Soil seed bank composition and regeneration potential in disturbed tropical dry forests

Tropical dry forests rank among the most threatened terrestrial ecosystems globally, yet their capacity for self-repair after disturbance depends heavily on what lies buried in the soil. The seed bank is the hidden reservoir from which regeneration begins once canopy removal, fire, or grazing ceases. This research assessed the composition, density, and vertical distribution of the soil seed bank across four disturbance categories in a tropical dry forest fragment near Manila, Philippines. Soil samples were collected from 48 plots at three depth intervals (0-5, 5-10, and 10-15 cm) during both wet and dry seasons between May 2023 and March 2024. Samples were spread in a greenhouse and monitored for seedling emergence over 16 weeks. A total of 8,741 seedlings representing 83 species germinated, with grasses and herbaceous pioneers dominating heavily disturbed plots (72.4% of germinants) and woody species more common in undisturbed forest soils (38.7%). Seed density ranged from 347 seeds/m² in undisturbed forest to 1,024 seeds/m² in heavily disturbed areas, but species richness showed the opposite pattern. The top 0-5 cm layer held 64.3% of all viable seeds. Seasonal comparison revealed 31% higher germination from wet-season samples than dry-season collections. These results indicate that while disturbed tropical dry forests retain a seed bank capable of initiating recovery, the dominance of ruderal species suggests that passive regeneration alone may not restore original tree composition. Active planting of native tree species alongside natural seed bank recruitment would likely accelerate recovery toward a more structurally complete forest.

Illustrations explain the paper; charts are replotted from the paper's own table values.

AI-generated illustration

↗ Enlarge

Disturbance vs. Seed Bank QualityClose ×

AI-generated conceptual illustration (Tropical dry forest soil profiles). It explains the idea and does not show study data.

Open full-size image ↗

Replotted from source data

↗ Enlarge

Density: Undisturbed vs Lightly dist. vs Moderately dist. vs Heavily dist.Close ×

Replotted from Table 1 (page 2) of the original paper. Every bar is the paper's own table value.

Open full-size image ↗

Replotted from source data

↗ Enlarge

Richness: Undisturbed vs Lightly dist. vs Moderately dist. vs Heavily dist.Close ×

Replotted from Table 1 (page 2) of the original paper. Every bar is the paper's own table value.

Open full-size image ↗

Introduction

A soil seed bank is the pool of viable but ungerminated seeds present in the soil at any given moment, and it represents the first line of ecological defense when vegetation above ground is damaged or destroyed [1] . In tropical dry forests, where up to six months of annual drought already impose strong selective filters on plant establishment, the seed bank takes on added importance because seedlings must emerge and anchor themselves before the next dry season arrives [2] . The Philippines lost an estimated 93% of its original dry forest cover during the twentieth century to agriculture, fuelwood harvesting, and urban expansion [3] . Remaining fragments are under continued pressure from edge effects, invasive species, and periodic burning. Whether these remnants can bounce back after disturbance stops depends largely on what seeds survive in the soil and whether those seeds belong to species capable of rebuilding forest structure [4] . Previous seed bank research in Southeast Asian dry forests has focused mainly on Thailand and Myanmar, with very little data from the Philippine archipelago [5] . The few Philippine surveys available examined lowland dipterocarp forests rather than the seasonally dry formations found on western Luzon and the Visayas. And because seed bank composition shifts with disturbance intensity pioneer species increase while shade-tolerant trees decline findings from undisturbed forests cannot simply be extrapolated to degraded sites [6] . This research was designed to fill that regional gap. We sampled the soil seed bank across four disturbance levels in a 230-hectare dry forest fragment, recorded germination over 16 weeks in a greenhouse, and compared wet-season versus dry-season seed bank composition. Our questions were: c) What proportion of the seed bank consists of woody species capable of contributing to canopy regeneration [7] ?

Study area description

Study area description

Methods

In the greenhouse, each pooled soil sample was spread to a depth of 2 cm over a 3 cm layer of sterile vermiculite in plastic germination trays (35 × 25 cm). Trays were watered daily to field capacity and maintained under natural light with shade cloth reducing irradiance by 30%. Emerging seedlings were counted and identified weekly for 16 weeks.

