Summary• No study of a single nickel (Ni) hyperaccumulator species has investigated the impact of hyperaccumulation on herbivores representing a variety of feeding modes.• Streptanthus polygaloides plants were grown on high-or low-Ni soils and a series of no-choice and choice feeding experiments was conducted using eight arthropod herbivores. Herbivores used were two leaf-chewing folivores (the grasshopper Melanoplus femurrubrum and the lepidopteran Evergestis rimosalis ), a dipteran rhizovore (the cabbage maggot Delia radicum ), a xylem-feeder (the spittlebug Philaenus spumarius ), two phloem-feeders (the aphid, Lipaphis erysimi and the spidermite Trialeurodes vaporariorum ) and two cell-disruptors (the bug Lygus lineolaris and the whitefly Tetranychus urticae ).• Hyperaccumulated Ni significantly decreased survival of the leaf-chewers and rhizovore, and significantly reduced population growth of the whitefly cell-disruptor. However, vascular tissue-feeding insects were unaffected by hyperaccumulated Ni, as was the bug cell-disruptor.• We conclude that Ni can defend against tissue-chewing herbivores but is ineffective against vascular tissue-feeding herbivores. The effects of Ni on cell-disruptors varies, as a result of either variation of insect Ni sensitivity or the location of Ni in S. polygaloides cells and tissues.

The European plant Thlaspi caerulescens hyperaccumulates zinc in its leaves to concentrations exceeding 2%. This may represent a defense against herbivory, as shown by reduced feeding on plants grown in high-zinc versus low-zinc substrates. However, the more evolutionarily relevant comparison involves plants whose metal content differs due to genetic rather than environmental causes. In our study, plants were grown in a uniform hydroponic solution containing 10 mg/l Zn. Foliar Zn concentrations were measured, and plants of contrasting Zn concentration were chosen for feeding trials using Pieris napi oleracea larvae. Comparisons revealed no statistically significant preferences when young larvae were used, but highly significant avoidance of high-Zn leaves by later-instar caterpillars, suggesting that hyperaccumulation could evolve in response to the selective pressures of herbivory.

Plants that contain elevated foliar metal concentrations can be categorized as accumulators or, if the accumulation is extreme, hyperaccumulators. The defense hypothesis suggests that these plants may be defended against folivore attack, and recent research has indicated that metal concentrations at or below the accumulator range may be defensively effective. This experiment explored the toxicity of four metals hyper-accumulated by plants (Cd, Ni, Pb, and Zn) and asked if combinations of metals, or metals and organic chemicals, might broaden the defensive effectiveness of metals. Metals were used alone and in certain metal + metal (Zn plus Ni, Pb, or Cd) and metal + organic defensive chemical (Ni plus tannic acid, atropine, or nicotine) combinations. Artificial diet amended with these treatments was fed to larvae of the crucifer specialist herbivore Plutella xylostella. Combinations of metals and metals + organic chemicals significantly decreased survival and pupation rates, compared to single treatments, for at least some concentrations in every experiment. Effects of combinations were additive rather than synergistic or antagonistic. Because Zn enhanced the toxicity of other metals and Ni enhanced the toxicity of organic defensive chemicals, our findings suggest that the defensive effects of metals are more widespread among plants than previously believed. They also support the hypothesis that herbivore defense may have led to the evolution of metal hyper-accumulation by increasing the preexisting defensive effects of metals at accumulator levels in plants.

We determined the effectiveness of Ni as an elemental defence of Streptanthus polygaloides (Brassicaceae) against a crucifer specialist folivore, diamondback moth (DBM), Plutella xylostella. An oviposition experiment used arrays of S. polygaloides grown on Ni-amended (high-Ni) soil interspersed with plants grown on unamended (low-Ni) soil and eggs were allowed to hatch and larvae fed freely among plants in the arrays. We also explored oviposition preference by allowing moths to oviposit on foil sheets coated with high-or low-Ni plant extract. This was followed by an experiment using low-Ni plant extract to which varying amounts of Ni had been added and an experiment using sheets coated with sinigrin (allyl glucosinolate) as an oviposition stimulant. Diamondback moths laid 2.5-fold more eggs on low-Ni plants than on high-Ni plants and larval feeding was greater on low-Ni plants. High-Ni plants grew twice as tall, produced more leaves, and produced almost 3.5-fold more flowers. Low-Ni plants contained more allyl glucosinolate than high-Ni plants and moths preferred to oviposit on foil sheets dipped in low-Ni plant extract. Moths showed no preference when Ni concentration of low-Ni extract was varied and overwhelmingly preferred sinigrin coated sheets. We conclude that Ni hyperaccumulation is an effective elemental defence against this herbivore, increasing plant fitness through a combination of toxicity to DBM larvae and decreased oviposition by adults.

Tissues of most plant species contain < 10 µg Ni g −1 but Ni hyperaccumulators contain more than 1000 µg Ni g −1 . Hyperaccumulated Ni can defend plants from some herbivores but the defensive role of lesser Ni concentrations is little explored. We raised five species of Streptanthus (Brassicaceae) native to ultramafic soils, one of which (S. polygaloides) is a Ni hyperaccumulator whereas the others are simply Ni-tolerant, on Ni-amended and unamended greenhouse soils to create plants differing in Ni concentrations. On high-Ni soil, leaves of the hyperaccumulator contained 3800 µg Ni g −1 whereas leaves of non-hyperaccumulator species contained 41−64 µg Ni g −1 . Plants of all species grown on low-Ni soils had < 14 µg Ni g −1 . Slugs (Limax maximus) were fed plant material in no-choice tests over a 50-day period and survival and mass changes were recorded. All slugs fed high-Ni leaves of the hyperaccumulator species died within 21 d. Slugs fed high-Ni leaves of the other species did not differ significantly in survival or mass change from those fed low-Ni leaves. In choice tests, slugs (Lehmannia valentiana) offered both high-and low-Ni S. polygaloides leaves did little damage to high-Ni leaves. We conclude that hyperacumulated Ni can defend S. polygaloides from slug herbivory via both toxicity and deterrence, but these defensive effects do not extend to Streptanthus species containing < 70 µg Ni g −1 .