Revegetation of Perennial Grasslands in the Peruvian Andes

Sept. 1, 2001

Recently I was able to participate in revegetation of grasslands that were disturbed by road construction in central Peru, about 150 mi. northeast of Lima, and I’d like to share some of the interesting results. The road revegetation was in an area that averaged about 14,000 ft. in elevation and ran through one of the great bunch grass grasslands of the world. In the Andes, forests are rare at this elevation, though eucalyptus plantations are common up to about 13,000 ft., and there are occasional stands of small native quinoa trees (Polylepis sp.). Because the area is only about 10º south of the equator, the climate at this high elevation is unpredictable. It can frost any day of the year, but killing frosts are very unusual, even during winter. The wet season is from November to March, and it rains at least a little almost every day during this period. The wet season is followed by a dry, cooler winter, though the growing season for perennial grasses is, in essence, 365 days per year.

The type of Andean grassland in the project area is known as the puna, and consists of dense stands of bunch grasses including fescue (Festuca), needle grass (Stipa), bluegrass (Poa), and reed bent grass (Calamagrostis). The major native vegetation type along the 70-mi. road was perennial grassland. The native grasslands have natural cover ranging from 90% to 100%, with very little bare ground; unvegetated areas are usually rock. The wildflower (forb) cover in the grasslands averages about 25%, and in some areas the forb/grass mix is 50/50. Moss and lichen cover is also significant in some grassland areas, contributing nearly 20% of ground cover. The grassland is mostly composed of bunch grasses of at least six different species, and there are at least six other species of grasses, including annual species, intermixed with the dominant bunch grasses.

Road improvement and new construction to serve new copper, silver, and gold mines and to connect isolated villages in remote parts of the Andes are common in Peru. Our revegetation project began in October, and I was able to assess results in May after the first peak growing season. The aim of revegetation along the entire access road was to establish grass species and one forb (clover) to stabilize the slopes, decrease soil erosion and protect water quality, improve the aesthetics along this public highway, and provide wildlife habitat and grazing areas.

Planting Methods

Before planting, natural contours were restored on the road edges, and where topsoil had been salvaged, it was spread. Two methods were used together to establish grasses: seeding, either by hydroseeding or broadcast seeding, and planting clumps of bunch grasses collected from nearby native stands. The seeding portion of the revegetation effort utilized non-native grass species, chiefly English rye grass (Lolium perenne, also called perennial rye grass, with seed imported from New Zealand). Mulch and fertilizer were also applied to the planted areas.

The bunch grass clumps planted during revegetation were a mixture of native species, chiefly fescue and needle grass, with an occasional bluegrass and reed bent grass. The local people commonly refer to the bunch grasses as “ichu” (pronounced e-choo), and they use the bunch grasses to thatch the roofs of their stone huts and barns. (An aside to this story is that we had to abandon the use of geotextile silt fence for sediment control and instead construct rock walls because the local people found the silt fence served as wonderful roofing material.) The scientific name of the native needle grass is Stipa ichu, and regardless of species, the local workers commonly called the clumps they collected from the local grasslands and planted in the revegetation areas “ichu.” This is somewhat confusing, because fescue appears to have been more widely planted in the revegetation areas than needle grass was; however, the terminology is so widely used by the local Spanish speakers that the term “ichu” will be used throughout this article to refer to the clumps of native bunch grasses planted in revegetation areas.


Hydroseeding and hydromulching, using commercially available cellulose mulch and a seeding rate of about 50 lb./ac., were utilized chiefly on steep slopes. The technique appears to be useful, but areas treated late in the growing season (after February) did not exhibit good vegetation cover by May. The most widely used mulch was rice straw spread by hand. The rice straw is an excellent mulch with coarse fibers often exceeding 12 in. The mulch was generally not crimped, but loss from washing or wind appeared minor in all areas. Areas mulched about seven months prior still had a good mulch layer during the May assessment.

