperate study has been discussed on the effect of fire protection forest ecosystems but also for tropical forest ecosystems in DDF. Kanzaki et al. (1995) reported the change and (e.g., Vitousek, 1984; Jaramillo and Sanford, 1995) in study increase of tree species by the practice of fire protection on nutrient cycling. Among forest nutrients, nitrogen (N) in DDF. Sakurai et al. (1998) suggested the recovering is known to be greatly affected the primary productivity effect on soil properties by the fire protection. Both 200 Tetsuya TODA, Hiroshi TAKEDA, Naoko TOKUCHI, Seiichi OHTA, Chongrak WACHARINRAT and San KAITPRANEET

of the approaches are quite recommendable for the dry evergreen forest, plantation forest, and grassland and comprehensive understanding of the relation between bamboo forest. There is 13.4 km2 of DDF in the SERS. In DDF and fire protection. However, the effect of fire the fire protection area of the DDF, fire was excluded by control on N cycling in DDF is not known. Here, we 10-m wide fire belts in a part of the SERS (Sahunalu and focused on the relationship between the N return via tree Dhanmanonda, 1995; Phongamfai, 1997). litterfall and history of forest fires. The climate type of the SERS is tropical monsoon In the present study, we performed a comparative (Fig. 2): mean annual temperature is 22.5 ˚C; annual study in DDFs with different forest fire histories. The precipitation is 1097 mm, characterized by a pattern of objective of this study was to compare N return of three two peaks of precipitation (one in May and the other in stands with different forest fire histories via tree litterfall September). The monthly precipitation is less than 100 to the forest floor. We hypothesized that: (1) the tree mm during the dry season, from November to March. litterfall would increase, and (2) the N return via tree Because of the extremely dry conditions and human litterfall would increase, with the time after the fire activities in the dry season, DDF without fire protection protection, due to the change of tree species composition. almost inevitably has annual surface fires. The soil in the study area is stony red-yellow podzolic soil, with a depth of 60 cm or less (Bos and Thunduan, 1968; Sakurai et al. MATERIALS AND METHODS 1998). The soil is equivalent to that of Orthic Acrisols Study area (FAO/UNESCO, 1974) or Haplustults (USDA, 2006). The study was performed in the Sakaerat Environmental Research Station (SERS) located in Nakhonrachasima Plot setting and tree census Province, approximately 180 km northeast of Bangkok, Three study stands with different fire histories were , , Thailand (lat. 14˚30 N, long. 101˚56 E, 200−800 m in established in the DDF of the SERS. The forest fire elevation, Fig. 1). SERS is an 81 km2 biosphere reserve history of each stand is as follows: (1) a stand without established in 1978 that consists of four forest types, DDF, fire protection that was burned in January 2000 (F0),

��

�0 �35 �10

0 2 4��

��

Fig. 1. Location of research site and the vegetation map of the Sakaerat Environmental Research Station. (Modified, after Sakurai et al., 1998) Comparison of nitrogen return by litterfall in three dry dipterocarp forests, Thailand 201

250 35

30 200 25 �

150 ℃ 20

15 100

10 ��

�� 50 5

0 0 �� 2000 2001 2002 Fig. 2. Monthly precipitation and temperature during March 2000−Februar y 2002 at the Sakaerat Environmental Research Station.

