SILICON AND OTHER ESSENTIAL ELEMENT COMPOSITION IN ROOTS USING X-RAY FLUORESCENCE SPECTROSCOPY: A HIGH THROUGHPUT APPROACH Ivan Hiltpold, Lanila Demarta, Scott N. Johnson, Ben D. Moore, Sally A. Power and Christopher Mitchell Hawkesbury Institute for the Environment, Western Sydney University, NSW, Australia Corresponding author: I. Hiltpold, ivan.hiltpold@gmail.com Keywords: Silicon, XRF, sample preparation, plant elemental composition, Introduction Silicon (Si) is the second most abundant element in soil, just after oxygen, making up 30% of the upper continental crust (w/w, ca. 146 times more abundant than carbon) (Hans Wedepohl, 1995). On a day-to-day basis, Si is used in countless households as building materials, in all electronic devices and even cosmetics (Vasanthi et al., 2012). This element is not essential only to human societies, but is also used by a large range of living organisms, including plants (Cooke and Leishman, 2011). Whereas Si concentrations vary considerably among plants, trace amounts of this element have been found in all tested species (Hodson et al., 2005). Intraspecifc variation between cultivars has also been documented (Deren, 2001). Poaceae have been shown to accumulate relatively large amounts of Si (Hodson et al., 2005), but rice, Oryza sativa L., has the highest Si concentrations, representing up to 10% of its dry mass (Epstein, 1999). Silicon contributes to plant strength, providing grazing resistance to stems and leaves (Ando et al., 2002; Osborne, 2008). Below-ground, silicon aids Distichlis spicata L. root penetration in the soil matrix (Hansen et al., 1976). Accumulation of Si in plant tissue is energetically inexpensive compared to carbon fxation (energy costs of Si vs. carbon 1:10- 1:20) (Raven, 1983). This could have been at the origin of the diversifcation of Poales during the Miocene, when atmospheric CO 2 levels were low (Craine, 2009) therefore favoring plants able to accumulate high levels of Si to substitute carbon in some essential functions, such as mitigation of abiotic stresses (e.g. Epstein, 1999; Currie and Perry, 2007) and defenses against herbivores (e.g. Reynolds et al., 2009). There is increasing evidence that Si accumulated in plants plays a role in physical and chemical defenses against herbivores. Massey and Hartley (2009) demonstrated that the performance of the herbivore Spodoptera exempta Walker was reduced when the insect was ofered grass species grown under Si-rich, compared to low Si conditions. The authors proposed that Si reduces the digestibility of the consumed plant tissues and that Si phytoliths cause wear to the insect mandibles, eventually impacting its performance (Massey and Hartley, 2009). Solubilized Si appears to be linked to the biochemistry of several plant secondary metabolites involved in plant defenses (Datnof et al., 2007). Additionally, grass leaf siliceousness (i.e. silica content) impacts the behavior of certain insect herbivores. For example, the white fy Bemisia tabaci Gennadius preferentially lays eggs on low-Si leaves, probably to ensure reliable food resources to its progeny (Correa et al., 2005). In plant roots, Si has also been found to mitigate the efects of abiotic stresses such as salinity and heavy metals (e.g. Kim et al., 2014; Vaculíková et al., 2014; Fialová et al., 2016) and improves resistance to pathogens (e.g. Cherif et al., 1994; Safari et al., 2012; Fortunato et al., 2014). Furthermore, as is seen with insect herbivores, Si accumulation in roots can afect rodent populations, via reduced digestibility and increased wearing) (Wieczorek et al., 2015a; Wieczorek et al., 2015b). To date, there is only one example of a demonstrated impact of Si on insect root herbivores (Frew et al., 2016). Typically, elemental analyses in plants are based on digestion protocols, that are both time consuming and potentially hazardous. Reidinger et al. (2012) reported a rapid, accurate and relatively cheap method for analyzing Si and phosphorus (P) in plant material using an X-ray fuorescence spectrometer (XRF). Here, we follow up this earlier work, presenting a protocol for analyzing the elemental composition of roots using XRF techniques. Because root material can be very scarce and laborious to collect, we have also tested various sample preparation methods to reduce the quantity of plant material required for the analysis. Materials and Methods All measurements were carried out on a PANalytical Epsilon 3 EDXRF spectrometer. Silicon was measured at a tube voltage of 5 kV, and a current of 60 μA in the presence of helium. Silicon standards were prepared by mixing SiO 2 with methyl cellulose (oven dried at 60°C overnight) at a range of Si concentrations (0.5%, 1%, 2%, 4%, 6%, 8%, 10%) and milling for 2 minutes at 30 Hz. Standards were measured as either big pellets (2 g of material, 33 mm diameter), small pellets (300 mg material, 13 mm diameter) or as a loose powder in a small mass holder (100 mg), and a standard curve was produced. To validate this standard curve, Si standards were run alongside several certifed plant samples, to produce a multi-element calibration allowing for simultaneous measurement of a range of elements (table 2). Of these standards, three included Si values (NJV 94-4 Energy grass 2.1%, IPE 132 Broccoli 0.2%, NCS ZC73018 Citrus leaves 0.41%). Calibration ranges and correlation coefcients are given in Table 3. Tests with root material To evaluate if the XRF protocols described above can accurately measure the elemental composition of root material, we randomly selected seven root samples of long rotation Italian ryegrass (Lolium multiflorum Lam.) from amongst a larger number of samples harvested during a feld experiment in 2016. Briefy, ryegrass plants were grown in pots for 80 days (September to December) before soil was carefully washed from the root balls. Roots were oven-dried at 40°C for three days and ground in a mixer mill (Retsch MM400) for 3 min (30 Hz) and stored at -20°C until processed. The large quantity of root material allowed for four replicates of each measurement per sample. Root material can be laborious to collect and, once dried, only minute amounts could remain. Therefore, in an attempt to reduce the quantity of plant material used for this analysis, three sample preparation methods were assessed: 1) big pellets were pressed (10 t for 2 min, Manual Hydraulic westernsydney.edu.au 191 PLANT RESISTANCE TO PESTS – NON-DIGESTIVE ELEMENTAL ANALYSIS IN ROOTS