Principles of Ecological Landscape Design PDF

Principles of Ecological Landscape Design Principles of Ecological Landscape Design Travis Beck Washington | Covelo | London © 2013 Travis Beck All rights reserved under International and Pan-American Copyright Conventions. No part of this book may be reproduced...

Principles of Ecological Landscape Design Principles of Ecological Landscape Design Travis Beck Washington | Covelo | London © 2013 Travis Beck All rights reserved under International and Pan-American Copyright Conventions. No part of this book may be reproduced in any form or by any means without permission in writing from the publisher: Island Press, Suite 300, 1718 Connecticut Ave., NW, Washington, DC 20009 ISLAND PRESS is a trademark of the Center for Resource Economics. Library of Congress Cataloging-in-Publication Data Beck, Travis. Principles of ecological landscape design / Travis Beck. p. cm. Includes bibliographical references and index. ISBN 978-1-59726-701-4 (cloth : alk. paper) -- ISBN 1-59726-701-5 (cloth : alk. paper) -- ISBN 978-1- 59726-702-1 (pbk. : alk. paper) -- ISBN 1-59726-702-3 (pbk. : alk. paper) 1. Ecological landscape design. 2. Ecosystem health. I. Title. QH541.15.L35B43 2012 577--dc23 2012022172 Printed using Franklin Gothic Condensed Typesetting by Lyle Rosbotham Printed by Printed on recycled, acid-free paper Manufactured in the United States of America 10 9 8 7 6 5 4 3 2 1 Keywords: biodiversity, biogeography, biomes, climate change, competition, disturbance, ecology, ecosystem management, edge effect, keystone species, landscape, landscape ecology, microclimate, plant communities, plant populations, soils, succession, Sustainable Sites, water  To those who taught me ecology and biology, especially Mr. Tolley, Richard Irwin, and Dr. Ralph Boerner.  Contents Acknowledgments xi Foreword xiii Introduction 1 1 Right Plant, Right Place: Biogeography and Plant Selection 7 2 Beyond Massing: Working with Plant Populations and Communities 33 3 The Struggle for Coexistence: On Competition and Assembling Tight Communities 65 4 Complex Creations: Designing and Managing Ecosystems 89 5 Maintaining the World as We Know It: Biodiversity for High-Functioning Landscapes 107 6 The Stuff of Life: Promoting Living Soils and Healthy Waters 125 7 The Birds and the Bees: Integrating Other Organisms 153 8 When Lightning Strikes: Counting on Disturbance, Planning for Succession 179 9 An Ever-Shifting Mosaic: Landscape Ecology Applied 209 10 No Time Like the Present: Creating Landscapes for an Era of Global Change 235 Bibliography 261 Index 273  Acknowledgments This book is the culmination of a long process of thought and discovery, reaching back to graduate school and before. Therefore, more people have offered ideas, assistance, and support than I can name here. Above all, this book stands on the shoulders of the many ecologists whose articles I consulted and of all those who generously shared their work and experiences with me as case studies. However, there are several individuals and institutions whom I would especially like to thank. Martin Quigley was an essential ally and made numerous contributions. Carol Franklin offered en- couragement and advice, as well as a case study and the Foreword. Erica Beade of MBC Graphics went above and beyond the call of duty in her preparation of illustrations for the book. The dedi- cated staff of the LuEsther T. Mertz Library at the New York Botanical Garden helped my research go smoothly. The main reading room at the New York Public Library’s Schwartzman Building provided a congenial atmosphere for many hours of writing. Sarah Paulson hunted down several key images and was endlessly patient and supportive during the final push. At Island Press, I am grateful to Heather Boyer, who adopted this project immediately, proved pa- tient during its maturation, and pushed me to complete it when the time came, and to Courtney Lix, who has kept tabs on the entire process. Kate Lu obtained numerous permissions to help finalize the art package. Writing a book proved to be a rewarding and demanding undertaking, and I am grateful to everyone who helped push the project forward.  Foreword The desire for informed sustainable, ecological, and regenerative design is increasing in every country, accentuated by recognition of the increasing severity of a wide variety of ecological crises (from the dead zones of many oceans to global climate change). A growing dissatisfaction with the ugliness and wastefulness of conventional development is accelerating this interest. Historically, plants for human landscapes were brought together for medicinal, economic, or aesthetic purposes. Now, a new para- digm is finally catching the popular imagination. In this era of threatened environmental Armageddon, ecological design, and in particular ecological planting design, is finally being understood as a critical tool for our ultimate survival. With the loss of almost all our undisturbed natural landscapes, there is also a growing appreciation of the beauty and function of our indigenous landscapes. Many professionals and nonprofessionals, from a variety of different disciplines and backgrounds, claim an expertise in creating natural plant communities to meet the growing demand for sustainable designs. Engineers, architects, landscape architects, restoration specialists, passionate volunteers, and others often find themselves in charge of the restoration of deteriorating plant communities or actual habitat re-creation. Engineering firms routinely turn out planting designs for floodplains and riparian corridors or planting plans for rain gardens and biotreatment swales, with little knowledge of the relation- ships of individual plant species to specific environmental conditions such as water tables or contours. However, there may be light at the end of the tunnel. Cities big and small, throughout the country, are adopting new form-based zoning and performance standards for stormwater management measures and are asking designers for smart growth plans and green infrastructure. As mandates from the Environmental Protection Agency and other government agencies encourage more environmentally respectful designs in our metropolitan areas, and as the public increasingly sees the need for less ugly and wasteful land use practices, attention is being focused on the successful design, installation, and establishment of native plant communities in appropriate environmental conditions that will sustain them. Additionally, nonprofit conservation and land management organizations, educational institutions, and public officials charged with evaluating and overseeing the implementation of mandated programs and practices want to see at- tractive, successful solutions, measured by clear performance standards. Until recently, the establishment or repair of native plant communities has been relatively unsophis- xiv Foreword ticated. Ecological restoration and habitat re-creation are very new disciplines. Research into many as- pects of plant ecology is either lacking entirely or discussed only in scientific articles, where the language is unfamiliar and the goal is not the translation of research into design actions. Compounding these problems, the older natural sciences often had a tradition of isolation. Even scientists within the same discipline often had difficulty communicating with their peers. (Until recently, soil scientists specialized in either soil structure, soil chemistry, or soil biology. They failed to communicate with each other and to understand soil as an interacting medium, where all these components are interdependent.) Only a few books have tackled the subject of linking the structure, function, composition, and organization of landscapes directly to ecological processes. Richard T. T. Forman, professor of land- scape ecology at the Harvard Graduate School of Design, has been one of the first authors to call our attention to the fact that spatial patterns reflect these processes. He has written a number of books and articles introducing designers to ecological ideas, particularly Landscape Ecology Principles in Landscape Architecture and Land Use Planning, with Wenche Dramstad and James Olson (1996), and Land Mosaics: The Ecology of Landscapes and Regions (1995). Principles of Ecological Landscape Design, by landscape architect Travis Beck of the New York Botanical Garden, is an excellent expansion of earlier books on this subject. This wonderful book is the most comprehensive exploration of a planting design approach based on the principles of plant ecol- ogy yet to be provided to designers. Principles of Ecological Landscape Design interweaves very clear descriptions of critical ecological processes to explain the effects of biogeography, foodwebs, nutrient cycles, plant and animal interac- tions, and many other factors on species composition, function, and spatial organization in natural plant communities. Each principle of plant ecology is paired directly with the implications for planting design, including the ecological processes that have shaped broad landscape configurations such as edges, centers, the fragmentation of a landscape, and the connections between landscapes. For