Lethal toxins in non-preferred foods: How plant chemical defences can drive microtine cycles

We hypothesize that periodic lethal toxin production by non-preferred food species can explain the precipitous decline phase of vole cycles in arctic and alpine tundra regions. For plants that cannot respond to grazing damage by compensatory shoot growth, periodic production of toxins can have an ad aptive advantage at the individual level. Several plants in the diet of cyc lical small mammals do produce lethal toxins and some production is known t o be cyclical. Despite the wealth of indirect and anecdotal observation in support of the hypothesis, there remains a lacuna in the hard core of evide nce: periodic production of lethal toxins and toxin-related deaths in micro tines. We argue that this is only because it has not been sought among like ly plants or has been sought at the wrong place or time. Strong candidate s pecies are non-preferred foods with circumpolar distributions such as Empet rum nigrum or Vaccinium uliginosum. The right place to expect lethal toxin production is the high altitude or latitude epicentres of population crashe s, in regions where recovery is by immigration as well as births; the right time is at the cusp of the crash. We propose an experimental design to tes t the hypothesis. (C) 1999 Academic Press.