The Puya clade is no longer considered to belong to the Pitcarnioideæ subfamily, but is a sister of the Bromelioideæ (Givnish et al. 2004, Horres et al. 2000, Terry et al. 1997).

Puya raimondii is rivalled only by members of the Ceroxylon palm genus as the most spectacular high-Andean plant. It occurs in often very isolated and usually small populations or rodales from Peru to Bolivia. Communities frequently number a few hundred individuals or less, but can range up to perhaps 30,000 plants in Paso Winchus as well as in Cashapampa, Pachacoto and sector Carpa, Huascarán National Park, Huaraz, which is probably Peru’s best known location. Populations reach 10,000 in Rodeo, Arani province, Bolivia’s largest population which may represent one third that country’s plants. In Titankayoc in southern Peru’s Ayacucho, however, there is an extraordinary site of several thousand hectares which contains, depending on source, an estimated 250,000 to at least 450,000 plants. With this arguable anomaly, the plant’s sporadic and scattered distribution and extreme genetic homogeneity, detailed below, suggest the vestigial remains of a species in decline. Outside its habitat, there appear to be no more than two dozen mainly small specimens in perhaps half a dozen botanical gardens.

The plant is monocarpic and in habitat seeds only once in about 80 years or more before dying. Although a mature Puya will produce 8–12 million seeds and viability is usually good, inclement montane conditions at the time of dispersal, which may inter alia affect pollinating insects, can result in few if any germinations. Moreover, seeds in less than ideal conditions can begin to lose germinating ability after a few months and are also susceptible to damping-off.  Because of these factors, a century-old plant may not reproduce at all and will, botanically, have lived in vain. This risk is exacerbated by global warming whose effects on Peru’s glaciers are well established. Climate change may already be impairing Puya raimondii’s ability to flower (Venero pers. comm.).

 
In the wild, these plants seem to be exceedingly choosy about where they grow. Their seeds are very small and by design easily wind-blown. Yet even in undisturbed locations P. raimondii can limit itself to one small spot although edafic, topographic and microclimatic conditions in the surrounding area appear identical.

The species is officially considered endangered in Peru (Law No 043- 2006-AG).  But in practical terms this means little, if anything and only the country’s best-known rodales (Puya communities) benefit from some protection.  Elsewhere, the plant can be the object of hostility and large stones are often thrown at them, lodging for years among the leaves. Cattle roam unfettered among many colonies which are not being regenerated because the animals either trample or may eat juveniles. In other sites, fires are set to create pastureland or the thorny leaves are burnt to facilitate access to the trunks' starch which becomes cattle fodder. Pith removal, of course, kills the plant.

Compounding all these issues, there is exceedingly little genetic variety within existing populations. A genomic DNA analysis (Sgorbati et al.) of the genetic structure of eight populations (including the huge one at Titankayoc) representative of P. raimondii in Peru detected just 14 genotypes in 160 plants. Only a few of the 217 AFLP marker loci screened were polymorphic and four populations were completely monomorphic. Less than 4% of the total genomic variation was within populations and genetic similarity among populations was as high as 98.3%.  Flow cytometry of seed nuclear DNA content and RAPD marker segregation analysis of progeny plantlets demonstrated that the extremely uniform genome of these populations is not compatible with agamospermy but the result of inbreeding. Many rodales have disappeared only recently, as proved by vernacular Puya names still in local use. Even though fully fitted to its harsh environment, P. raimondii lacks sufficient variability in its genome and, possibly also phenotype variability to allow it to adapt to both anthropic pressure and present climate change.

This plant usually occurs at around 4,000 m in the Andes of Peru and Bolivia, but it ranges from 3,000 m up to 4,800 m (both extremes occur in Bolivia). Its communities are often very isolated from each other and can be found in pockets from Calipuy in northern Peru, Huaraz in the centre to Apurimac and Ayacucho in the south, crossing into western Bolivia’s La Paz province, Cochabamba in the centre and Potosí in the south. The most important site by far in Peru is Titankayoc - Chanchayoc in Ayacucho, while Bolivia’s largest known community is in Rodeo, Arani province. A single reference to Puya species in southern Ecuador and northern Chile has not been verified.

Thanks to a single enormous subpopulation, which could represent most of the world’s population of P. raimondii, the number of these plants in Peru may number 800,000 individuals. Bolivia is estimated to have 30,000-35,000 plants.

Typical habitat for this species occurs at about 4,000 m but can extend from 3,000–4,800 m. At these levels, air temperatures range from very cold (as low as -20ºC or less) to an estimated maxima of 8–24ºC. Precipitation, as rain, hail or snow falls mainly from October to March. The ground is almost invariably rocky and usually sloping and friable. Drainage requirements or an inability to compete with other flora in more fertile land may account for the rarity of P. raimondii in damp gullies although a few specimens have been seen close to standing water. Still less clear is why a plant known to thrive in very different ex-situ conditions confines itself to a single spot on a mountainside when surrounded by seemingly similar terrain; or why population densities can vary considerably among and within communities.

The Puya’s ability to grow and even thrive in quite different conditions (low altitude, high humidity, high temperature) in which maturity can be reached in half the time needed in the wild, makes the plant a candidate for a range of ex situ environments.

The Puya is susceptible to threatening events because its communities are generally small and very isolated from each other. This has apparently rendered populations extremely homogeneous genetically due probably to autogamous inbreeding depression and so at greater risk to disease, parasites or predators accompanying climatic change. The latter is well documented and manifest in Peru’s rapidly receding glaciers.  For possibly related but as yet unknown reasons, at least one population has not flowered for decades.

The major risk in most communities, however, is due to human impact including repeated fires to generate or maintain pasture land and usage as fuel or building material by local populations. An added incentive is the fear that Puyas may ‘capture’ grazing animals with their leaves’ (fearsome) inward-curving spines. This is probably very rare but not inconceivable. Native birds have been ensnared and killed this way.

The Puya trunk’s pith is at least occasionally harvested to feed domestic animals. Simple furniture is sometimes made from the plant and its woody parts used for fuel. Leaves are buried upright in the ground to form rudimentary fences in a few Bolivian villages and can sometimes be inserted in the tops of stone walls to enhance separation in Peru where they may also be used to shed rain from adobe walls. Such usages are strictly local and usually small-scale.

Puyas are officially considered endangered and protected by legislation in Peru (Law No 043- 2006-AG). But the practical enforcement of this measure are not evident outside of one national park and the law must be strengthened and enforced. Likewise, Peru’s pertinent authorities should be urged to promote Puya communities as a tourist attraction and to raise awareness of the plant’s value among both children and adults living near existing groves. Acción Ambiental plan to select at least one reserve for specific conservation measures. This may be the huge Titankayoc site or one nearer the major tourist destination, Cusco.

The plant also needs to be far better known and steps should be taken to strengthen communities of P. raimondii genetically. To this end, we plan to collect seeds from the most divergent Puya sites, altitudinally and latitudinally, to reproduce in nurseries as a reserve. Following further research and if endorsed by experts seeds might ultimately be used to cross-seed (some) existing populations. Reproducing and extending the plant, which in a sparse habitat has significant ecological importance, is another priority. Establishing one or more new rodales in promising sites will be evaluated.

The ‘Queen of the Andes’ is also of considerable ornamental value and introducing it as a dramatic landscaping element internationally as well as nationally is another goal. The author is aware of only a few specimens being used for such purposes in Peru’s Puno and Cusco provinces. Preliminary contacts have already been established with several foreign botanical gardens.