ecological planting design to be more than greenwashing, this book provides much-needed and long-awaited access to principles, strategies, and specific directions. It allows us to understand, with scientific rigor, the full requirements of establishing thriving plant communities in appropriate habitats, with appropriate plant companions and in requisite numbers and densities. It offers us both an over- view of the central issues and a concise, easy-to-use reference. In many ways, this book is recognition of our newfound sophistication and of how far we have come since the era of “progress” of the 1950s to the 1980s and since Ian McHarg wrote Design with Nature in 1969. With our increasing need to change the destructive plans and practices of the recent past, and with our growing familiarity with ecological ideas, this book is an indispensable next step, firmly linking the breadth and depth of es- sential ecological processes directly to design actions. In doing so, Beck has given us a new and better toolbox for ecological planting design. Carol Franklin, RLA, FASLA Founding Principal, Andropogon Associates, Ltd. Philadelphia, PA May 2012  Introduction Here, at the beginning of the twenty-first century, we find ourselves in an unprecedented situation. More than seven billion humans dominate the planet in ways we never have before. Our ever-expanding megalopolises creep out into landscapes cut over for timber, mined for fuel, bisected by roads, grazed by livestock, drained and plowed for farming, put back to cover, abandoned and regrown, parceled for houses, or opened for recreation. Even the most pristine wilderness areas are subject to our legislated forbearance. The rain that falls on them is enriched and polluted by our activities elsewhere, and the climate they live under is shifting by our hand. As human influence over the planet grows, and as the built environment increases in prominence, the landscapes we design and manage will play an increasingly important role. From now on, the ecological function of our planet can come only from a network of preserved, restored, managed, and constructed landscapes. To maintain the function of this network, and the quality of life that it offers, we will have to change the way we think about landscape design. A landscape, in its first meaning, is a depiction of scenery, and this has been our conventional approach to landscape design. Think of New York City’s Central Park, a site to which the origins of landscape architecture in the United States are often traced, and the High Line, one of the most talked- about contemporary landscapes. In these master works, art imitates nature or perhaps an idealized nature already represented in art. Some assume that Central Park, with its pastoral fields and tangled woodlands, preserves a remnant of the farmlands and wilds that once occupied the center of Manhattan. In fact, in 1857, when the competition for the design of Central Park was announced, the site was a tract of rocky swamps. In their winning entry, Frederick Law Olmsted and Calvert Vaux conjured both English and American scenes. The “Greensward” Plan, as the designers called it, featured broad meadows and contoured lakes, similar to those found in the English countryside or, more accurately, in the English countryside as reimagined by “landscape improver” Capability Brown. The plan also included dramatic rock outcroppings, cascades, and dense woods, like those in New York’s Hudson River Valley and Catskill Mountains, or again, more accurately, like those in the landscape paintings of the Hudson Valley School. Olmsted and Vaux’s object was to evoke in the visitor a range of emotions, from tranquility and deliberation at the edge of T. Beck, Principles of Ecological Landscape Design, 1 DOI 10.5822/978-1-61091-199-3_0, © 2013 Travis Beck 2 Principles of Ecological Landscape Design The Lake, to excitement and rapture in the midst of The Ramble. The romantic place names complete the vision of an untrammeled world apart from the city’s grid. To achieve this vision required a massive reengineering of the site, carried out over 20 years. Existing rock was blasted out, and some of the rubble used to build other features. About 500,000 cubic feet of topsoil was brought in from New Jersey. Four million trees, shrubs, and plants were acquired. In the end, ten million cartloads of material had been hauled in or out. In her appreciative history of Central Park, Sara Cedar Miller (2003: 13) wrote, The 843-acre Park seems natural because it is composed of real soil, grass, trees, water, and flowers that need constant tending. In reality, however, it is naturalistic—an engineered environment that is closer in essence to scenes created in Hollywood than it is to the creation of Mother Nature. The appeal of such naturalism is still strong. A century and a half later, and just a few miles away, it has taken contemporary form in the High Line. Before it became a celebrated public park, this abandoned rail line on the west side of Manhattan drew urban explorers onto its elevated decks, where an unexpected wilderness had emerged. Botanist Richard Stalter (2004) describes passing from an artist’s loft, across an adjacent roof, then via ladder and rope to the train tracks, where he cataloged 161 species, more than half of them native, growing in a dry grassland punctuated by the occasional tree of heaven (Ailanthus altissima) (fig. I.1). Haunting photographs by Joel Sternfeld (2001) of the overgrown industrial infrastructure captured the public imagination, helped garner support for saving the space from demolition, and set the tone for the park that has emerged. The converted High Line, by James Corner Field Operations, Diller Scofidio + Renfro, and planting designer Piet Oudolf, enthralls crowds with its offset walkways, clever details, and staccato views out over the Hudson River and below to city streets. Exuberant plantings grow between relaid steel tracks from gravel mulch meant to recall railroad ballast (fig. I.2). When the first section of the High Line opened in 2009, New York Times architecture critic Nicolai Ouroussoff noted the resonance between the new plantings and what grew naturally before: And those gardens have a wild, ragged look that echoes the character of the old abandoned track bed when it was covered with weeds, just a few years ago. Wildflowers and prairie grasses mix with Amelanchier, their bushes speckled with red berries.... On Saturday the gardens were swarming with bees, butterflies and birds. I half expected to see Bambi. (Ouroussoff 2009: C1) Hollywood should be proud. As at Central Park, creating such a peaceable kingdom on the High Line was an enormous undertaking. A total of $152 million was spent on the first two sections’ few slender acres, to upgrade the infrastructure and make it safe for visitors, to design and mix specialty soils and lift them by the bagful onto the platform, and to procure and plant the thousands of grasses, flowers, and trees that evoke the spontaneous vegetation they replaced. Two spectacular parks call on us to imagine unfettered nature, yet they took prodigious human  Introduction 3 Figure I.1 Joel Sternfeld. Looking East on 30th Street on a Late September Morning, 2000. (©2000, Joel Sternfeld; image courtesy of the artist, Luhring Augustine, New York, and The Friends of the High Line, New York.) effort to construct, not to mention the ongoing exertions needed for their maintenance. Let these parks represent conventional landscape design. We need not even think of the acres of aspirational suburban yards, the turf-filled office parks, and the windy municipal plazas. Conventional landscapes, both the masterpieces and the mass produced, are intended to accommodate human functions while achieving a certain look and evoking certain feelings. They rely on a designer’s vision, the alteration of the site as necessary to achieve that vision, and an often lengthy period of maturation and care. What if, instead of depicting nature, we allowed nature in? What if, instead of building and maintaining artistic creations, we worked to develop and manage living systems? What could we learn from the wild and pastoral landscapes that Central Park imitates, and from places such as the undeveloped High Line, about how nature works? Could we create landscapes that were more efficient, more connected, more effective, and ultimately more valuable? In other words, could we create ecological landscapes? An ecological landscape is a designed landscape based on the science of ecology. To clarify one point immediately, when ecologists say “landscape” they mean an area comprising multiple 4 Principles of Ecological Landscape Design Figure I.2 The High Line, Section 1, July 2009. (Photo by Travis Beck.) patches that differ from one another. An ecological landscape is a landscape in this sense, but typically in this book we will emphasize designed or constructed when we refer to landscapes, such as ecological landscapes, that are imagined and assembled by people. Ecological landscapes may abut or include natural ecosystems, but above all they are human creations. An ecological design may incorporate restoration of degraded ecosystems, but it does