Figure 5 . The feedback loop of plant response to complete submergence. Here we consider floodwater with high turbidity so that the light intensity for underwater photosynthesis is limited. The general plant responses shown in black arrows deteriorate plant oxygen status during complete submergence and may even lead to carbon starvation. The “escape” strategy (pink box and pink arrows) imposes an overall negative feedback loop on shoot oxygen but an overall positive feedback loop on carbohydrate and energy reserve, while “quiescence” strategy (purple box and purple arrows) imposes an overall negative feedback loop on carbohydrate and energy reserve. Depending on the species and ecotype, plants can either self-rescue through “escape” strategy or self-preserve through “quiescence” strategy.
Discussion and concluding remarks
Our proposed model framework considers the Soil-Plant-Atmosphere-Continuum as a “backbone”, coupling hormone-mediated flooding-response processes to extend the theoretical framework to the flooding scenarios, including waterlogging and complete submergence. Efforts of integrating flooding response to plant hydraulic and photosynthetic model have been made previously. Compared to these efforts, e.g. , the model developed by Feddes (1982) that phenomenologically simulated stomatal activity in response to soil water content from drought to waterlogging and extended by more process-based knowledge on oxygen stress (e.g. Bartholomeus, 2009), our proposed model framework attempts to step further by introducing biological mechanistic processes under both drought and flooding stress. These processes include oxygen dynamics in the root and shoot and consequences for metabolism in these organs. Also, we include hormonal signaling processes, in which the central hormone ethylene accumulates within a couple of hours after waterlogging or submergence, prior to occurrence of plant oxygen deficit. The hormonal signaling processes resulting in either “escape” or “quiescence” impose a negative feedback loop upon the main skeleton or abate the positive feedback loop of the main skeleton, indicating a self-rescuing mechanism or self-preservation mechanism, respectively.
For simplicity, plants in the proposed framework were assumed to be mature. Except for adventitious root development and shoot elongation that serve as specialized flooding-response morphological developments, the growth of root and shoot is currently not considered. Furthermore, we included a simplified set of hormonal signaling processes. The plant stress response signaling network involved is highly complex, involving a diverse set of hormones, secondary messengers, and genes, between which there is a complex network of crosstalk (Sasidharan et al., 2018). Classically the major phytohormones are subdivided into two categories, the stress response hormones ABA, ethylene, salicylic acid (SA), and jasmonate (JA), and the growth hormones gibberellin, cytokinin, and auxin (Verma et al., 2016). Phenotypically plastic stress responses often arise through the crosstalk between these two categories of hormones. For instance, auxins play a key role in root development, and their transport and signaling is affected by ABA under drought (Rock & Sun, 2005), while under flooding stress, ethylene directly and/or indirectly affects gibberellin signaling, triggering “escape” or “quiescence” strategies, respectively (Bashar et al., 2019). Here we propose to model the dynamics of the stress response hormones and secondary messengers explicitly, incorporating their effects on auxin, cytokinin, and gibberellin only implicitly. Specifically, we propose to incorporate ABA, ethylene, and ROS, since the stress hormones JA and SA mainly function against biotic stresses (Verma et al., 2016).
Given the aforementioned simplifications, we suggest that the proposed framework can be used to model short-term (i.e. , days to a couple of weeks) responses to drought and flooding stress. A key hypothesis we propose this framework can test is that prior exposure to water stress alters future stress responses through the occurrence of a “memory”, which can be formed through both plant morphological changes and preconditioning of hormone levels. For instance, when soil water conditions shift from waterlogging to drought, aerenchyma formed during waterlogging still exist inside the plant and are expected to impact water transport (Van der Weele et al., 1996; Yang et al., 2012). Importantly, hormonal and physiological crosstalk is highly tissue and species specific. As an example, it is found that an ethylene pre-treatment enhances ROS scavenging capacity and thereby tolerance to ROS mediated oxidative stress in roots (Peng et al., 2014). On the other hand, in aerenchyma forming species, ethylene reduces ROS scavenging to enhance ROS levels and induce aerenchyma formation in specific cell types (Steffens et al., 2011). It is exactly this complexity that makes predicting plant responses and their potential dependence on prior conditions soil moisture conditions across the full spectrum from drought to flooding impossible without modeling the processes presented in our mechanistic framework.
A possible future extension of the mechanistic framework is to incorporate the life history of plants and the changes in plant architecture and physiology this results in, thus enabling the simulation at longer time scales, say months to years or even decades. Besides this extension on the temporal scale, we also envision straightforward extensions one the spatial scale. This proposed framework can serve to represent individuals within individual-based models that scale up to the community and ecosystem level. Incorporating only horizontal water transport this would already enable one to investigate competition for water and the impact of species composition and distribution on ecosystem water stress responses. Additionally, these models can be extended to incorporate, e.g. , interplant differences in their efficiency of nutrient and light acquisition and their competition for these resources.
A natural application for our model framework is its incorporation into land-surface models. So far, the incorporation of plant hydraulics into land-surface models has been used to investigate the feedback between climate and vegetation in terms of water cycle during non-stressed conditions and drought, yet mechanistic plant responses to excessive soil water content are still lacking in these models (Li et al., 2021; Nguyen et al., 2020). With climate change expected to enhance surface evaporation and atmospheric vapor accumulation, the intensity and duration of droughts as well as the chances of flooding are both increased, underlining the demand for a framework capable of integrating both drought and flooding responses of plants (Trenberth, 2011). Therefore, through incorporation of the model framework proposed here or simplifications thereof into larger-scale land-surface models we aim to contribute to the improvement of our understanding of how plants help shape the climate.
Acknowledgements
This work was supported by the funding from the Complex Systems Fund of Utrecht University, with special thanks to Dr. Peter Koeze.ORCID
Siluo Chen https://orcid.org/0000-0003-2152-9109
Kirsten H.W.J. ten Tusscherhttps://orcid.org/0000-0002-1945-7858
Rashmi Sasidharan https://orcid.org/0000-0002-6940-0657
Stefan C. Dekker https://orcid.org/0000-0001-7764-2464
Hugo J. de Boer https://orcid.org/0000-0002-6933-344X