The majority of areas were seeded by broadcast seeding, either by hand or using a seed spreader. This technique was very successful, and plant densities and distribution appeared acceptable seven months after planting. English rye grass was the most commonly used grass for revegetation because of likelihood of success (cover and root growth), cost of seed, availability of seed, and the species’ ability to serve as a “nurse crop” for native grasses, forbs, and shrubs that will subsequently invade the areas. Overall survival and growth of the rye grass has been excellent, though only an occasional plant was found to flower in the first year, but this may be attributed to heavy grazing by livestock. Regrowth in subsequent years from the established plants should be significant unless the plants are overgrazed.

Several other species of grasses were planted in small trial plots. In general, any cool-climate—adapted grass appears to be able to grow in this area (which is, after all, a grassland). Oat grasses (Avena sp.) have done very well and, in some places, form thick stands more than 3 ft. tall. Wheat (Triticum sp.) also grows well. One of the more impressive performers, though slow to establish, is orchard grass (Dactylis glomerata). It is recommended that future revegetation seed mixes in this area include a mixture of rye grass and orchard grass.

Two different species of clover (Trifolium sp.), red and white, were used during revegetation. In general, the red clover has performed the best on good sites (with topsoil and adequate moisture), but it is too early to judge which species works best on different sites. The use of red clover on better sites and the inclusion of white clover in seed mixes on poorer soils are recommended.

Workers collected clumps of native bunch grasses from nearby stands, and the collected clumps appear to be a random mixture of the species available, mostly the large fescue (Festuca orthophylla), a lesser amount of needle grass (Stipa ichu), and an occasional bluegrass and reed bent grass. Natural bunch grass clumps are typically about 3 ft. tall and 2 ft. across. These were dug up and divided by shovel or machete into smaller clumps about 3 in. across and trimmed to about 15-20 in. long. Often the clumps included few roots and sometimes only a few basal buds in place of roots. Frankly, I was skeptical these would survive, but the Peruvians assured me this would work. Shallow holes were dug or punched with a dibble, and the small clumps were stuck into the soil on about a 3- x 3-ft. spacing.

After seven months, overall survival of the planted clumps averaged about 90%, with about 5% of the plants flowering at the time of the assessment. Few of the clumps had increased in size, but root systems were often extensive, and much effort was needed to dislodge the rooted clumps from the soil, indicating they were successfully preventing erosion. There appeared to be little grazing damage to the clumps, possibly because the livestock could select seedling grasses instead and avoid the coarse foliage in the clumps. Also of note, grass seedlings, generally rye grass, survive and grow larger amidst the planted bunch grass clumps, probably because the clumps absorb more sunlight and warm the soil and because the clumps act to protect the seedlings from livestock.

The single most important factor identified to favor revegetation success was the presence of topsoil. Areas where topsoil was spread prior to revegetation have the greatest density and cover of vegetation. It is recommended that all areas of revegetation be provided with at least 4 in. of topsoil prior to planting. Of note, topsoil stockpiles that were not planted had little vegetation cover (less than 1%) after one year. The primary species that occurred on these unseeded areas was annual bluegrass (Poa annua), a non-native, potentially noxious weed that commonly occurs in the natural grasslands and reportedly is a common pest in local gardens. This finding suggests seeding or ichu planting is a necessity for disturbed areas to prevent accelerated erosion.


Invasion, or introductions of weedy plant species, does not appear to be a problem in any of our revegetation areas. Only three weedy forb species were common on any revegetation area, an annual mustard (possibly Brassica), chickweed (possibly Cerastium sp., Caryophyllaceae), and prickly lettuce (Latuca sp.). None of these contributes more than a trace of cover to any area, and none has inhibited revegetation success. Annual bluegrass is relatively common in all revegetation areas but contributes less than 1% of total cover on average. It does not appear to be interfering with revegetation of the perennial grass species along the road, though examination of other revegetated areas and some fields where local farmers planted wheat revealed rather extensive stands of annual bluegrass growing with the planted species. The annual bluegrass is common in the natural grasslands (though never covering much area) and could become a problem in some revegetation sites in the future if it naturally invades. The best prevention is to establish a robust stand of perennial grass species as quickly as possible.