not principally seek to put things back the way they were. Ecological landscape design is for the growing number of areas where there is no going back to the way things were. It aims instead to go forward, to apply our knowledge of nature to create high-performing landscapes in which our design goals and natural processes go hand in hand. The science of ecology offers our most rigorous and accurate understanding of how nature works at the scales most relevant to landscape designers. It is a growing understanding based on more than one hundred years of observation, experiment, and debate. The scientific side of this book draws from academic articles, both classic and recent, and aims to present an overall picture of the current state of knowledge. The state of ecological knowledge may surprise you. It does not describe webs of exquisite interconnectedness and balance, with every creature in its place. Rather, it outlines a world ruled by  Introduction 5 change and chance, in which life self-organizes and persists. This is the world we must deal with as designers and managers of landscapes. The design side of this book applies ecological understanding to answer practical questions. How do we set up a planting so that it will thrive with a minimum of care? How many different species should we include, and how do we select them? What do we do with the animals that show up? In what ways should a project we are designing relate to what is around it? How can a constructed landscape live through a catastrophe and recover? Can the landscapes we design help us face the environmental challenges of the twenty-first century? In places the answers are speculative, suggesting strategies for a theoretical ecological landscape. Often, however, they are based on actual projects in a range of sizes from multiple regions of the United States. Increasingly, landscape professionals are taking an ecological approach to their work. For instance, there has been a large shift toward more natural approaches to managing stormwater. Landscape architects and landscape designers have also explored ecological methods of plant community assembly and managing the changes in plant communities over time. Notably, the American Society of Landscape Architects has taken a lead role in developing the Sustainable Sites Initiative, which offers a set of guidelines and benchmarks for sustainable land development practices centered around the idea of providing ecosystem services (Sustainable Sites Initiative 2009a, 2009b). To be sustainable means to perform these indispensable services while demanding fewer resources, which we might think of as doing more with less. The best way to do more with less is to harness ecological processes. An ecological landscape knits itself into the biosphere so that it both is sustained by natural processes and sustains life within its boundaries and beyond. It is not a duplicate of wild nature (that we must protect and restore where we can) but a complex system modeled after nature. Above all, to be sustainable is to continue functioning, come what may. An ecological landscape is based on self-organized patterns, which are more robust than patterns imposed according to some external conceit. It is flexible and adaptive and continually adjusts its patterns as conditions change and events unfold. We know such systems are beautiful and arousing because we have been imitating them in our designed landscapes for so long. Now that humans have co-opted so much of the planet, the time has come to cease representation and to partner with nature instead in acts of vital co-creation. 1. Right Plant, Right Place: Biogeography and Plant Selection It is often said that the secret to good horticulture is putting the right plant in the right place. By match- ing plants to their intended environment, a designer helps to ensure that the plants will be healthy, grow well, and need a minimum of care. Too often designers force plants into the wrong places, putting large trees that thrive in extensive floodplains into confining tree pits or planting roses that need full sun in spindliness-inducing shade. Or we try to create a generically “perfect” garden environment, with rich soils and regular moisture, for a wide-ranging collection of plants, some of which may actually prefer more stringent conditions. Whether we do these things from ignorance, in conformance with established practices, or because our focus is on aesthetic qualities or our associations with certain plants, the too common result is struggling plantings, ongoing horticultural effort, and the dominance of familiar generalist species. An ecological approach to landscape design takes the fundamental horticultural precept—right plant, right place—and views it through a biogeographical lens. Where do plants grow, and why do they grow there? How many degrees of native are there? What are the relative roles of environmental adap- tation and historical accident? Selecting plants according to biogeographical principles can help us create designed landscapes that will thrive and sustain themselves. Such landscapes celebrate their region and fit coherently into the larger environment. Of course, these landscapes can also be beautiful. Let us begin, then, with a fundamental ecological question: Why is this plant growing here? Plants Are Adapted to Different Environments Five hundred million years ago the earth’s landmasses were devoid of life. Then, scientists speculate, ancestral relatives of today’s mosses began to grow along moist ocean margins and eventually on land itself. To survive out of water, these primitive plants had to evolve structures to support them- selves out of water, ways to avoid drying out, and the ability to tolerate a broader range of tempera- tures. As they evolved, plants diversified and spread into every imaginable habitat, from deserts to T. Beck, Principles of Ecological Landscape Design, 7 DOI 10.5822/978-1-61091-199-3_1, © 2013 Travis Beck 8 Principles of Ecological Landscape Design wetlands, and from the tropics to the Arctic. Today, there are more than 300,000 plant species on our planet (May 2000). Plant diversity and the diversity of habitats on Earth are closely related. Natural landscapes are composed of heterogeneous patches, each of which presents a different environment (see chap. 9). At the largest scale are deserts and rainforests. At the smallest scale are warm, sunny spots and wet depressions. Charles Darwin proposed in The Origin of Species (1859: 145), The more diversified the descendants from any one species become in structure, constitution, and habits, by so much will they be better enabled to seize on many and widely diversified places in the polity of nature, and so be enabled to increase in numbers. Plants have been able to move into so many different environments because they have developed many means of adapting. Consider plant adaptations to two critical environmental variables: tempera- ture and the availability of water. Temperature affects nearly all plant processes, including photosynthesis, respiration, transpiration, and growth. Very high temperatures can disrupt metabolism and denature proteins. Low temperatures can reduce photosynthesis and growth to perilously low levels and damage plant tissues as ice forms within and between cells. Plants that grow in high-temperature regions may have reflective leaves or leaves that orient themselves parallel to the sun’s rays in order to not build up heat. Some use the alternate C4 photosynthetic pathway, which can continue to operate efficiently at high temperatures. Plants in cold regions have developed bud dormancy and may grow slowly over several seasons before producing seed. They have high concentrations of soluble sugars in their cells to act as natural anti- freeze, and they are able to accommodate intercellular ice without experiencing damage. Plants that are adapted to grow well in wet, moist, and dry conditions are called, respectively, hydro- phytes, mesophytes, and xerophytes. Hydrophytes have to provide oxygen to their flooded roots, which they do through a variety of mechanisms, including by developing spongy, air-filled tissue between the stems and the roots or by growing structures like knees that bring oxygen directly to the roots (fig. 1.1). Many hydrophytes also have narrow, flexible leaves to avoid damage from moving water. Xerophytes, on the other hand, exhibit adaptations to lack of water such as small leaves, deep roots, water storage in their tissues, and use of an alternate photosynthetic pathway that allows them to open the stomata on their leaves only in the cool of night (fig. 1.2). Because of 500 million years of evolution and the diversity of habitats open to colonization, the planet is now filled with plants adapted to nearly every combination of environmental variables. Choose Plants That Are Adapted to the Local Environment Because plants exhibit such a wide range of natural adaptations, we need not struggle—expending both limited resources and our collective energy—against the environment we find ourselves in to make it a better home for ill-suited plants. Using biogeography as our guide, we can always identify plants ready-made for the conditions at hand. Gardeners, nursery owners, and landscape designers have long recognized that plants ill-suited Right Plant, Right Place: Biogeography and Plant Selection 9 Figure 1.1 Knees bring oxygen to the roots of some hydrophytes, such as these bald cypress (Taxodium distichum) growing in a swamp at the Lacassine National Wildlife Refuge in Louisiana. (Photo courtesy of the US Fish and Wildlife Service.) Figure 1.2 Tree cholla (Cylindropuntia imbricata), a xerophyte native to the southwestern United States and northern Mexico, photosynthesizes with its stems, rather than with leaves, and stores water from periodic rainfall in succulent tissues protected with spines. (Photo by Gary Kramer, USDA Natural Resources Conservation Service.) 