(2) a stand protected for 10 years (F10), and (3) a stand The oven-dried and sorted litter samples collected protected for 35 years (F35). In each stand, a 50 m × 50 m in January, April, June, August, October, and December plot was established in February 2000. 2001 were ground up separately. The amount of Each plot was further divided into 25 subplots (10 m reproductive organs was so small that we mixed them × 10 m). In each subplot, we recorded the tree species with the miscellaneous litter. For the same reason, we and measured the diameters at breast height (DBH, 1.3 also mixed the leaf litter of the three species Mitragyna m above the ground) of all trees with a DBH of 3 cm or brunonis (Rubiaceae), Sindra siamensis (Leguminosae), lager. and Bauhinia spp. (Leguminosae) with the other species The lower layer was composed of saplings of canopy in F0. trees. A dwarf bamboo-like grass, Arundinaria pusilla (= Vietnamosasa pusilla, Poaceae), occupied the understory Statistical analysis both in F10 and F0. The grass was dense in F0. We found The total litterfall and weights of the four sorted no understory grass in F35. litter samples were analyzed for variability using non- parametric tests. Significant differences of the litterfall Litter trap setting and litter collection among the three plots were examined by the Kruskal- In each plot, litterfall was measured over a 2 year period Wallis test and Mann-Whitney U-test (SPSS, 1999). from March 2000 to February 2002. Thirteen litter traps, 1 m high with a diameter of 60 cm, were arranged in Chemical analysis the centers of the subplots in a staggered pattern. In F0, Total N concentrations of the sorted litter samples in 2001 forest ﬁres occurred at the end of January 2001 and 2002. were analyzed with an NC analyzer (Sumigraph NC900, We successfully collected litter samples and removed Sumika Chemical Co. Ltd., Japan). litter traps from F0 2 to 6 days before burning and reset the traps immediately after burning. Estimation of nitrogen content in litterfall Trapped litter was collected at monthly intervals We calculated N content in the tree litterfall in 2001 as and then oven-dried at 105 ˚C for 48 h. The litter was follows: (monthly N content in litter component) = (dry sorted into the following categories, (1) leaves; (2) weight of monthly litter component) × (N concentration woody material including branches, twigs, and bark; in each litter component). For February, March, May, (3) reproductive organs such as flowers and seeds; July, September, and November, we used the average and (4) miscellaneous material such as insects, feces, concentrations of the values of the two neighboring and amorphous material. Further, we sorted leaf litter months (e.g., the value of November was obtained from into each of the six most dominant species and another those of October and December). species category in each plot. 202 Tetsuya TODA, Hiroshi TAKEDA, Naoko TOKUCHI, Seiichi OHTA, Chongrak WACHARINRAT and San KAITPRANEET

ha−1, for F0, F10, and F35, respectively. RESULTS The three plots were dominated by the major Thai Forest structure DDF tree species. Shorea obtusa (Dipterocarpaceae) was The tree density was 444, 2020, and 2408 trees ha−1, for the dominant species in F0 and F10, but was not found in F0, F10, and F35, respectively (Table 1). The average F35. In F0, S. obtusa accounted for 62.3 % of the total basal DBH was 17.3, 8.44, and 8.41 cm, for F0, F10, and F35, area. In F35, Pterocarpus macrocarpus (Leguminosae) respectively. The basal area was 14.3, 19.4, and 22.4 m2 was the dominant species, occupying 22.9 % of the total

Table 1. Forest condition of study plots based on the census of trees with DBH ≥ 3 cm. F0 F10 F35

Fire frequency Every year No ﬁre for 10 years No ﬁre for 35 years

Tree density (ha−1) 444 2020 2408

Diameters at breast height 17.3 8.4 8.4 (DBH, cm)

Basal area (m2 ha−1) 14.3 19.4 22.4

Relative basal area of tree Shorea obtusa 62.3 Shorea obtusa 32.8 Pterocarpus macrocarpus 22.9 species to total basal area (%) Xylia xylocarpa 13.3 Dipterocarpus intricatus 17.6 Dipterocarpus intricatus 10.8 Pterocarpus macrocarpus 6.0 Pterocarpus macrocarpus 14.8 Irvingia malayana 10.1 Mitragyna brunonis 2.7 Xylia xylocarpa 6.5 Cratoxylum formosum 8.7 Sindora siamensis 2.7 Shorea siamensis 5.4 Sindora siamensis 6.8 Bauhinia spp. 0.7 Gardenia sootepensis 3.9 Xylia xylocarpa 5.7 Bauhinia spp. 2.4 Lannea coromandelica 5.6 Mitragyna brunonis 1.1 Gardenia sootepensis 3.2 Sindora siamensis 1.0 Clausena harmandiana 2.2 Quercus kerrii 1.0 Dalbergia nigrescens 2.0 Erythrophleum succirubrum 0.6 Bombax anceps 1.8 Lannea coromandelica 0.6 Albizia myriophylla 1.0 Morinda coreia 0.2 Bauhinia spp. 0.8 Cratoxylum formosum 0.2 Terminalia glaucifolia 0.4 Dalbergia oliveri 0.4 Terminalia bellirica 0.4 Antidesma acidum 0.3 Others 12.3 Others 11.9 Others 16.9

