Effects of leaf silicon on drought performance of tropical tree seedlings

Abstract

Elevated leaf silicon (Si) concentrations improve drought resistance in cultivated plants, suggesting Si might also improve drought performance of wild species. Tropical tree species, for instance, take up substantial amounts of Si, and leaf Si varies markedly at local and regional scales, suggesting consequences for seedling drought resistance. Yet, whether elevated leaf Si improves seedling drought performance in tropical forests is unknown. To manipulate leaf Si concentrations, seedlings of seven tropical tree species were grown in Si-rich and -poor soil, before exposing them to drought in the forest understorey. Survival, growth and wilting were monitored. Elevated leaf Si did not improve drought survival and growth in any of the species. In one species, drought survival was reduced in seedlings previously grown in Si-rich soil, contrary to our expectation. Our results suggest that elevated leaf Si does not improve drought resistance of wild tropical tree species. Elevated leaf Si may even reduce drought performance, suggesting differences in soil conditions influencing leaf Si may contribute to soil-related variation of tropical seedling performance. Furthermore, our results are at odds with most studies on cultivated species and show that alleviative effects of Si in crops cannot be generalized to wild plants in natural systems.

1. Introduction

Tropical forests harbour an enormous diversity of tree species and offer important ecosystem services, e.g. as globally important carbon sink [1]. In tropical forests, drought is a dominant environmental factor [2], shaping tree species performance and distribution as well as forest composition, structure and biomass [3–6]. Furthermore, rainfall is becoming increasingly variable and unpredictable with climate change, leading to longer and/or more intense drought periods in the tropics [7]. Thus, understanding drought resistance and performance of tropical tree species and its ecological consequences, has been a major research focus.

Seedlings are particularly vulnerable to drought, because their root system is shallow and they must compete over water with neighbouring adults [8]. To withstand drought, they evolved dessication tolerance and avoidance mechanisms [4,9], including leaf shedding, delayed turgor loss and high resistance to embolism [4,9,10]. However, whether elevated leaf silicon (Si) concentrations improve drought performance of tropical tree seedlings, as is often observed in cultivated species [11,12], has not been investigated.

Elevated leaf Si of droughted cultivated species can improve tissue hydration, water-use efficiency and/or photosynthesis, with positive effects on growth [11,12]. The mechanisms of how leaf Si interacts with the plant water status and thereby improves drought resistance remain debated [13]. Effects on cuticular resistance and stomatal closure [14] and on molecular processes, including gene regulation, antioxidant and phytohormone activity have been invoked [15]. Furthermore, most work on Si-induced plant drought resistance has been done in highly controlled environments, e.g. hydroponics or greenhouses [12] and only few field studies were undertaken (e.g. [16]), precluding conclusions about the relevance of Si-mediated drought resistance for wild tropical seedlings. Particularly, the very low water potentials of droughted tropical soils [17] have not been reached. Furthermore, drought performance of seedlings is affected by herbivory, pathogens, light and nutrients, all of which might exacerbate the negative effects of drought on seedling performance [18]. If elevated leaf Si was an ecologically meaningful mechanism to improve drought resistance in tropical tree species, seedlings with elevated leaf Si should have higher drought performance compared to conspecifics with low leaf Si under natural conditions.

Besides direct effects of tissue Si on plant physiology, elevated soil Si might modulate soil processes, potentially explaining the positive growth responses of Si-fertilized plants under drought [19]. Experiments have shown that increasing biogenic amorphous Si in soils improves its water holding capacity [20], increasing water availability to plants [21]. Thus, to fully appreciate the role of elevated leaf Si as a drought-resistance mechanism it needs to be evaluated in isolation from potential soil Si effects.

Considering that leaf Si concentrations vary across and within wild tropical tree species [22,23], e.g. due to variation in soil Si, moisture, and nutrient availability [23–25], Si-mediated drought resistance might have pervasive consequences for species performance and distribution, especially in the light of climate change. Thus, we tested whether elevated leaf Si improves drought performance of tropical tree seedlings in the forest understorey. We hypothesized that if Si improves drought resistance of tropical tree species, seedlings with high leaf Si should show higher growth and survival under drought than seedlings with low leaf Si. To experimentally manipulate leaf Si, we raised seedlings on Si-rich versus Si-poor soil.