10 Principles of Ecological Landscape Design to the temperature extremes of the place where they are planted are unlikely to survive their first year in the ground. The US Department of Agriculture has codified this knowledge in a map of hardiness zones, which was updated in 2012 (fig. 1.3). Hardiness zones represent the average annual minimum temperature, that is, the coldest temperature a plant in that zone could expect to experience. There are thirteen hardiness zones, ranging from zone one in the interior of Alaska (experiencing staggering winter minimums of below –50°F) to zone thirteen on Puerto Rico (experiencing winter minimums of barely 60°F). Plants are rated as to the lowest zone in which they can survive. Balsam fir (Abies balsamea), for instance, is hardy to zone three. The hardiest species of Bougainvillea are hardy only to zone nine. Plants are sometimes given a range (e.g., zones three to six). Strictly speaking, hardiness refers only to ability to survive minimum temperatures, but the practice of indicating a range serves as shorthand for the overall temperatures in which a plant will grow. The American Horticultural Society (2012) has also prepared a map of heat zones for the United States, indicating the number of days above 86°F that a region experiences on average per year. Catalog descriptions of landscape plants may include reference to these heat zones and to the more common hardiness zones. Given the wide acceptance of hardiness zones, it is somewhat surprising that similar thinking ap- plied to water requirements for plants has developed only within the past few decades. Perhaps this is because of the ease of meeting the needs of some plants for more water with irrigation. Or perhaps it Figure 1.3 The 2012 USDA Plant Hardiness Zone Map. Note the fairly regular progression of zones from north to south in the center of the continent and the irregular zone boundaries related to mountain ranges and the moderating effects of large bodies of water (including the Great Lakes) in the west and east. (US Department of Agriculture.)  Right Plant, Right Place: Biogeography and Plant Selection 11 is because of the deep influence of English gardening traditions in the United States and expectations of what a cultivated landscape should look like. Regardless of local conditions, our nationwide default residential landscape is water-hungry lawns and summer-flowering borders. Many regions of North America are in fact too dry, or receive precipitation too unevenly, to support this kind of designed land- scape without major inputs of water. In San Diego, for example, more than half of all residential water is used to irrigate lawns and landscapes (Generoso 2002). Using water this way can deplete aquifers, damage habitat in areas from which water is drawn, decrease local agricultural production, and leave our landscapes vulnerable to desiccation when water restrictions go into effect. The negative consequences of landscape irrigation and the countervailing benefits of water conser- vation motivated Denver Water (the water department in Denver, Colorado) to introduce xeriscaping in 1981. Xeriscaping, from the Greek word xeros, for “dry,” emphasizes grouping plants in the landscape according to their water needs (Weinstein 1999). Not surprisingly, many xeriscapes feature xerophytes, plants with low water needs. Denver exists in a semiarid environment, getting on average around 14 inches of precipitation a year, as compared to about 35 to 40 inches a year in most areas east of the Mississippi. Kentucky bluegrass (Poa pratensis) lawns, shade trees, and most common garden plants need additional water to survive. At their former home, Panayoti and Gwen Kelaidis ambitiously replaced their entire front lawn with plants well adapted to Denver’s semiaridity. These include sulfur-flower buckwheat (Eriogonum umbellatum), soapweed (Yucca glauca), and partridge feather (Tanacetum densum ssp. amani). Today, 20 years later, these plants are still thriving with no supplemental irrigation (fig. 1.4). Selecting plants that are adapted to the temperatures and available water of the environment in which they will be placed is a fundamental step in creating an ecological landscape. Biomes Describe the Broad Character of a Region’s Vegetation In late July 1799, Alexander von Humboldt, a German naturalist traveling aboard a Spanish vessel with French botanist Aimé Bonpland, came to the bow for his first glimpse of the shore of South America (Humboldt and Bonpland 1818: 175–76). He wrote, Our eyes were fixed on the groups of cocoa-trees that border the river, and the trunks of which, more than sixty feet high, towered over the landscape. The plain was covered with tufts of cassias, capers, and those arborescent mimosas, which, like the pine of Italy, extend their branches in the form of an umbrella. The pinnated leaves of the palms were conspicuous on the azure of a sky, the clearness of which was unsullied by any trace of vapors. The Sun was ascending rapidly toward the zenith. A dazzling light was spread through the air, along the whitish hills strewed with cylindric cactuses, and over a sea ever calm, the shores of which were peopled with alcatras, egrets, and flamingoes. The splendor of the day, the vivid coloring of the vegetable world, the forms of the plants, the varied plumage of the birds, everything announced the grand aspect of nature in the equinoctial regions. For a couple of newly arrived Europeans, these were truly stunning sights. After 5 years of travel throughout Latin America, Humboldt (1805: 56) was able to organize some of his observations in 12 Principles of Ecological Landscape Design Figure 1.4 Xeriscape by Panayoti and Gwen Kelaidis, teeming with plants that need no supplemental irrigation, in Denver, Colorado. (Photo by Travis Beck.) an essay on what he called “the geography of plants,” a foundational work in the field we now know as biogeography: Plant forms closer to the equator are generally more majestic and imposing; the veneer of leaves is more brilliant, the tissue of the parenchyma more lax and succulent. The tallest trees are constantly adorned by larger, more beautiful and odoriferous flowers than in temperate zones.... However the tropics never offer our eyes the green expanse of prairies bordering rivers in the countries of the north: one hardly ever has the gentle sensation of spring awakening vegetation. Nature, beneficial to all beings, has reserved for each region particular gifts. A tissue of fibers more or less lax, vegetable colors more or less brash depending on the chemical mixture of elements and the stimulating strength of solar rays: these are just some of the causes that give each zone of the globe’s vegetation its particular character.  Right Plant, Right Place: Biogeography and Plant Selection 13 What Humboldt recognized as the particular character of each zone of the globe’s vegetation we today call a biome. Biomes are large geographic areas dominated by certain types of plant and animal life: rainforests in the wet tropics, for example, or prairies in drier temperate zones. Although today hu- man impacts, particularly agriculture and urbanization, have somewhat obscured the nature and extent of biomes, underlying climatological realities still inform what will grow where. In fact, a simple graph uniting temperature and precipitation shows us what conditions give rise to what biomes (fig. 1.5). Where there are high temperatures and high levels of precipitation, tropical rainforests grow, capturing the sun’s energy in the substantial biomass of large standing forests. At the opposite extreme, where temperature and precipitation are both low, we find arctic and alpine tundra full of slow-growing, di- minutive plants. Thus, temperature and precipitation drive the adaptations of plants in each area and determine the character of the vegetation in different regions of the globe. Figure 1.5 Graph of biomes in relation to precipitation and mean annual temperature. (Courtesy of http:// www.thewildclassroom.com/biomes.) Design Different Landscapes for Different Biomes Throughout the United States many cultivated landscapes are out of touch with their surroundings. Woodland trees line scrubland streets, and broad lawns hacked from forests pretend at grassland or pasture. An ecological landscape should respond to the same environmental realities that give rise to biomes. Areas of moderate temperature and moderate precipitation tend to be forest. Areas with less water tend to be grassland, desert, or in some cases chaparral. Designed landscapes that match their plants and the communities in which they are grown to the prevailing climate should take less effort to create and maintain. They will also be better able to provide habitat for local wildlife (see chap. 7), better connect to regional landscape networks (see chap. 9), and better bounce back after predictable disturbances such as fire, windstorms, or floods (see chap. 8). 