Table 2. Annual litterfall, its components (Mg ha−1 yr−1) and proportions to total litterfall (%, in parentheses) in the three plots in dry dipterocarp forest, in Thailand. Reproductive Miscellaneous Leaf Woody Total organs material Mar. 2000-Feb. 2001 3.23 (79.8) 0.56 (13.8) 0.08 (0.2) 0.25 ( 6.2) 4.12 (100) F0 Mar. 2001-Feb. 2002 2.83 (76.0) 0.53 (14.2) 0.07 (2.0) 0.29 ( 7.8) 3.72 (100) Mean 3.03a (77.9) 0.55a (14.0) 0.07a (1.1) 0.27a ( 7.0) 3.92a (100) Mar. 2000-Feb. 2001 5.09 (70.1) 1.02 (14.0) 0.28 (3.9) 0.87 (12.0) 7.26 (100) F10 Mar. 2001-Feb. 2002 5.15 (73.5) 1.00 (14.2) 0.15 (2.2) 0.71 (10.1) 7.01 (100) Mean 5.12b (71.8) 1.01b (14.1) 0.21b (3.1) 0.79b (11.0) 7.13b (100) Mar. 2000-Feb. 2001 5.92 (68.6) 1.61 (18.7) 0.17 (2.0) 0.93 (10.7) 8.63 (100) F35 Mar. 2001-Feb. 2002 5.92 (66.2) 2.36 (26.3) 0.10 (1.1) 0.57 ( 6.4) 8.95 (100) Mean 5.92c (67.4) 1.98c (22.5) 0.13b (1.6) 0.75b ( 8.5) 8.79c (100) The values denoted by the same letter are not signiﬁcantly different from each other among the three plots at P < 0.05 (the Mann-Whitney U-test after the Kruskal-Wallis test). Comparison of nitrogen return by litterfall in three dry dipterocarp forests, Thailand 203

basal area. P. macrocarpus was � � � a common species among the � 2002 � � � three plots. In addition to these � � � � � � � � � species, the following species �� occupied over 10 % of the � � � � � � � � total basal area, Dipterocarpus � � � � � � � � � intricatus (Dipterocarpaceae) �� a n d Irvingia malayana � � � � � � � � � (Irvingiaceae) in F35, D. 2001 � � � � �� intricatus in F10, and Xylia � � � � � � � � xylocarpa (Leguminosae) in � � � � � � � � F0. The number of tree species � � � � increased as the duration of ﬁre � � � � � �� protection increased. �� 2000 0 0 0 0

800 400 800 400

200 100 800 600 400 200

2400 2000 1600 1200 1200

� ��

1 1 1 1 � � � Annual litterfall and litterfall � components Mean annual litterfall during � � � � 2002 � � � the 2 year period was 3.92, � � � � −1 � � � � 7.13, and 8.79 Mg ha for F0, � �� F10, and F35, respectively � � � � ( T a b l e 2 ) . T h e a n n u a l � � � � � � � � litterfall increased with the � = 13 ). Rainy season: April-October, Dry season: November-March. � � � � �� year after fire protection. � n � � � � � � � In components, the annual � � � � � 2001 � � � litterfall, leaf litters, and woody � �� litters varied by more than an � � � � � � � � order of magnitude. There � � � � � � � � was no significant difference � � � � � �� between reproductive organs � �� and miscellaneous materials 2000 0 0 0 0

800 400 800 400

200 100 800 600 400 200

2400 2000 1600 1200 1200

� ��

1 1 1 1 � � � among the plots. In all the � plots, the leaf litter was the largest component, comprising 77.9 %, 71.8 %, and 67.4 % of the � � � � 2002 � � � total litterfall for F0, F10, and � � � � � � � � �

F35, respectively. The woody �� litter in F35 was 22.5 % of the � � � � � � � � total litterfall, greater than in � � � � � � � � the other plots. The litter of � �� reproductive organs in each plot � � � � � � � � � made only a small contribution 2001 � � � � � � � to the total litterfall, 1.1 % to 3.0 � � � �� %. � � � � � � � � � � � � � � � � � S e a s o n a l c h a n g e s i n � � � � � �� litterfall � 2000 �� The leaf litter amount was � 0 0 0 0 Fig. 3 . Seasonal changes in litterfall. Bars represent the standard deviations (

800 400 800 400

200 100 800 600 400 200

2400 2000 1600 1200 1200

� ��

1 1 1 1 � � � greater in the dry season � (November-March) than in 204 Tetsuya TODA, Hiroshi TAKEDA, Naoko TOKUCHI, Seiichi OHTA, Chongrak WACHARINRAT and San KAITPRANEET

