Introduction
Leaf hairs, formally known as trichomes, are epidermal appendages on average present in >50% of species in given ecosystems and varying widely in morphology among and even within species (Perez-Estrada et al., 2000). Leaf hairs are thought to provide several functional roles. They provide mechanical protection against phytophagous insects (Levin, 1973; Dalin et al., 2008), and some plant species can produce leaf hairs with glandular secretion which also serve a defensive purpose (Werker, 2000). Leaf hairs can also provide protection from disease pressures such as downy mildew (Kortekamp and Zyprian, 1999). Leaf hairs also influence energy balance and gas exchange. The measured effect of hairs on leaf temperature varies, from warming up to several degrees C (Woolley, 1964; Ehleringer and Mooney, 1978; Ripley et al., 1999; Peng et al., 2015) to cooling in summer and warming in winter (Ehleringer and Mooney 1975). The overall effect of hairs on water use efficiency (WUE, the ratio of net photosynthetic rate to transpiration rate) differs across studies but is most commonly reported to be positive (Ehleringer and Mooney, 1978; Ripley et al., 1999). The influence of trichomes on leaf temperature and gas exchange arises from effects on reflectivity and the boundary layer. Trichomes can lower leaf absorptance to visible light (Wuenscher, 1970a; Holmes and Keiller, 2002) and provide protection from photoinhibition, ultraviolet radiation and overheating (Ehleringer and Mooney, 1978; Karabourniotis et al., 1992; Karabourniotis et al., 1995; Ntefidou and Manetas, 1996).
Trichomes are also thought to influence the diffusive exchange of heat and mass across the leaf boundary layer – that is, the zone adjacent to the leaf surface where wind speed is substantially reduced by friction with the leaf surface, and in which transport is therefore dominated by diffusion rather than by advection. Because diffusion is much slower than advection at macroscopic scales (e.g., over distances > ~21 µm at a wind speed of just 1 m s-1), the boundary layer slows heat and gas exchange between the leaf and air (Raschke, 1960; Grace and Wilson, 1976). It is most often assumed that hairs increase the effective thickness of the boundary layer by entraining a layer of relatively still air, and thereby impede heat and mass transport, and that boundary layer thickness relates to the height of the trichome canopy (Wuenscher, 1970a; Ehleringer and Mooney, 1978; Nobel, 2005). However, some have speculated the converse, on the premise that hairs break up air flow and thereby promote turbulence (Wolpert, 1962; Schreuder et al., 2001).
These contrary theoretical speculations on the influence of leaf trichomes on the boundary layer, and potentially on gas exchange, have have not been adequately tested experimentally. This may be partly because leaf boundary layer conductance is perceived to be very large relative to stomatal conductance (Nobel, 2005; Jones, 2014) and thus relatively unimportant. Indeed, many models assume leaf and air temperatures are always equal (Dong et al., 2017), which is equivalent to assuming boundary layer conductance is infinite (resistance is zero), so that leaf temperature is strongly coupled with air temperature. Yet many studies have reported large leaf-air temperature differencesin situ (Leigh et al., 2017; Fauset et al., 2018; Guha et al., 2018; Xu et al., 2020): e.g., ΔT (leaf minus air) = +5.1 ± 2.1oC (mean ± SD) for 68 Proteaceae species (Leigh et al., 2017); –6 to +7 oC for three woody species in a tropical dryland (Dong et al., 2017); +3.7 to + 4.7 oC in wheat (Ehrler et al., 1978; Ayeneh et al., 2002); –4.1 to +4.1oC in almond canopies (Gonzalez-Dugo et al., 2012). Leaf-air temperature difference is strongly influenced by boundary layer conductance because it directly influences both convective heat exchange and transpiration (and thus latent cooling).
Full resolution of precisely how trichomes affect the boundary layer has not been achieved, because of the technical difficulty of estimating boundary layer conductance experimentally in real leaves, rather than in physical models of leaves. It is common, for example, to estimate boundary layer conductance to water vapor in a gas exchange chamber using a wet filter paper or other wet fabric surface, to represent a leaf without an epidermis and thus without stomatal resistance (Monteith and Unsworth, 2013). Another approach, based on using rigid, heated metal plates, contributed to the validation of phenomenological equations relating g bh to wind speed and to leaf size (specifically, the characteristic dimension related to downwind width) (Gates, 1968; Nobel, 1975; Schuepp, 1993; Brenner and Jarvis, 1995; Schlichting and Gersten, 2017). Another set of methods is based on surface treating leaf surfaces with water or a water-impermeable substance to modulate the contribution of evaporation to energy balance (Impens, 1966; Impens et al., 1967; Linacre, 1972; De Parcevaux and Perrier, 1973; Paw U and Daughtry, 1984). While potentially powerful, each of those approaches is difficult to apply to the specific question of how leaf hairs affect boundary layer conductance. The most direct experimental test of this question was performed by Wuenscher (1970), who examined how the removal of leaf hairs affected transpiration rate in Verbascum thapsus . Although Wuenscher concluded that hairs reduced boundary layer conductance, later reanalysis by Parkhurst (1976) suggested that cuticular damage caused by shaving led to enhanced transpiration and evaporative cooling, which confounded Wuenscher’s results.
Dynamic thermography – heating a leaf and then measuring its rate of cooling by convection (Jones, 2014) – can quantify boundary layer conductance independent of the effect of hairs on light absorption. Dynamic thermography using infrared cameras has recently been used successfully to quantify both boundary layer conductance (Albrecht et al., 2020) and stomatal conductance (Vialet-Chabrand and Lawson, 2019) in intact leaves. Here we adapt and extend the approach of Albrechtet al . (2020) to quantify the influence of leaf hairs on boundary layer conductance to heat (g bh; see Table 1 for a list of symbols) in five broadleaved species, by measuring the time constant for leaf cooling (τ ) after transient radiative heating in leaf patches with hairs and without hairs and then inferringg bh from τ using a dynamic energy balance model, accounting for stomatal conductances measured in each patch. We hypothesized that leaf hairs will add a component of transport resistance that is proportional to the depth of the hair layer, thereby reducing g bh compared to an otherwise similar leaf surface without leaf hairs.