14 Principles of Ecological Landscape Design In the rolling hills of northern Delaware, set between the formal gardens and the native forest of Mt. Cuba Center, the Woods Path is a landscape aptly suited to its biome (fig. 1.6). The moderate tem- peratures and rainfall in this region naturally give rise to a deciduous forest, the structure and feel of which the Woods Path captures in a subtly designed landscape. Tulip trees (Liriodendron tulipifera) pro- vide an enclosing canopy and strong vertical architecture. Flowering dogwood (Cornus florida), hollies, and rhododendrons encroach gently on the meandering path. Herbaceous plants carpet the ground plane. This simple structure—canopy, understory, herbaceous groundcover—expresses the essence of a woodland. It also provides the cool, humid, partially shaded growing conditions to which many plants of the Appalachian Piedmont are adapted. The Woods Path is a designed garden that feels absolutely appropriate to its place. Figure 1.6 Woods Path at Mt. Cuba Center, Delaware. Tulip trees, rhododendron, and herbaceous groundcovers create a three-story landscape that encloses the visitor. (Photo by Rick J. Lewandowski.) Across the continent, at the base of southern California’s San Gabriel Mountains, the coastal sage scrub community at Rancho Santa Ana Botanic Garden looks completely different (fig. 1.7). Here, be- fore agriculture and development, the dry, warm climate supported an open scrub, also known as soft chaparral, intermixed with oak savannas and riparian woodlands. At Rancho Santa Ana an oak wood- land provides a backdrop against which nestles a mix of loosely spaced drought-tolerant shrubs, in- cluding California sage (Artemisia californica), black sage (Salvia mellifera), and wild lilac (Ceanothus  Right Plant, Right Place: Biogeography and Plant Selection 15 Figure 1.7 Coastal sage scrub community at Rancho Santa Ana Botanic Garden, Claremont, California. An oak woodland forms a backdrop to California sage and flowering pinebush (Ericameria pinifolia). (Photo by Travis Beck.) spp.). Living in a low-productivity environment, these shrubs discourage herbivory by filling their leaves with unpalatable aromatic compounds that give the landscape a pungent fragrance. Open ground in between the scrub is home to spring and summer wildflowers. An ecological landscape need not be a slavish imitator of the biome in which it is situated, but the more carefully it responds to the regional climate, the more its structure and features are likely to express what Humboldt called that region’s particular gifts (Darke 2002; Francis and Reimann 1999). Plants Are Adapted to the Seasonal Cycles of the Climates Where They Evolved It is not just the average temperature and the annual precipitation that determine the character of a biome, of course; it is also the seasonality of these factors. Compare the average monthly precipitation and low temperatures in Wilmington, Delaware and Claremont, California (fig. 1.8). Wilmington demonstrates the characteristics of a humid temperate climate: regular, moderate amounts of precipitation and temperatures that cycle with the seasons. Claremont experiences a clas- sic Mediterranean climate, with a cool, wet winter and a warm, dry summer. This is no travel brochure, 16 Principles of Ecological Landscape Design Figure 1.8 Average monthly precipitation and low temperature (a) in Wilmington, Delaware, near Mt. Cuba Center, and (b) in Claremont, California, the location of Rancho Santa Ana Botanic Garden.  Right Plant, Right Place: Biogeography and Plant Selection 17 however. Plants in both locations have to be prepared for the worst of these seasonal extremes: freezing cold winters in Wilmington and bone-dry (zero inches of precipitation in July!) summers in Claremont. The duration of stress—such as heat or cold—may be even more important than a one-time extreme. Plants are adapted to climate extremes anatomically and physiologically, as we have already seen, but also through their phenology, the timing of events such as flowering, setting seed, leafing out, and going dormant. In deciduous forests, such as those found natively around Wilmington, as the days get shorter, signaling the onset of winter, trees and shrubs begin to drop their leaves. Evergreen plants such as American holly (Ilex opaca) and Christmas fern (Polystichum acrostichoides) have thicker leaves with a waxy coating and cell physiology that allows them to survive the winter. Most herbaceous plants survive the coldest months underground, either in perennial roots or as seed. Interestingly, plants from the Mediterranean climate of California exhibit some of the same adapta- tions, only timed differently to handle the stress of summer drought. A few plants, such as California buckeye (Aesculus californica), are deciduous and drop their leaves during the summer drought. Many more plants, such as coast live oak (Quercus agrifolia) and manzanitas (Arctostaphylos spp.) have tough, leathery leaves (sclerophyll foliage) that they can keep year-round, even during a month with no rain. Many California wildflowers (such as California poppy [Eschscholzia californica]) are annuals, which bloom and set seed during the warm, moist spring and then die as the summer drought pro- gresses, surviving underground as seeds until spring returns. The seasonal life cycles of plants are part of what create the character of each biome. Create Seasonal Interest by Showcasing Plant Adaptations to Climate Designers are forever striving to create, and clients are ever clamoring for, year-round interest in the landscape. Nurseries have answered the call with spring-, summer-, and fall-flowering bulbs and pe- rennials, ornamental evergreens of every size and hue, and frost-tolerant pansies in crayon colors. In a more naturalistic aesthetic, such as that of Dutch designer Piet Oudolf, plants are allowed to grow and die back in their own time, offering, among other delights, gorgeous winter tableaux of frost-coated grasses. An ecological landscape can integrate all these elements. The key is simply to recognize the adaptations behind plants’ seasonal displays and match them to the changing environmental condi- tions of a particular site throughout the year. This will ensure the maximum seasonal effect and reduce the inputs and effort needed to maintain plants that are out of sync with the cycles of the environment in which they are planted. By necessity, this will create different landscapes for different places. Looking once again at the Woods Path at Mt. Cuba in Delaware and the coastal sage scrub at Ran- cho Santa Ana Botanic Garden in California, we see that both create seasonal interest by showcasing the life cycles of plants that are adapted to their different climates. On the Woods Path, the highlight of the year is the appearance of the spring ephemerals (fig. 1.9). Called ephemerals because they seem to appear and disappear in a matter of weeks, flowers such as trillium and bloodroot (Sanguinaria canadensis) take advantage of an early-season window in the deciduous forest. When temperatures have risen sufficiently, but before the canopy trees have leafed out to shade the forest floor, these plants emerge, creating a spectacle of showy white, pink, yellow, and red flowers. Through summer, after the ephemerals have set seed and died back, the ground plane of 18 Principles of Ecological Landscape Design Figure 1.9 Before the canopy fully leafs out in the spring, along the Woods Path at Mt. Cuba Center, flowering dogwood, dwarf larkspur (Delphinium tricorne), and wild blue phlox (Phlox divaricata) capture the eye at a lower level. (Photo by Rick J. Lewandowski.) the garden is a leafy, mostly green space. In autumn, the Woods Path turns shades of yellow, orange, and red as the trees of the eastern deciduous forest stop producing chlorophyll in their leaves before shedding them, exposing the other photosynthetic pigments left behind. Come winter, the trunks of the trees and the evergreen rhododendrons and hollies preside over a quiet, sometimes snow-covered ground plane as the herbaceous plants wait dormant for spring.  