Table 3. The comparisons of leaf litter production by major species (kg ha−1) and seasonal proportions to annual productions (%, in parentheses) between the rainy and dry seasons in 2000 and 2001. Rainy season: April-October, Dry season: November-March. 2000 2001 Average Rainy season Dry season Rainy season Dry season Rainy season Dry season (kg ha−1) (%) (kg ha−1) (%) (kg ha−1) (%) (kg ha−1) (%) (kg ha−1) (%) (kg ha−1) (%) F0 Total 315.3 (9.7) 2916.4 (90.3) 452.6 (16.0) 2373.6 (84.0) 384.0 (12.7) 2645.0 (87.3) S. obtusa 156.3 (9.2) 1542.5 (90.8) 315.8 (15.7) 1693.3 (84.3) 236.1 (12.2) 1617.9 (87.3) X. xylocarpa 0.3 (0.1) 269.2 (99.9) 5.7 (3.5) 156.0 (96.5) 3.0 (1.4) 212.6 (98.6) P. macrocarpus 16.3 (4.7) 328.3 (95.3) 34.3 (17.5) 162.2 (82.5) 25.3 (9.4) 245.3 (90.6) M. brunonis 7.8 (2.6) 294.3 (97.4) 15.4 (8.9) 157.6 (91.1) 11.6 (4.9) 226.0 (95.1) Sindora siamensis 7.5 (10.3) 65.6 (89.7) 6.5 (9.3) 63.5 (90.7) 7.0 (9.8) 64.6 (90.2) Bauhinia spp. 3.3 (6.4) 48.3 (93.6) 5.4 (8.8) 56.1 (91.2) 4.4 (7.7) 52.2 (92.3) Other species 123.8 (25.7) 368.2 (74.3) 69.5 (45.0) 84.9 (55.0) 96.7 (29.9) 226.6 (70.1) F10 Total 1030.7 (20.2) 4061.6 (79.8) 1225.8 (23.8) 3920.6 (76.2) 1128.3 (22.0) 3991.1 (78.0) S. obtusa 261.8 (18.2) 1178.1 (81.8) 210.4 (15.8) 1117.9 (84.2) 236.1 (17.1) 1148.0 (82.9) D. intricatus 83.2 (20.1) 329.8 (79.9) 156.8 (19.0) 668.3 (81.0) 120.0 (19.4) 499.1 (80.6) P. macrocarpus 109.2 (29.3) 263.2 (70.7) 77.6 (21.3) 286.3 (78.7) 93.4 (25.4) 274.8 (74.6) X. xylocarpa 3.6 (0.7) 478.2 (99.3) 19.9 (4.4) 428.4 (95.6) 11.8 (2.5) 453.3 (97.5) Shorea siamensis 4.0 (3.6) 105.6 (96.4) 13.1 (6.3) 193.9 (93.7) 8.6 (5.4) 149.8 (94.6) G. sootepensis 111.1 (31.6) 240.3 (68.4) 60.7 (15.9) 321.5 (84.1) 85.9 (23.4) 280.9 (76.6) Other species 457.8 (23.8) 1466.4 (76.2) 687.3 (43.2) 904.3 (56.8) 572.6 (32.6) 1185.4 (67.4) F35 Total 1839.6 (31.1) 4084.5 (68.9) 1584.5 (26.8) 4335.2 (73.2) 1712.1 (28.9) 4209.9 (71.1) P. macrocarpus 190.3 (28.0) 488.6 (72.0) 121.4 (14.7) 703.9 (85.3) 155.9 (20.7) 596.3 (79.3) D. intricatus 20.6 (15.7) 111.0 (84.3) 57.2 (21.0) 214.5 (79.0) 38.9 (19.3) 162.8 (80.7) I. malayana 51.7 (14.8) 298.2 (85.2) 100.9 (19.3) 422.3 (80.7) 76.3 (17.5) 360.3 (82.5) C. formosum 2.4 (1.2) 201.7 (98.8) 171.8 (49.3) 176.6 (50.7) 87.1 (31.5) 189.2 (65.8) Sindora siamensis 23.9 (19.2) 100.6 (80.8) 35.9 (17.8) 166.0 (82.2) 29.9 (18.3) 133.3 (81.7) X. xylocarpa 19.1 (6.0) 299.1 (94.0) 38.9 (11.8) 291.0 (88.2) 29.0 (8.9) 295.1 (91.1) Other species 1531.6 (37.2) 2585.3 (62.8) 1058.4 (31.0) 2360.9 (69.0) 1295.0 (34.4) 2473.1 (65.6)