2. Material and methods

The study was conducted in Gamboa, Panama (09°07′ N, 79°42′ W). The climate is moist tropical with mean annual rainfall of 2100 mm, a pronounced four-month dry season and mean temperature of 27°C [26,27]. The area has late secondary semi-deciduous moist forests and was selected due to its Si-poor soils (see electronic supplementary material, table S1; [22]).

(a) . Study species and plant material

We studied seedlings of seven shade-tolerant tree species from seven families (see electronic supplementary material, table S2). We refer to the species by their genus name.

Seeds were collected in forests of the Panama Canal area in August and September 2021 from at least three trees per species with a minimum distance of 100 m between them. Seeds from different parent trees were mixed to avoid genotype effects, germinated and raised until cotyledon stage or development of first foliage leaves on a Si-poor substrate (see below).

(b) . Leaf Si manipulation

Seedlings were then transplanted into individual pots (diameter: 6.5 cm, depth: 36 cm) with a Si-poor (−Si) or -rich (+Si) experimental substrate, in which they grew for 2.5 months. Both experimental substrates were a 1 : 1 mixture of Si-poor local forest soil and washed river sand. The +Si substrate contained an additional 18 g l⁻¹ hydrophilic pyrogenic silicon dioxide (Aerosil 300, Evonik Industries AG; for details see Klotz et al. [23]). The resulting plant-available soil Si concentrations were 2 and 5 mg kg⁻¹ for the −Si and +Si substrate, respectively. Plants were regularly watered and kept under intermediate light conditions (10% light), protected from rainfall and herbivory in a screenhouse. The position of species and treatments was randomized. Per species we selected 60 healthy seedlings of similar size (see electronic supplementary material, table S3), i.e. 30 seedlings per species and Si treatment, for transplantation to the field.

(c) . Drought treatment

We transplanted the seedlings without the experimental substrates (i.e. with bare roots) to 30 plots (60 × 60 cm) in the forest understorey, and monitored their performance over the dry season. Plots were covered with transparent rain-out shelters (1.5 × 1.5 m) to exclude occasional rainfall and protected from vertebrate herbivores by wire mesh (1.2 × 1.2 cm mesh size). We did not implement an irrigated control treatment because previous studies in the area showed that most seedling mortality in the dry season is due to direct effects of drought [18]. Plots were established in an area of about 100 × 100 m, with a minimum distance of 5 m between them. One individual per Si treatment and species was transplanted to each plot, with individuals randomly assigned to positions in a 4 × 4 uniform grid. Individuals that did not survive transplantation (two Lacistema, one Herrania) were not considered further. On average 7.1% light reached the plots (assessed with hemispherical photographs). Seedlings were transplanted at the end of the wet season in December 2021, and regularly watered for about three weeks to allow for establishment. Seedlings were then kept unwatered under dry season conditions for 18 weeks.

(d) . Assessment of growth, survival and wilting stages

Visual wilting stages of the seedlings, which reflect leaf water potentials [28] were monitored (electronic supplementary material, table S4). Growth was assessed as leaf area change after the drought relative to the beginning. Before the onset of the drought leaf area was calculated from leaf width and length measurements [26]. Leaf area of the seedlings remaining after the drought was measured with a leaf area meter (Li-3100C, LICOR, Bad Homburg, Germany). The proportion of desiccated leaf area was estimated and subtracted from the measurements. Seedling survival was examined weekly based on aboveground living biomass, which is indicative of whole plant survival in the area [26]. Before the onset of drought seedling height was measured.

(e) . Leaf Si concentrations

Leaf Si concentrations were obtained after the screenhouse and the field phase (both n = 5) following Schaller et al. [22]. Measurements after the screenhouse phase were done on extra seedlings not used in the field experiment. For one species (Laetia) leaf Si data were not obtained due to insufficient biomass. Si fertilization significantly increased leaf Si concentrations in five of the six species for which leaf Si was measured (electronic supplementary material, table S2). The effect sizes (maximum twofold increase) were within the typical range of leaf Si responses to changing Si availability of tropical trees [23].