Right Plant, Right Place: Biogeography and Plant Selection 19 The seasonal cycle at Rancho Santa Ana is very different. In spring the wild lilacs and manzanitas and currants burst into flower, along with annual wildflowers (fig. 1.10). As temperatures soar and precipitation ceases, most plants go dormant, though the evergreen oaks and shrubs maintain their leaves. Seed heads dry on the sages and remain that way through autumn, when they are joined in display by the dramatic flowering of the pinebush. In winter the still green (or gray) oaks, lilacs, and sages soak up moisture and push new growth. These two landscapes create very different spectacles of seasonal interest—spectacles based on the seasonal cycles of plants adapted to two very different climates. Figure 1.10 In early spring at Rancho Santa Ana Botanic Garden, seedheads of pinebush shine alongside flowers of manzanita and currant. (Photo by Travis Beck.) Environmental Differences at Small Scales Create Microclimates So far we have looked at environmental heterogeneity at the broadest scale, that of biomes and climate. An individual plant grows in a particular biome, under a particular climatic regime, but what really matters to its survival are the conditions in its immediate environment. Small-scale varia- tions in environmental conditions create what are called microclimates. A microclimate for a group of trillium could be the shade of a single forest maple; the microclimate for a redwood (Sequoia sempervirens) grove might be several square miles in a deep canyon with moister soils than the hillsides above. Every place in a landscape experiences a combination of environmental variations; thus, one way of understanding a landscape as a collection of heterogeneous patches is to view it as a mosaic of microclimates. In the midst of the Illinois prairie, protected from fire by the Salt Fork of the Vermillion River, a 16-square-kilometer mixed mesophytic forest known as the Big Grove once grew. After settlement, much of this forest was chopped down for development, agriculture, and woodlots. Trelease Woods 20 Principles of Ecological Landscape Design and Brownfield Woods are two remnant parcels of the Big Grove, each about 400 by 600 meters, surrounded variously by roads, farm fields, fences, and a prairie restoration project. These woods served as the study site for Sophia Gehlhausen and her colleagues (2000) to investigate microcli- matic differences between forest edge and forest interior and the effect of those microclimates on understory vegetation. Gehlhausen established linear transects from the north, east, south, and west edges of each forest to the interior. Along each transect she took measurements of relative humidity, air temperature, and soil moisture, and she sampled the herbaceous vegetation, shrubs, and saplings. Gehlhausen’s results are complex, showing the interplay of numerous factors, and demonstrate a clear link between microclimate and vegetation. Moving from forest edge to interior, canopy openness declined, and soil moisture and relative humidity increased in most cases. The direction (aspect) each edge faced and what bordered it made an important difference in the results, especially by affecting the amount of sun and wind to which the edges were exposed (see chap. 9). Vegetation too changed from edge to interior. Samples from the forest edges were least similar to samples from the forest interior, and samples became more similar the closer to the interior they were taken. Gehlhausen suggested that competition (see chap. 3) and disturbance played a role in these differences, but to a large extent vegetation was responding to microclimate. Some species such as wood nettle (Laportea canadensis) and great waterleaf (Hydrophyllum appendiculatum) increased in coverage as one approached the interior. Wood nettle did poorly in full light, and both plants thrive where soil moisture and relative humidity were higher. The more sun- and drought-tolerant black snake- root (Sanicula odorata), on the other hand, decreased in cover toward the interior. Gehlhausen’s results demonstrate that even at the small scale governed by microclimates, environ- mental heterogeneity drives what plants grow where. Match Plants to Microclimates The same factors that govern the distribution of herbaceous plants in Trelease and Brownfield Woods influence what plants will grow best in different areas of a designed landscape. Relative humidity, air temperature, soil moisture, available sunlight, and exposure to wind are all important factors for a plant’s well-being, even when that plant is sufficiently cold hardy or heat tolerant and adapted to the natural precipitation of a region. One of the most detailed efforts to test, observe, and describe the relation between small-scale en- vironmental variables and the success of garden plants has been that undertaken by Richard Hansen and Friedrich Stahl in Germany. In their book Perennials and Their Garden Habitats (Hansen and Stahl 1993: 33) they wrote, Those who have seen the effect of a particular perennial growing wild in its characteristic environment, and compared it with the same plant’s sad destiny in some of our gardens, planted here and there according to some arbitrary conceit, will surely understand this attempt to place perennials in their correct garden habitat.... Conditions vary tremendously within a garden. The correct choice of plants for any given spot requires detailed investigation of the planting position  Right Plant, Right Place: Biogeography and Plant Selection 21 and careful consideration of all possible candidates. If this is done well, then a stable and long- lasting plant community can develop, often requiring just a bare minimum of maintenance, yet forming a convincing feature in the garden design. Hansen and Stahl began by defining the main garden habitats for their region: woodland, woodland edge, open ground, rock garden, border (traditional garden bed), water’s edge and marsh, and water (fig. 1.11). These habitat classifications correspond to sets of microsite variables. Open ground is obvi- ously sunnier than woodland, water’s edge wetter than a rock garden. It is with their more specific clas- sifications, however, that things get really interesting. Within woodland edge, for instance, some of the subcategories they identified are “perennials for shade and bright shade on moist, nutrient-rich soils,” “perennials for sun or bright shade on an open woodland edge (moderately dry, loamy, alkaline soils),” and “perennials for cool, damp, sunny or lightly shaded sites.” As Hansen and Stahl’s experience shows, it is really the set of all these environmental variables for which plants are adapted and according to which they should be selected for each place within a landscape or garden. Figure 1.11 An open, sunny woodland edge. (Adapted from Hansen, R., and F. Stahl. 1993. Perennials and Their Garden Habitats, translated by R. Ward. New York: Cambridge University Press.) Take Advantage of Microclimates in the Built Environment In urban and suburban environments, constructed elements of the landscape can play a role as large as or larger than that of natural features in determining microclimates. For instance, a stand-alone building creates its own patterns of sun and shade, as influential for growing plants as the direction a woodland edge faces. Materials such as concrete, stone, and steel absorb the sun’s energy and re- 22 Principles of Ecological Landscape Design radiate it, raising the temperature of their immediate environment (fig. 1.12). Areas that are surrounded or overhung by structures remain drier than they would if they were out in the open. The edges of road- ways and impermeable parking lots, on the other hand, can have significantly higher soil moisture than the surrounding landscape. Landscape designers must be attentive to these differences in microclimate when selecting and siting plants. We should also look at microclimatic differences as providing horticultural and design opportunities. Planting beds backed by south-facing walls will warm up earlier in the season and stay warmer longer, effectively lengthening the growing season. Paradoxically, marginally hardy plants that would be killed by midwinter soil warming and daily fluctuations in temperatures can be safer in the shade on the north side of a structure, where they will remain safely dormant until spring fully arrives. In semiarid and arid climates, the protection afforded by structures can create a sufficiently cooler and more humid microclimate on north and east exposures to allow plants to be grown that otherwise would not survive. Of course, designers are also not limited to making use of microclimates provided accidentally by architects and engineers. Walls and landscape structures can be built in order to create warmer and cooler microclimates and to serve other functions. Small-scale contouring of the landscape can also Figure 1.12 A thermal image of downtown Atlanta, Georgia in May reveals microclimatic differences in the built environment. Flat rooftops are scorching hot, whereas areas in the shade of buildings remain cool. (Image courtesy of NASA/Goddard Space Flight Center Scientific Visualization Studio.)  