Odor Communication with Green Leaf Volatiles for Stress Signalling in the Internet of Plants

TL;DR

Numerical results show that (Z)-3-hexenol is the primary driver of the system and that plant perception generally operates in a non-linear region, which provides a framework for understanding the evolution of plant-plant communication and for developing next-generation precision farming technologies.

Abstract

This paper develops an end-to-end odor communication model for stress signaling between plants using Green Leaf Volatiles (GLV). A damaged transmitter plant emits (Z)-3-hexenal, (Z)-3-hexenol, and (Z)-3-hexenyl acetate, which propagate through a time-varying diffusion-advection channel and undergo multiplicative loss at the receiver. The sink plant is modeled with a biochemical receiver network that converts the received GLVs into the defensive metabolite (Z)-3-hexenyl $β$-vicianoside, and an alarm decision is defined based on its concentration level. Numerical results show that (Z)-3-hexenol is the primary driver of the system and that plant perception generally operates in a non-linear region. These findings provide a framework for understanding the evolution of plant-plant communication and for developing next-generation precision farming technologies.

Figures (8)

• Figure 1: End-to-end odor communication model used in this work. A damaged plant emits three GLVs (HAL, HOL, and HAC) that propagate through a time-varying diffusion–advection channel. The received air concentrations are subject to multiplicative loss and drive a receiver-side biochemical network (HAL/HAC conversion to HOL and downstream conversion to HEXVic), whose output $c_v(t)$ is used for the alarm decision BioRender_Merdan_2026_3.
• Figure 2: Receiver nonlinearity results across various simulation times, input signal configurations, and scaling scenarios. For (i) constant inputs \ref{['LinearityPilot_SingleVOC_HAL_mode_1']} scales HAL only, \ref{['LinearityPilot_SingleVOC_HAC_mode_1']} scales HAC only, \ref{['LinearityPilot_SingleVOC_HOL_mode_1']} scales HOL only, and \ref{['LinearityPilot_AllVOC_mode_1']} scales all three input amplitudes together, and the horizontal axis shows the common scaling factor (SF). For (ii) single initial pulses, \ref{['LinearityPilot_SingleVOC_HAL_mode_2']} scales HAL only, \ref{['LinearityPilot_SingleVOC_HAC_mode_2']} scales HAC only, \ref{['LinearityPilot_SingleVOC_HOL_mode_2']} scales HOL only, and \ref{['LinearityPilot_AllVOC_mode_2']} scales all three input amplitudes together. For (iii) periodic on-off cycling, \ref{['LinearityPilot_SingleVOC_HAL_mode_3']} scales HAL only, \ref{['LinearityPilot_SingleVOC_HAC_mode_3']} scales HAC only, \ref{['LinearityPilot_SingleVOC_HOL_mode_3']} scales HOL only, and \ref{['LinearityPilot_AllVOC_mode_3']} scales all three input amplitudes together. The horizontal dashed lines indicate the $0.02$ threshold used to define the LTI operating region.
• Figure 3: Frequency-domain response of the HOL output to different odor inputs. The top row shows the magnitude responses for HAL, HOL, and HAC inputs in \ref{['freq_response_input_hal_to_HOL_magnitude']}, \ref{['freq_response_input_hol_to_HOL_magnitude']}, and \ref{['freq_response_input_hac_to_HOL_magnitude']}, respectively. The bottom row presents the corresponding phase responses in \ref{['freq_response_input_hal_to_HOL_phase']}, \ref{['freq_response_input_hol_to_HOL_phase']}, and \ref{['freq_response_input_hac_to_HOL_phase']}.
• Figure 4: Sensitivity analysis of the linear operating region with respect to enzyme abundances. The horizontal and vertical axes show the scaling factors for each enzyme abundance.
• Figure 5: $1$-transmitter $1$-receiver simulation results. The channel output air concentrations are shown in \ref{['fig_channel_rx_air_1tx1rx']}. The internal leaf concentrations are shown for \ref{['fig_rx_c_a_hal_1tx1rx']} HAL, \ref{['fig_rx_c_t_hac_1tx1rx']} HAC, \ref{['fig_rx_c_o_hol_1tx1rx']} HOL, \ref{['fig_rx_c_g_hexglc_1tx1rx']} HEXGlc, and \ref{['fig_rx_c_v_hexvic_1tx1rx']} HEXVic. The red dashed lines indicate the linearity limit for that molecule, and the black dashed line indicates the alarm decision threshold.