Seedlings that could not be identified at the cotyledon stage were transplanted and grown until diagnostic features appeared. Control trays filled with sterile vermiculite only were included to check for greenhouse contamination; fewer than 4 seedlings appeared in controls across all batches [8] . Species were classified into life-form categories: grasses, herbaceous forbs, woody pioneers, and woody non-pioneers. Seed density was expressed as seeds/m² by extrapolating core area. Species richness, Shannon diversity (H′), and

Sørensen similarity between disturbance classes were calculated. Differences in seed density and richness among disturbance levels and depths were tested with two-way ANOVA followed by Tukey HSD in R v4.3.2 [9] .

Seasonal Variations

Wet-season collections (July 2023) coincided with the peak fruiting period of most canopy tree species in the fragment, while dry-season collections (February 2024) followed five months of negligible rainfall. Mean soil moisture at the time of wet-season sampling was 34.8% (gravimetric), compared with 11.2% during the dry-season campaign. Temperature in the top 5 cm averaged 26.1°C in July and 28.7°C in February. These contrasting conditions allowed assessment of whether seed viability and composition differ between the two seasonal extremes. Seed density increased with disturbance intensity (F₃,₄₄ = 24.7, p<0.001), driven by a proliferation of grass and herbaceous forb seeds in open areas. But species richness moved in the opposite direction, peaking in undisturbed forest (22 species in the 0-5 cm layer) and declining to 13 species in heavily disturbed plots. The proportion of woody species in the seed bank dropped from 38.7% in undisturbed to 8.4% in heavily disturbed top-layer samples. Depth stratification was consistent across all categories: the 0-5 cm layer contained 64.3% of all viable seeds, the 5-10 cm layer held 24.8%, and the 10-15 cm layer accounted for 10.9%. Wet-season samples produced 31% more seedlings than dry-season samples overall, with the difference most pronounced for small-seeded pioneer species. Sørensen similarity between undisturbed and heavily disturbed plots was only 0.28, indicating that the two extremes share fewer than a third of their seed bank species. The most frequently encountered species in heavily disturbed soils were Imperata cylindrica (present in 92% of plots), Chromolaena odorata (83%), and Lantana camara (71%) all aggressive invaders that suppress native seedling establishment [10] .

Results

Results

Discussion

The inverse relationship between seed density and species richness across disturbance levels is a pattern reported in

International Journal of Biology Sciences https://www.biologyjournal.net tropical dry forests elsewhere but hadn't been documented for the Philippines before this research [4] . High seed densities in disturbed areas largely reflect the prolific output of a few ruderal species rather than a diverse seed rain. Imperata cylindrica alone accounted for 34% of all germinants in heavily disturbed plots, and its rhizomatous growth habit means that even a modest seed bank presence can translate into rapid ground coverage that shades out slower-growing tree seedlings [11] . The near-absence of woody non-pioneer species from the seed bank of disturbed areas (3.8-8.4%) is concerning for natural regeneration prospects. Most large-seeded tropical dry forest trees have transient seed banks that lose viability within weeks of dispersal [12] . Once the parent trees are removed, the soil reservoir of their seeds is quickly exhausted. This means that passive recovery in heavily disturbed plots would likely stall at a grass-dominated or scrub stage unless seed sources are reintroduced through planting or by attracting frugivorous dispersers from nearby intact forest. The 31% seasonal difference in germination highlights a methodological point: single-season seed bank surveys may substantially underestimate true diversity. Seeds of some species require specific cues such as fluctuating temperatures or smoke chemicals that only arise during or shortly after the dry season [13] . Combining wet-and dryseason sampling, as we did here, gives a more complete inventory.

Conclusion

This research showed that disturbed tropical dry forests in the Philippines retain a numerically large but compositionally impoverished soil seed bank dominated by invasive grasses and herbaceous pioneers. While this reservoir can initiate rapid ground cover after disturbance, it is unlikely to restore native tree composition without active intervention. Conservation managers working in Philippine dry forest fragments should consider enrichment planting of native tree species alongside invasive species control to complement natural seed bank recruitment. The finding that over 64% of viable seeds sit in the top 5 cm of soil also argues against management practices such as deep plowing or topsoil removal that would destroy this shallow regeneration layer.