the rainy season (April-October) (Fig. 3). This pattern species, Cratoxylum formosum (Guttiferae) in F35 and was similar among the three plots. The other litterfall Gardenia sootepensis (Rubiaceae) in F10 had large components had no clear seasonality. seasonal variations in the proportion of litterfall between Table 3 shows the leaf litter distributions in the the 2 years. The minor species, which were placed in the rainy and dry seasons. In F0, 84.0 % to 90.3% of the other species category, shed over 23 % of their leaves in annual litterfall fell in the dry season. In F10 and F35, the the rainy seasons, irrespective of plot and year. percentage of leaf litter in the dry season ranged from 68.9 % to 79.8 %. Litterfall N content The seasonal changes in leaf litter of the major The N concentrations of leaf litter had a seasonal trend: species (>10 % on a basal area basis) were as follows: high in the rainy season and low in the dry season (Table S. obtusa shed over 80 % of annual leaf litter in the dry 4). Dry season N concentration was lowest in S. obtusa in seasons both in F10 and F0; P. macrocarpus shed over 70 every plot, and rainy season N concentration was highest % of annual leaf litter in the dry season; and D. intricatus in P. macrocarpus in every plot. The N concentration of shed about 80 % of annual leaf litter in the dry season both leaf litter in the other species category in F35 was the in F35 and F10. On the other hand, two of the nonmajor highest in the three plots, and reached a maximum (2.62 Comparison of nitrogen return by litterfall in three dry dipterocarp forests, Thailand 205

Table 4. Seasonal changes in nitrogen concentration (% dry weight) of litterfall component in 2001. †Other species in F0 include M. brunonis, Sindra siamensis, and Bauhinia spp.. § Other components indicates the nitrogen concentrations analysed from the mixed samples of reproductive organs and miscellaneous material litterfall. N.A means data is not available due to insufﬁcient sample size for chemical analysis.

―Dry season→ ←――――Rainy season――――→ ←Dry season― Component January April June August October December

F0 Leaf S. obtusa 0.61 1.25 1.49 0.67 0.51 0.65 X. xylocarpa 1.09 N.A N.A 1.40 N.A 1.08 P. macrocarpus 1.27 2.30 1.46 1.09 1.33 1.25 Other species† 0.95 N.A 1.65 1.38 0.97 1.20 Woody 0.72 N.A 0.80 0.78 1.11 0.70 Other components§ N.A 1.72 2.19 N.A 1.64 1.34

F10 Leaf S. obtusa 0.72 1.61 1.54 0.75 1.40 0.65 D. intricatus 0.81 1.58 N.A 1.25 0.99 0.87 P. macrocarpus 1.15 N.A 2.35 2.21 1.72 1.14 X. xylocarpa 1.26 N.A 1.84 1.68 N.A 1.28 Shorea siamensis 0.69 N.A 1.35 N.A N.A 0.61 Gardenia sootepensis 0.79 N.A N.A 0.77 0.84 0.73 Other species 0.83 1.54 1.37 1.65 1.40 0.97 Woody 0.56 0.76 0.63 0.63 0.73 0.54 Other components§ N.A 2.00 1.61 1.09 1.71 1.91

F35 Leaf P. macrocarpus 1.44 N.A 2.15 1.52 1.66 1.06 D. intricatus 1.35 N.A 1.25 N.A N.A 1.08 I. malayana 1.34 1.45 N.A N.A 1.30 1.27 C. formosum 1.22 1.29 1.62 0.98 0.93 1.01 Sindora siamensis 1.29 1.50 1.88 N.A 1.30 1.13 X. xylocarpa 1.14 1.16 1.21 N.A N.A 1.39 Other species 1.16 2.62 1.79 2.16 1.44 1.53 Woody 0.91 0.84 0.76 0.69 1.14 0.75 Other components§ N.A 2.52 2.45 N.A 2.11 1.93

Table 5. Nitrogen content in litterfall in 2001 (kg ha−1) and proportions to total content (%, in parentheses). § Other components calculated from the mixed samples (reproductive organs and miscellaneous materials). F0 F10 F35 Tree -Leaf 23.3 (69.7) 52.7 (70.2) 87.9 (71.0) -Woody 4.1 (12.3) 6.2 (8.2) 19.6 (15.8) -Other components§ 6.0 (18.0) 16.2 (21.6) 16.3 (13.2) Total 33.4 (100.0) 75.1 (100.0) 123.8 (100.0) 206 Tetsuya TODA, Hiroshi TAKEDA, Naoko TOKUCHI, Seiichi OHTA, Chongrak WACHARINRAT and San KAITPRANEET