Right Plant, Right Place: Biogeography and Plant Selection 23 create drier mounds or wetter pockets where these are desired (see chap. 6). Because at a small scale it is the microenvironment that determines what will grow where, design of the physical site and plant- ing design have to go hand in hand. Ecotypes Are Adapted to Local Environments We have seen how plant species have evolved to grow in particular climatic and microclimatic environ- ments. In fact, adaptation is even more site specific. In the late 1910s and early 1920s, Swedish bota- nist Göte Turesson noted that some plants, such as hawkweed (Hieracium umbellatum), have multiple forms. Where it grows in open woods, hawkweed is erect and broad-leaved. Where it grows in nearby sand dunes the same species is narrow-leaved and less erect. Turesson transplanted specimens of these different forms into a common garden, first at his house and later at the Institute of Genetics at Åkarp, Sweden. When these different forms persisted in the common garden, it was clear that they were not simply morphological changes caused by differences in the environment but true genetic dif- ferences. Turesson (1922) called these genetically distinct groups within a species, each adapted to its local environment, ecotypes. In the 1940s a trio of scientists in California—Jens Clausen, David Keck, and William Hiesey—in- vestigated ecotypes more thoroughly using common yarrow. Yarrow (Achillea millefolium) is one of the most widely distributed plants in the northern hemisphere. In California alone, it grows in nearly every conceivable environment, from seaside bluffs to the arid edges of the Great Basin. Where yarrow grows in the seasonally arid interior foothills it has thin leaves and gray foliage, and it goes dormant in summer. Where it grows in meadows amid conifers on the slopes of the Sierra Nevada, it is midsized with greener leaves, has winter dormancy, and is slow to mature. Where it grows above timberline it is short, frost-resistant, and early to flower after a long winter dormancy. Even when the researchers grew the different ecotypes of yarrow in a common garden at Stanford, these differences persisted (fig. 1.13). Clausen, Keck, and Hiesey (1948) set out to find out what would happen to these ecotypes if they were each grown in the natural environment of the others. In addition to the garden at Stanford, which is located in a low valley between the inner and outer Coast Ranges of central California, they established a garden at Mather, near 1,400 meters in elevation on the western slope of the Sierra Nevada, and another garden they called Timberline, at 3,050 meters. The foothill ecotypes of yarrow developed and flowered quickly at Stanford but performed poorly in the mountain gardens. At Timberline the foothill plants were nearly all wiped out during the first winter of the experiment. The yarrow that were collected in the coniferous belt near Mather grew well at Stanford but reached a greater height and put out more stems at Mather, outperforming any of the other California ecotypes in that garden. At Timberline many of the Mather plants survived and grew, but none successfully set any seed. The extreme alpine ecotype of yarrow grew and flowered at Stanford and Mather but was tallest and best developed at Timberline. Taken together, the results of these experiments demonstrate that in yarrow the adaptations of each ecotype to its local environment increase its growth, reproduction, and overall fitness in that environ- ment, relative even to other plants of the same species. The best-performing plants in each of the test gardens were those collected locally. 24 Principles of Ecological Landscape Design Figure 1.13 Different ecotypes of yarrow and the environments in which they grow in California. (Adapted from Clausen, J., D. Keck, and W. Hiesey. 1948. Experimental Studies on the Nature of Species III. Environmental Responses of Climatic Races of Achillea. Publication 581. Washington, DC: Carnegie Institute of Washington.) Use Local Ecotypes Where Possible Horticulturists have long been aware of the importance of provenance, or the ultimate source of plant material. Especially in wide-ranging species (such as yarrow), the origin of the seed from which plants are grown and even the environment in which they are propagated can make a difference in the estab- lishment and success of the plants. If you are planting in hardiness zone three, for instance, you want to use plants whose provenance is from an equally cold area; otherwise those plants might not survive the winter, even if other members of their species do. Savvy gardeners know that choosing sources with conditions similar to where the plants will ultimately reside increases the odds that these plants will perform well in their new environment. Sourcing plants by ecotype takes this approach to another level of specificity. On the edge of Freshkills Park (the former Fresh Kills Landfill) on Staten Island in New York Harbor, a tidy collection of buildings, greenhouses, and nursery fields houses the Greenbelt Native Plant Center, a project of the New York City Department of Parks and Recreation (fig. 1.14). Greenbelt is committed to growing local plant ecotypes for use in restoration and management projects throughout the city. Ed Toth, Greenbelt’s director, began his work with the production of native plants back in the 1980s for a revegetation project at Prospect Park in Brooklyn. As Toth continued his work, his definition of  Right Plant, Right Place: Biogeography and Plant Selection 25 Figure 1.14 Plots of ecotypic plants being grown at the Greenbelt Native Plant Center. (Photo courtesy of NYC Parks.) native became more and more stringent as he realized that the genetics of local plant populations represent thousands of years of adaptation to particular sites. From a conservation standpoint, these are irreplaceable resources and form the basis for the future evolution of the species. From a restora- tion standpoint, local ecotypes can give the best assurance of success because of how well they are matched to site conditions. To propagate local ecotypes, Greenbelt collectors identify existing, healthy populations of native plant species in natural areas throughout the five boroughs of New York City and, in the case of plants that have been extirpated within the city limits, in nearby New Jersey, Connecticut, Long Island, and upstate New York. Where possible they collect from populations with at least fifty individual plants, in order to capture the local range of genetic traits (see chap. 2). Collectors record local environmental information, such as the type of soil the plants are growing in, the slope and aspect of the site, whether the site is disturbed, and what other plants are growing nearby. Then they visit the site multiple times and collect seed from randomly selected plants. Some seeds are stored in a midterm seedbank; others are sown to produce live plants. At any one time, Greenbelt has 400,000 to 500,000 plants in produc- tion, representing more than three hundred species. When asked what one should do if one is seeking plants for a project and there is no such careful 26 Principles of Ecological Landscape Design nursery nearby, Toth points out, “There wasn’t one here before either!” He acknowledges that we all have to make pragmatic choices to get projects done. The best source of seeds and plants, he suggests, is the site itself, and then one should search outward from there. Under Toth’s leadership, Greenbelt has joined with the Bureau of Land Management, other government agencies, botanic gardens, and nonprofit organizations to support the Seeds of Success program (www.nps.gov/plants/sos). Seeds of Success aims to collect seed from every plant taxon native to the United States and to identify the thousand species most important for restoration and produce multiple ecotypic releases of seed for each of those species to support local restoration efforts throughout the country. Plants from Distant Regions May Have Similar Adaptations While ecotypes of plants are finely adapted to their local environment, in some cases plants from dis- tant corners of the globe have much in common. Since the first botanical expeditions to Japan and China, western botanists have noted the striking similarity between the floras of eastern Asia and eastern North America. Numerous genera occur in these two regions only, with distinct species on the separate continents. Examples include Magnolia, Wisteria, Pachysandra, hickories (Carya), witch hazels (Hamamelis), and mayapples (Podophyllum) (figs. 1.15, 1.16) (Li 1952). These genera are said to be disjunct. Figure 1.15 (left) The North American mayapple, Podophyllum peltatum. Figure 1.16 (right) The Chinese mayapple, Podophyllum pleianthum. Tens of millions of years ago, in the Tertiary period, a mesophytic deciduous forest spanned the two continents. Broadleaf trees and shrubs spread a dense canopy over rich, moist soils across a terrain of hills, mountains, and lowlands. In their shade grew herbaceous groundcovering perennials, and up their branches clambered aggressive vines. At the end of the Tertiary period, however, global cooling made the northern reaches of the continents uninhabitable for these forest plants, and they became separated on the two continents. There were probably several episodes of separation and reconnection across the Bering land bridge between what is now Russia and Alaska. Isolated on distant continents,  Right Plant, Right Place: Biogeography and Plant Selection 27 the species continued to evolve, a process known as vicariance, which has created the separate but related species we know today. Recent studies of the phylogenetic relationships of these plants have complicated this picture, how- ever (Wen 