Figures and tables from the paper

Category	Depth	Density	Richness	H′	Woody %
Undisturbed	0-5 cm	248	22	2.74	38.7
Undisturbed	5-10 cm	71	14	2.18	41.3
Undisturbed	10-15 cm	28	7	1.63	44.1
Lightly dist.	0-5 cm	312	19	2.51	29.4
Lightly dist.	5-10 cm	98	12	2.04	31.8
Lightly dist.	10-15 cm	41	6	1.47	28.6
Moderately dist.	0-5 cm	483	16	2.12	18.3
Moderately dist.	5-10 cm	147	10	1.78	14.7
Moderately dist.	10-15 cm	64	5	1.31	11.2
Heavily dist.	0-5 cm	718	13	1.84	8.4
Heavily dist.	5-10 cm	224	8	1.42	6.1
Heavily dist.	10-15 cm	82	4	0.97	3.8

Table text extracted from the original PDF; superscript letters are shown inline. The original table image is below.

View original table image

↗ Enlarge

Table 1Close ×

Soil seed bank density and species richness across four disturbance categories and three soil depth layers

Open full-size image ↗

Original paper tableSource PDF · page 2 ↗

↗ Enlarge

Figure 1Close ×

Box plot of soil seed bank density (seeds/m²) across four disturbance categories, showing increasing density but decreasing

Open full-size image ↗

Original paper figureSource PDF · page 3 ↗

↗ Enlarge

Figure 2Close ×

Violin plot showing the distribution of bryophyte species richness per sample across three soil depth layers, with the highest values

Open full-size image ↗

Original paper figureSource PDF · page 3 ↗

1. Thompson K, Grime JP. Seasonal variation in the seed banks of herbaceous species in ten contrasting habitats. Journal of Ecology. 1979;67(3):893-921.
2. Vieira DLM, Scariot A. Principles of natural regeneration of tropical dry forests for restoration. Restoration Ecology. 2006;14(1):11-20.
3. Kummer DM. Deforestation in the postwar Philippines. Geography Research Forum. 1992;12:1-18.
4. Garwood NC. Tropical soil seed banks: a review. In: Leck MA, Parker VT, Simpson RL, eds. Ecology of Soil Seed Banks. Academic Press; 1989:149-209.
5. Marod D, Kutintara U, Yarwudhi C, Tanaka H, Nakashizuka T. Structural dynamics of a natural mixed deciduous forest in western Thailand. Journal of Vegetation Science. 1999;10(6):777-786.
6. Dalling JW, Swaine MD, Garwood NC. Dispersal patterns and seed bank dynamics of pioneer trees in moist tropical forest. Ecology. 1998;79(2):564-578.
7. Ceccon E, Huante P, Rincon E. Abiotic factors influencing tropical dry forests regeneration. Brazilian Archives of Biology and Technology. 2006;49(2):305312.
8. Ter Heerdt GNJ, Verweij GL, Bekker RM, Bakker JP. An improved method for seed-bank analysis: seedling emergence after removing the soil by sieving. Functional Ecology. 1996;10(1):144-151.
9. R Core Team. R: a language and environment for statistical computing. Vienna: R Foundation for Statistical Computing. 2023.
10. Swamy V, Pinedo-Vasquez M. Secondary forest seed banks in the Amazon region. Biotropica. 2014;46(2):226-235.
11. MacDonald GE. Cogongrass (Imperata cylindrica): biology, ecology, and management. Critical Reviews in Plant Sciences. 2004;23(5):367-380.
12. Shen Y, Liu W, Baskin JM, Baskin CC, Cao M. Persistent soil seed banks of 14 species from a tropical rain forest of southwest China. Seed Science Research. 2006;16(2):151-157.
13. Baskin CC, Baskin JM. Seeds: Ecology, Biogeography, and Evolution of Dormancy and Germination. 2nd ed. San Diego: Academic Press. 2014.
14. Janzen DH. Management of habitat fragments in a tropical dry forest: growth. Annals of the Missouri Botanical Garden. 1988;75(1):105-116.
15. Powers JS, Becknell JM, Irving J, Perez-Aviles D. Diversity and structure of regenerating tropical dry forests in Costa Rica. Forest Ecology and Management. 2009;258(6):959-970.