%) in April. protection was 8.65 Mg ha−1, suggesting that the practice The N content in the litterfall was calculated for of fire protection in DDF increases litter production. the three plots based on the litter amounts and their N The litterfall composition was characterized by high concentrations (Table 5). The N content in tree litter percentages of leaf litter in every plot (Table 2). Seasonal was 33.4, 75.1 and 123.8 kg ha−1, for F0, F10, and F35, change in leaf litter production in DDF is strongly respectively. The N content in the litterfall increased with affected by the dry conditions of air and soil (Thaiutsa the year after fire protection. et al., 1978). Moore (1989) reported that water stress triggers the synthesis of abscisic acid in the foliage of plants, which in turn stimulates senescence of leaves and DISCUSSION other parts. Litterfall of tropical deciduous species peaks Effects of fire protection on forest structure and in the dry season in tropical dry forests (e.g., Kumar litterfall in DDF and Deepu, 1992; Muoghalu et al. 1993; Sundarapandian Trees in DDF are often and regularly affected by forest and Swamy, 1999). In the present study, the litterfall fires, and thus have adapted to this severe stress. Sakurai peaked in the dry season, especially in F0 (Table 3). This et al. (1998) described the adaptation of DDF trees to is partly because of the tree species composition of F0, the forest fire environment. By frequent forest fire, where S. obtusa accounted for 60 % or more on a basal selection of tree species occurred first, and adaptation of area basis. The seasonal pattern of litterfall in F0 was the trees to the harsh environment, such as thickened mainly represented by S. obtusa. Whereas in F10 and F35, bark, occurred subsequently. The dominant species in the concentration of litterfall in the dry season gradually F0, S. obtusa (Table 1), is also a fire-resistant tree species decreased for the variety of tree species. By the practice with thick bark (Sahunalu and Dhanmanonda, 1995). In of fire protection in DDF, the seasonal pattern of litterfall F0, the tree density was lower than in F10 and F35, and changed to be dispersed through the year. consisted of trees with a large DBH, whereas F35 and The litterfall pattern of some same species also F10 consisted of trees with a smaller DBH (Table 1). The tended to change by fire protection. S. obtusa, which sum of the relative basal area of the minor tree species (< appears in F0 and F10, shed in the dry season in F0 grater 10 % on a basal area basis) in F35 was 56.2 % of the total than in F10 (Table 3). P. macrocarpus and X. xylocarpa, basal area and higher than those in F0 and F10. Tree common species among the three plots, shed more of species composition might have been changed due to the their leaves in the dry seasons in F0 than in F35. The fire protection. practice of forest fire protection in DDF might not only Kanzaki et al. (1995) reported the canopy species increase species number, but also change the seasonal composition of F35 and F10 was similar in 1993. Both pattern of leaf litter production. plots were predominated by canopy trees of S. obtusa and P. macrocarpus. The long-term fire protection in F35 from Effect of fire protection on N content in litterfall in , the 1960 s, however, excluded the grass species A. pusilla DDF and allowed for an increase of less fire-tolerant pioneer The N concentration of leaf litter in F35 was higher and species such as C. formosum and Clausena harmandiana that in F0 lower in the dry season (Table 4). Firstly, N (Rutaceae) (Table 1), and invasion of some evergreen concentration of leaf litter is regulated by N concentration tree species such as Walsura trichostemon (Meliaceae) of fresh leaves. Additionally, N retranslocation from and Memecylon ovatum (Merastomataceae) (Kanzaki et senescing leaves occurred, and is affected by complex al., 1995). By the time we studied, additionally, all of the S. factors e.g., each tree character of nutrient conservation obtusa trees in F35 disappeared (Chongrak Wacharinrat, mechanism at the species level or relationship with unpublished data). neighboring plants, etc. (Aerts and Chapin, 2000). Annual litterfall ranges from 3 to 10 Mg ha−1 in N retranslocation is effective in poor environmental tropical dry forests (Murphy and Lugo, 1986). The annual conditions where there is nutritional constraint, like in litterfall observed in the present study was within that F0 (Ernst and Tolsma, 1989; Tripathi and Singh, 1994). range. There was a significant difference in the annual Tripathi and Singh (1994) reported the substantial rates litterfall among the three plots (Table 2). The annual of N retranslocation in some vegetation types in Indian litterfall increased with an increasing duration of fire tropical dry forest. They suggested the relation between , protection. Wacharinrat and Takeda (2003) reported vegetation and site s N