(f) . Data analysis

To test the effect of the Si treatment on drought survival for each species we ran Cox proportional hazard (COXPH) models. We calculated the change in risk of dying during the experimental period for the +Si seedlings compared to the −Si seedlings following Landes et al. [29]. We checked the proportional hazard assumption. To test the effect of the Si treatment on growth for each species we ran generalized least square (GLS) models. If necessary we included variance structures allowing for heterogeneous residuals in untransformed data and selected the best fit using AIC. We ran graphical model diagnostics to ensure normal, homogeneous and independent residuals. In all models Si treatment and initial height were the fixed effect and covariate, respectively. Plots were included as random effect and cluster in the GLS and COXPH models, respectively. We also tested, within species, whether the Si treatment influenced the time until the seedlings' worst wilted leaf reached at least the second, third and fourth wilting stage using Kruskal–Wallis tests. Statistical analyses were performed in R version 4.2.2 [30], using the packages ‘nlme’ [31] and ‘survival’ [32].

3. Results

The drought treatment led to wilting in all species (electronic supplementary material, figure S1), indicating that they experienced low leaf water potentials and drought stress [28]. Within all species, +Si and −Si seedlings reached the different wilting stages (i.e. slightly wilted, strongly wilted and leaf necrosis) at similar points into the drought (p > 0.39, H = 0.002–3.888). Seedlings of all but one species (Gustavia) showed necrosis and aboveground death.

Si addition did not influence drought survival in six of the seven species we studied (figure 1a and table 1a), suggesting elevated plant Si did not improve drought resistance. In one species (Ocotea) +Si seedlings had lower survival and about double the risk of mortality during the drought compared to −Si seedlings, opposite to our expectation. Furthermore, in most species, Si addition did not influence growth over the drought period (figure 1b, table 1b). We found weak evidence for a positive effect of elevated leaf Si on growth in only one species (Herrania).

Figure 1.

Difference in (a) survival probabilities and (b) growth (leaf area change) at the end of an 18 week drought treatment between seedlings raised in Si-rich (yellow) and Si-poor (blue) soils prior to the drought. Shown are means and 95% confidence intervals for five and six species in (a) and (b), respectively. In (a) results of Cox proportional hazard models are given by filled (p < 0.05) and empty (p > 0.05) circles. In (b) results of generalized least square models are given by empty (p > 0.05) circles. For species names see electronic supplementary material, table S1.

Table 1.

Effect of Si fertilization on subsequent (a) drought survival and (b) growth of seedlings of seven tropical tree species. In (a) results of Cox proportional hazard models are given and in (b) results of generalized least square models (*p < 0.05). Please note that in (a) a positive sign for β indicates an increase in mortality risk (i.e. lower survival). Number of replicates (N) are given. For species names see electronic supplementary material, table S1.

a)	+Si					height				N
	β	SE	z	p	change in risk (%)	β	SE	z	p
CALOLO	−0.043	0.345	−0.191	0.849	−4.170	0.001	0.003	0.296	0.767	30
HERRPU	−0.617	0.628	−1.031	0.303	−46.063	0.034	0.024	1.556	0.120	29
LACIAG	0.174	0.344	0.472	0.637	18.961	−0.041	0.022	−1.512	0.131	27
LAETTH	0.523	0.504	1.017	0.309	68.680	−0.009	0.025	−0.482	0.630	30
OCOTWH	0.702	0.341	2.566	0.010*	101.820	−0.003	0.003	−1.746	0.081	30
THEVAH					a
GUSTSU					a
b)
	β	SE	t	p		β	SE	t	p
CALOLO	0.044	0.055	0.806	0.457		−0.000	0.001	−0.390	0.713	7
HERRPU	0.137	0.065	2.106	0.054		0.004	0.004	0.961	0.353	16
LACIAG					b
LAETTH	−0.160	0.174	−0.922	0.374		−0.014	0.009	−1.632	0.127	15
OCOTWH	−0.432	0.551	−0.785	0.490		−0.001	0.002	−0.377	0.731	5
THEVAH	0.053	0.055	0.955	0.348		−0.001	0.001	−1.435	0.163	29
GUSTSU	−0.005	0.022	−0.236	0.815		0.000	0.000	0.938	0.356	29

^aModels for Thevetia and Gustavia were discarded because their mortality was near zero and caused singular fits.

^bSample size too small.