One of the more interesting “weeds” that has appeared in one revegetation area has been a dense stand of reed canary grass (Phalaris arundinacea), which forms thick clumps up to 6 ft. tall. It is unknown how this species was introduced here, but it appears well adapted to the area and should be watched to make sure it does not spread to unwanted areas. Reed canary grass might be a candidate to plant in some revegetation areas in Peru, but its propensity to spread as a weed should first be evaluated.

Grazing by livestock, including cows, sheep, goats, horses, burros, mules, pigs, and occasionally alpacas, has had a detrimental effect on almost all revegetated areas. Herders sometimes drive their livestock to the revegetation areas, either because of the easy access along the road or to take advantage of the tender new grasses. Comparison of ungrazed revegetated areas, located on the tops of a few isolated roadside mounds with steep sides inaccessible to livestock, and nearby grazed areas showed that grazing has reduced vegetation cover by at least 50%. The impact of grazing on plant density is much less, with perhaps only 10% fewer plants in grazed areas.

There is no easy solution to this problem because grazing is so prevalent in this area, and I do not recommend fencing or trying to exclude access to these traditionally public lands. Establishment of the revegetated grasses will be greatly retarded by the grazing. Fortunately the livestock do not appear to eat the ichu clumps, so the native grasses should have a chance to become well established.

A Word About Shrubs

Shrubs were not generally planted during revegetation; however, a test using quart-size containerized shrubs of quinual (Polylepis racemosa) and colle (Buddleia coriacea) grown from locally collected seed was established in January. An assessment of this test in May found about 80% of the shrubs had survived. A few of the shrubs had almost doubled their planting height and were more than 12 in. tall. Other native shrub species in the area that might have some use in future revegetation efforts are lupine (Lupinus sp.) and a groundsel or ragwort (Senecio sp.). Incorporating these species into the revegetation program could be done by collecting and propagating seed in a nursery, direct seeding, or transplanting shrubs from undisturbed areas.

Some Final Thoughts

The success of the ichu grass clump plantings was so encouraging that I highly recommend this technique for other revegetation projects in the Andes, and trials of the method in other bunch grass grasslands around the world. Some colleagues have asked about the potential for the rye grass or other non-native grasses used in revegetation to spread as weeds in the native grassland. I think the potential is very low because the undisturbed grasslands are so thoroughly dominated by well-adapted native grasses. Also, wheat has been widely grown by mountain farmers for many decades; it does very well when used as an erosion control species, yet it has not spread beyond where it has been planted. It might be prudent, however, to use a sterile hybrid grass species along with a native planting, such as ichu, in areas where the native grasslands are disturbed.

I’ve also been asked about using ichu plantings by themselves or planting native seeds collected in the local areas with the ichu instead of using non-native seed. If erosion control is a major goal of the project, the ichu establishes too slowly, and it is important to have a crop to fill the spaces between the ichu. Also, without the seedling grasses, the ubiquitous grazing animals might take a heavier toll on the ichu. The labor-intensive effort of collecting local seed might be possible in Peru. I have noticed, however, that in any given year, few of the native bunch grasses flower and set seed, so planning a revegetation project around local seed collection may take some lead time to ensure that sufficient seed is obtainable.

We were favored with a hard-working labor force and skilled field managers, all Peruvians, who gave great care to the details necessary to carry out a project of this magnitude. Revegetation is a standard practice in Peru, and timely application of techniques that work will serve to maintain this beautiful country. 

About the Author

Roy A. Woodward

Roy A. Woodward, Ph.D., CPESC, works with the California Department of Parks and Recreation in Sacramento, CA.