fertility. Although the degree of that the amount of litterfall in a DDF with 37 year fire N retranslocation was not measured in the present study, Comparison of nitrogen return by litterfall in three dry dipterocarp forests, Thailand 207 the variations in the N concentration in leaf litter among the plots might reflect differences in N availability and REFERENCES cycling among the plots. Aerts, R. & Chapin III, F.S. 2000. The mineral nutrition The increased N return by leaf litter in F35 might of wild plants revisited: A re-evaluation of processes depend on the increase of leaf biomass and N-rich leaf and patterns. Advances in Ecological Research, 30: litter. As the duration of fire protection increased, the 1−67. abundance of S. obtusa decreased (Table 1), while P. Akaakara, S. 2002. Special Report on Forest Fire. macrocarpus, D. intricatus, and various tree species with International Forest Fire News, 26: 100−105. higher N concentrations than S. obtusa increased (Table Bos, F. & Thunduan, V. 1968. Detailed soil survey of 4). Thus, the practice of fire protection in DDF increases the north-eastern portion of ASRCT Sakaerat the litterfall N content and N return. Experimental Station, ASRCT Report No.1. 22p. Comprehensive nutrient data for tropical dry forests ASRCT, Bangkok. are limited. According to a review of the limited number Ernst, W.H.O. & Tolsma, D.J. 1989. Mineral nutrients in of studies available, the range of annual N return by some Botswana savanna types. In: Proctor, J. (ed.), litterfall in tropical dry forests is between 48 and 156 Mineral Nutrients in Tropical Forest and Savanna kg ha−1 (Jaramillo and Sanford, 1995). The N content in Ecosystems. Blackwell Scientific Publications, litterfall for F35 and F10 was within the range of values Oxford, pp. 97−120. recorded in other tropical dry forests (Table 5), whereas FAO/UNESCO 1974. Soil map of the world Vol. 1. the N content in F0 was lower than the range recorded in Regend. UNESCO, Paris. other tropical dry forests. Gillon. & Rapp, M. 1989. Nutrient losses during a winter In F0, the tree density was 20 % that of F35 (Table low-intensity prescribed fire in a Mediterranean 1) and the forest floor was densely covered by A. pusilla. forest. Plant and Soil, 120: 69−77. Although the N return by tree litter in F0 was small, Jaramillo, V.J. & Sanford, Jr., R.L. 1995. Nutrient cycling compared with the other plots, the aboveground biomass in tropical deciduous forests. In: Bullock, S.H., of A. pusilla reached 5.8 Mg ha−1 and N accumulated Mooney, H.A. & Medina, E. (eds.) Seasonally Dry more than 70 kg ha−1 (Toda et al. 2007). It is estimated the Tropical Forests. Cambridge University Press, significant amount of N return if we added the pathway Cambridge, pp. 346−361. through the grass to the tree in F0. The N return by Kanzaki, M. Kagotani, Y., Kawasaki, T., Yoda, K., tree litter had a clear correspondence relationship with Sahunalu, P., Dhanmanonda, P., Prachaiyo, B. the duration of fire protection in DDF, however, a better & Thoranisorn, S. 1995. Forest structure and understanding of the N dynamics of grass is necessary to composition of tropical seasonal forests of Sakaerat assess N cycling in disturbed forest ecosystems that are Environmental Research Station and the effects of affected by forest fire, like F0. fire protection on a dry deciduous forest. In: Yoda, K., Sahunalu, P. & Kanzaki, M. (eds.) Elucidation ACKNOWLEDGEMENTS The authors thank of the Missing Sink in the Global Carbon Cycling Messrs. Rattana Thaingam, Thiti Visarattana, and Suchat –Focusing on the Dynamics of Tropical Seaonal Nimpila of Thai Royal Forest Department, for site Forests–. Osaka City University, Osaka, pp. 1−20. arrangement. We wish to thank Miss Wandi Jettanadi, Komkris, T. 1971. Forest policy. 317p. Faculty of forestry, the all staffs of the Sakaerat Field Station, and Sakaerat Kasetsart University, Bangkok. Environmental Research Station for assistance in the field Kumar, B.M. & Deepu, J.K. 1992. Litter production and and in the laboratory. We thank Dr. Mamoru Kanzaki at decomposition dynamics in moist deciduous forests Kyoto University for the site introduction and valuable of the Western Ghats in Peninsular India. Forest comments. And our thanks also to Mr. Tetsuji Iwama Ecology and Management, 50: 181−201. of Forestry Agency of Japan for help during the field Marafa, L.M. & Chau, K.C. 1999. Effect of hill fire on working. This research was supported by the Japan upland soil in Hong Kong. Forest Ecology and Ministry of Education, Science, Sports, and Culture Management, 120: 97−104. through its Creative Basic Research (No. 09NP1501). Moore, T.C. 1989. Biochemistry and physiology of plant hormones. 330p. Springer Verlag, New York. Muoghalu, J.I., Akanni, S.O. & Eretan, O.O. 1993. Litterfall and nutrient dynamics in a Nigerian 208 Tetsuya TODA, Hiroshi TAKEDA, Naoko TOKUCHI, Seiichi OHTA, Chongrak WACHARINRAT and San KAITPRANEET

