Rinorea Aublet (Violaceae) is a pantropical genus of forest shrubs and trees. It is the second most species-rich genus in the family after Viola L., with an estimated total of 225–275 species throughout the tropics (Wahlert and Ballard 2012). In the Malesian region, only 11 species are recognized in the genus, with four species attributed to the Philippines (Jacobs and Moore 1971). However, the very broad circumscriptions of the widespread Rinorea bengalensis (Wall.) Kuntze and Rinorea javanica (Blume) Kuntze in the taxonomic revision of Jacobs and Moore (1971) needs a closer re-examination. A few new taxa from Borneo have subsequently been added for the region (e.g. Forman and Ahmad 1996, Jarvie and Stevens 1998, Stevens 2000).

In Rinorea, at least three species are known to hyperaccumulate the heavy metal nickel. Rinorea bengalensis (Wall.) Kuntze was the first nickel hyperaccumulator species of Rinorea discovered with up to 17, 500 µg g^-1 (dry weight) based on herbarium specimens from throughout Southeast Asia, including the Philippines (Brooks and Wither 1977). A subsequent analysis of herbarium material of 70 other species of Rinorea from Central and South America, Africa, and Asia also revealed another species, Rinorea javanica (Blume) Kuntze, as a nickel hyperaccumulator with up to 2, 170 µg g^-1 in its leaf tissues (Brooks et al. 1977). More recently, Proctor et al. (1994) reported another, yet unnamed, nickel-hyperaccumulating species of Rinorea from Mt Piapi on Karakelong Island, northeast of Sulawesi in Indonesia with up to 1, 830 µg g^-1 foliar Ni.

The Violaceae also includes two other genera that hyperaccumulate nickel. In Agatea A.Gray, one species from New Caledonia, Agatea longipedicellata (Baker f.) Guillaumin & Thorne, has been recorded to accumulate up to 2, 500 µg g^-1 of nickel in its foliar tissues (Jaffré 1980, Boyd and Jaffré 2009). In Hybanthus Jacq., at least six taxa from New Caledonia, Western Australia, and Sri Lanka are also known nickel hyperaccumulators (Reeves 2006) with maximum recorded foliar nickel levels of 1, 860 µg g^-1 in Hybanthus enneaspermus (L.) F.Muell. (Rajakaruna and Bohm 2002) to 25, 500 µg g^-1 in Hybanthus austrocaledonicus Melch. (Brooks 1994, Reeves 2003).

The ability to absorb certain metals and metalloids (chemical elements with properties in between those of metals and non-metals, also referred to as semi-metals) from the soil and to accumulate them in shoot tissues in exceptionally high and normally toxic concentrations without any evidence of physiological stress is rather rare among plants (Baker and Brooks 1989; Reeves and Baker 2000; Kramer 2010). Metal hyperaccumulation has recently been suggested to have had multiple origins within the angiosperms (Cappa and Pilon-Smits 2013). The more than 500 plant taxa thus far recorded as metal hyperaccumulators represent only a very small portion of all known angiosperms (Reeves and Baker 2000, Kramer 2010, van der Ent et al. 2012). The largest number of species, approximately 450, distributed in a wide range of angiosperm families, hyperaccumulate the metal nickel and generally occur on serpentine or ultramafic soils (van der Ent et al. 2012, Pollard et al. 2014).

Apart from their unusual and interesting ecology and physiology, hyperaccumulator plants have received considerable attention owing to the possibility of exploiting their accumulation traits for practical applications, especially in the development of so-called environmentally green technologies, e.g. phytoextraction, phytoremediation of heavy metal in contaminated soils, or phytomining to recover commercially valuable metals in plant shoots from mineralized sites (Chaney et al. 1997, Brooks and Robinson 1998, McGrath and Zhao 2003, Reeves 2003, Pilon-Smits 2005, Rascio and Navari-Izzo 2011).

In the Philippines, much of the forest flora on ultramafic or serpentine soils (Fernando et al. 2008) remain underexplored. Field surveys in a number of sites in the archipelago have revealed some new species (e.g. Hoffmann et al. 2003, Fernando and Rodda 2013), including several species that are able to accumulate heavy metals in their above-ground tissues (Baker et al. 1992, Hoffmann et al. 2003, Fernando et al. 2013, Gotera et al. 2014). In this paper, we describe a new species of Rinorea discovered in remnant forest on ultramafic soils that is also a nickel-hyperaccumulator. This species is, thus far, known only from small populations in the northern section of Zambales Province on Luzon Island in the Philippines. This area is part of the Zambales Ophiolite Complex (Rossman et al. 1989, Zhou et al. 2000, Yumul 2004) which is host to several metallic mineral deposits (e.g. chromium, nickel) (Osberger et al. 1988, Bacuta et al. 1990, Yumul et al. 2003).

The morphology of the species presented here was based on field, vegetative, and reproductive characters. Field characters were recorded on site. Vegetative characters were observed and measured from press-dried specimens and seedlings and reproductive characters from fresh specimens and from material preserved in 70% ethanol. Detailed morphological measurements were made using digital calipers and a calibrated eye piece under a dissecting microscope. Herbarium specimens were also consulted and compared at CAHUP, LBC, PNH, and PUH, including additional material, e.g. images of type specimens of Southeast Asian and Philippine Rinorea available online at BISH, K, L, MO, NY, and US. All photographs, except where indicated, were taken in the field in the natural habitat of the species. Conservation threat assessment follows IUCN Categories and Criteria (IUCN 2012).

Field semi-quantitative screening for nickel accumulation in this species was performed on site on leaf samples, thoroughly washed in distilled water, crushed in a mortar and pestle, and tested on filter paper previously impregnated with 1% of the nickel-specific colorimetric reagent, dimethylglyoxime, dissolved in 95% ethanol (Baker et al. 1992; Reeves et al. 1996, 1999). Formation of pink or magenta color indicated exceptionally high (above 1, 000 µg g^-1) concentration of Ni in the dry plant matter. Tissue samples of roots, stems and leaves, and of soil from the rhizosphere (c. 30–100 cm) of each plant sampled were also collected. These were subsequently subjected to laboratory elemental analyses for nickel (Ni) and two other heavy metals, copper (Cu) and cobalt (Co). The plant samples were thoroughly washed in distilled water and then oven-dried at 60 °C. Each sample was later weighed into borosilicate test tubes and ashed in a muffle furnace for 4–5 hours, with the final temperature of 500 °C being maintained for the last 2 hours. The ash was then taken up in 5 ml of warm 2 M HCl and the digest finally made up to an appropriate volume (5–20 ml) then analyzed for Ni, Cu, and Co content using atomic absorption spectrophotometer. The soil samples were digested with aqua regia (3:1 concentrated hydrochloric acid: nitric acid), then diluted appropriately for metal analyses of Ni, Cu, and Co using an atomic absorption spectrophotometer. Details of this method follow in general that described by Reeves et al. (1996).