4. Discussion

No positive effect of elevated leaf Si on seedling growth and survival under drought was found in the seven tropical tree species we studied, not consistent with our hypothesis that Si should improve their drought resistance and performance. In one species +Si seedlings even had lower drought survival than −Si seedlings, opposite to our expectation. A trend towards better growth in +Si seedlings emerged in another species. By manipulating leaf Si prior to the field drought experiment, we were able to focus on direct effects of leaf Si and exclude potential indirect effects of Si addition on soil water holding capacity [20].

To our knowledge we were the first to experimentally test whether Si improves drought performance in wild tropical tree species and potentially wild plant species in general. However, our results are at odds with most previous studies on cultivated plants, mostly grasses, showing improved growth in Si-enriched plants under drought [11,12]. Initial transient positive effects of Si in the 18-week field experiment can be ruled out, since the progression of wilting, indicative of leaf water potentials [28], did not differ between treatments. Furthermore, in a separate 10-week drought experiment with five of our species and standardized nutrient and light conditions in the screenhouse also no effect of elevated leaf Si on survival emerged (electronic supplementary material, table S5, method S1). In our field experiment additional environmental factors might have overruled the positive effects of Si on seedling water status and drought performance [18]. Overall, our data suggest that Si accumulation is not an ecologically relevant drought resistance strategy in wild tropical tree species.

Previous studies on droughted grasses demonstrated that elevated leaf Si was associated with reduced stomatal and cuticular conductance, higher root weight ratio and specific leaf area, and improved plant drought resistance [12,14,33]. However, in the screenhouse experiment (see above), these parameters did not differ between −Si and +Si seedlings. This could be because most trees take up less Si than grasses [34]. At lower leaf Si concentrations these mechanisms, e.g. deposition of sub-epidermal phytoliths reducing cuticular conductance, increased sensitivity of stomatal guard cells reducing transpiration and/or modulations of molecular mechanism improving stress resistance [14,15], may not occur.

Unexpectedly, in one of our species +Si seedlings showed lower drought survival than −Si seedlings. This result implies that seedling drought performance of some species may be reduced in forest sites where abiotic conditions lead to elevated leaf Si concentrations. In fact, factors driving leaf Si concentrations, such as soil Si, water and nutrient availability [23–25], show pronounced spatial gradients in tropical forests [4,22,35,36]. This may add another, previously unexplored, dimension to soil-related variation of seedling performance, which influences species regeneration success, interactions and distributions in tropical forests [37].

Although elevated leaf Si may not play a beneficial role in tropical trees under drought, soil Si might still be important. Amorphous soil Si varies substantially at local and regional scale in tropical forests [22] and can enhance soil water availability [19–21]. Sites with higher amorphous Si might therefore offer soil water conditions in which drought-sensitive species could survive. Furthermore, a recent study showed that even in rice, which is probably one of the best studied species with respect to Si effects on drought resistance, improved drought performance with Si addition was not due to plant Si uptake, but indirect soil effects [19].

The adverse effects of elevated leaf Si we found in one of our species and possible indirect soil Si effects on water availability suggest that the ecological relevance of Si for seedling drought performance, including effects on species distribution, warrants further research in tropical forests. Furthermore, based on our results, we recommend that the numerous findings of Si-mediated improvements of drought resistance in cultivated species should not be generalized to wild plants in natural systems.

Ethics

Research and collection permits were obtained from MiAmbiente (permit number SE/AP-23-19).

Data accessibility

Data, R code and description of the data are available from the Dryad Digital Repository: https://doi.org/doi:10.5061/dryad.vdncjsz1g [38].

Supporting information are available as electronic supplementary material [39].

Declaration of AI use

We have not used AI-assisted technologies in creating this article.

Authors' contributions

M.K.: conceptualization, formal analysis, investigation, methodology, project administration, visualization, writing—original draft, writing—review and editing; J.S.: conceptualization, funding acquisition, methodology, project administration, supervision, writing—review and editing; A.M.K.: formal analysis, investigation, visualization, writing—original draft, writing—review and editing; B.C.: investigation, methodology, resources, writing—review and editing; B.M.J.E.: conceptualization, funding acquisition, methodology, project administration, supervision, writing—review and editing.

All authors gave final approval for publication and agreed to be held accountable for the work performed therein.

Conflict of interest declaration

We declare we have no competing interests.

Funding

The study was funded by the German Research Foundation (EN 699/5-1, SCHA 1822/15-1).