rainforest seven years after a ground fire. Journal of Tokyo. (in Japanese). Vegetation Science, 4: 323−328. USDA. 2006. Keys to soil taxonomy 10th edition. 333p. Murphy, P.G. & Lugo, A.E. 1986. Ecology of tropical dry USDA, Washington, DC. forest. Annual Review of Ecology and Systematics, Vitousek, P.M. 1984. Litterfall, nutrient cycling, and 17: 67−88. nutrient limitation in tropical forests. Ecology, 65: Pongampai, S. 1997. Plant exploration in Sakaerat 285−298. Environmental Research Station. 408p. National Vitousek, P.M. & Howarth, R.M. 1991. Nitrogen Statistical Office, Bangkok. (in Thai). limitation on land and in the sea: How can it occur? Proctor, J. 1983. Tropical forest litterfall. I. Problems of Biogeochemistry, 13: 87−115. data comparison. In: Sutton, S.L., Whitmore, T.C. & Wacharinrat, C. & Takeda, H. 2003. Litterfall pattern of Chadwick, A.C (eds.) Tropical Rain Forest: Ecology dry deciduous and evergreen forests in Sakaerat and Management. Blackwell Scientific Publications, and vicinity area, North eastern Thailand. In: Oxford, pp. 267−273. Takeda, H., Khamyong, S. & Wiwatiwitaya, D. (eds.) Royal Forest Department. 2000. Forestry statistics of Decomposition and nutrient cycling processes Thailand 1999. 150p. Royal Forest Department data in tropical seasonal forests in Thailand. NRCT, center information office, Bangkok. Bangkok, pp. 7−23. Sahunalu, P. & Dhanmanonda, P. 1995. Structure and dynamics of dry dipterocarp forest, Sakaerat, Received 7th Feb. 2007 northeastern Thailand. In: Box, E.O., Peet, R.K., Accepted 18th Jan. 2008 Masuzawa, T., Yamada, I., Fujiwara, K. & Maycock, P.F. (eds.) Vegetation Science in Forestry. Kluwer Academic Publishers, Dordrecht, pp. 465−494. Sakurai, K., Tanaka, S., Ishizuka, S. & Kanzaki, M. 1998. Differences in soil properties of dry evergreen and dry deciduous forests in Sakaerat Environmental Research Station. Tropics, 8: 61−80. Spain, A.V. 1984. Litterfall and the standing crop of litter in three tropical Australian rainforests. Journal of Ecology, 72: 947−1061. SPSS. 1999. SPSS Base 10.0J for Windows. 484p. SPSS Inc., Chicago. Sundarapandian, S.M. & Swamy, P.S. 1999. Litter production and leaf-litter decomposition of selected tree species in tropical forests at Kodayar in the Western Ghats, India. Forest Ecology and Management, 123: 231−244. Thaiutsa, B., Suwannapinunt, W. & Kaitpraneet, W. 1978. Production and chemical composition of forest litter in Thailand. 32p. Forest Research Bulletin 52, Kasetsart University. Toda, T., Takeda, H., Tokuchi, N., Ohta, S., Wacharinrat, C. & Kaitpraneet, S. 2007. Effects of forest fire on the nitrogen cycle in a dry dipterocarp forest, Thailand. Tropics, 16: 41−45. Tripathi, S.K. & Singh, K.P. 1994. Productivity and nutrient cycling in recently harvested and mature bamboo savannas in the dry tropics. Journal of Applied Ecology, 31: 109−124. Tsutsumi, T. 1989. The life of forest -Nutrient cycling between trees and soils-. 197p. Chuoukouronsha,