1999). Some of the presumed sister species pairings based on morphological similarities between east Asian and eastern North American species (e.g., in the genus Aralia, the genus Hama- melis, and the Magnolia section Rhytidospermum) have not been confirmed genetically. Although the species are indeed distantly related, the morphological similarities may be seen as examples of convergent evolution. In many cases plants and animals of different genetic origins growing in similar habitats develop similar adaptations. The most famous example of convergent evolution in plants is probably that of the Old World euphorbia family and the New World cactus family, both of which show reduced leaf area and store scarce water in fleshy stems that they protect from herbivores with sharp spines. In the case of woodland herbaceous plants, very broad, thin green leaves that can collect light under a shady canopy have been selected for across many disparate genera and families. This adaptation is shared by the unrelated Asian endemic Japanese wood poppy (Glaucidium palmatum) and the North American en- demic hairy alumroot (Heuchera villosa), for example. Whether because of disjunction and vicariance or convergent evolution, plants are sometimes well adapted to grow in distant but environmentally similar regions. Plants from Other Regions Can Be Included in an Ecological Landscape Disjunct taxa and convergent evolution pose a puzzle for the ecological designer: Is it ecologically justifiable to use well-adapted plants from other regions of the globe? Many aspects of this question are discussed later in the book, including what makes a plant community, ecosystem function, specific plant–animal relationships, the importance of diversity, the threat posed by invasive species, and the consequences of global change. At the level of choosing plants that are suited to their environment, however, it is clear that for every region there are plants from other regions that will grow with a mini- mum of care. Some of these plants are related to our natives, and once they even may have coexisted. Some have developed similar adaptations from evolving in similar climatic regimes on other parts of the planet. The Asian Woods at Chanticleer Garden in Wayne, Pennsylvania demonstrates how well plants from other regions can grow alongside natives (fig. 1.17). Beneath a canopy of red maple and other trees native to the eastern North American deciduous forest, the gardeners at Chanticleer have created a woodland garden filled with plants from China, Japan, and Korea. The structure of the Asian Woods is similar to the Woods Path at the nearby Mt. Cuba Center, with canopy, understory, and an herbaceous ground plane. Only here, among other substitutions, the east Asian kousa dogwood (Cornus kousa) takes the place of the native flowering dogwood. The seasonal spectacle of early spring flowers and intense autumn color is similar as well. Instead of the native trillium emerging through the leaf litter, though, the garden is graced by the early flowers of the Asian buttercup (Adonis amurensis). One would not mistake the Asian Woods for a native landscape. Even without the Asian-inspired ar- 28 Principles of Ecological Landscape Design Figure 1.17 Asian Woods at Chanticleer Garden, Wayne, Pennsylvania. (Photo by Travis Beck.) chitecture and site furnishings, the bamboo and broad-leafed hostas give the garden a decidedly Asian feel. But it is equally a celebration of the temperate deciduous forest, filled with plants well suited to their environment. If one is tempted to cringe at the use to which a remnant Pennsylvania woodland was put, consider that before the Asian Woods was created, this area was overrun by the native trouble- maker poison ivy (Toxicodendron radicans) and a truly invasive Asian honeysuckle. In the right circumstances, adapted nonnative plants can be useful plants to include in an eco- logical landscape, whether for their toughness in urban conditions, their resistance to disease, their ability to perform needed functions (see chap. 4), or their aesthetic appeal. Several of the plants included in the yard of Panayoti and Gwen Kelaidis, for example, are not Colorado or even North American natives. It may not be desirable, or even possible, to grow pure communities of native plants in every landscape situation (see chap. 10). Consider, for instance, the work of the American Chestnut Foundation (2012), which is close to being able to reintroduce chestnuts that are resis- tant to the Asian fungus that decimated their stands in Appalachian forests, thanks to a breeding program using genes from the Chinese chestnut (Castanea mollissima). Both the cause of and the likeliest solution to chestnut blight stem from the close relation between eastern North American and east Asian forests.  Right Plant, Right Place: Biogeography and Plant Selection 29 History of Dispersal Affects Biogeography As the last section suggests, what plants grow where is not simply a matter of adaptation to the envi- ronment. Geologic and landscape history play roles as well. It is well understood that North America has experienced a series of massive glaciations in recent geologic history. We saw how these glacia- tions separated plants from a once-extensive mesophytic temperate forest into disjunct eastern Asian and eastern North American groups. The most recent glaciation was the Wisconsin, which reached its maximum about 18,000–20,000 years ago. As the massive glacier retreated over the course of a few thousand years, it left in its wake familiar features such as the terminal moraine we know as Long Island and the giant pools of meltwater called the Great Lakes. Less well understood is the instability that the retreat of the glacier has caused in the vegetation of the region. Margaret Davis has made a career of studying changes in northeastern forests. She has gained insight into the history of these forests by examining grains of pollen trapped in layers of sedi- ment at the bottom of lakes. Year after year the pollen of whatever wind-dispersed plant species are nearby drifts onto the surface of a lake, sinks to the bottom, and becomes trapped in the accumulating mud. By taking deep cores of this sediment, radiocarbon dating the layers, and counting the pollen grains of different species found in each layer, Davis and other researchers can identify the shifting abundance of different species. One might imagine that as the glaciers retreated, bands of tundra, boreal, and temperate forests would follow the improving climate north. In fact, however, individual species migrated by different paths and at different rates (Davis 1981). Hickory and beech (Fagus grandifolia), which grow together in eastern North American hardwood forests today, appear to have survived the last glaciation in the lower Mississippi valley. From there hickory migrated quickly through the Midwest, reaching areas of Minnesota and Michigan as early as 10,000 years ago, but took another 5,000 years to penetrate New England. Beech, on the other hand, migrated east of the Appalachian mountains and arrived in Upper Michigan only 3,500 years ago (see fig. 1.18). Davis (1981: 152) concluded, Much of the time the rate of spread [of tree species] was not controlled by climate, and the geographic distributions of many species were not in equilibrium with climate, depending instead on the availability of propagules and the ability of seedlings to survive in competition with plants already growing on the site. In other words, the distribution of plants is affected not only by climate but by the history of their dispersal. Dispersal is simply the expansion of the geographic range of a species as individuals move into new areas. Biogeographers recognize several types of dispersal (Morrone 2009). Geodispersal is the si- multaneous movement of many species brought about by the removal of a geographic barrier, followed by the emergence of a new barrier that creates vicariance. Geodispersal was involved in the evolution of the eastern Asian and eastern North American forests. Diffusion is the gradual movement of a spe- cies into suitable habitats over several generations. The migration of tree species after the Wisconsin glaciation is an example of diffusion. Jump dispersal was promoted by Darwin, among others. It is the 30 Principles of Ecological Landscape Design random overcoming of a geographic barrier, as when a seed washes up on an island. Together with vicariance, dispersal helps explain the historical aspects of why plants grow where they do. Figure 1.18 Migration of (a) hickory and (b) beech from the lower Mississippi Valley after the retreat of the Wisconsin glaciation. Hatching indicates the species’ current range. Lines show extent of range at different times. Numbers indicate thousands of years ago. Hickory reached the upper Midwest 10,000 years ago but extended into New England only 5,000 years ago. Beech migrated east of the Appalachians and reached Upper Michigan only 3,500 years ago. (Redrawn from figs. 10.12 and 10.13 from Davis, M. B. ©1981. Quaternary history and the stability of forest communities. In Forest Succession: Concepts and Applications, edited by D. C. West, H. H. Shugart, and D. B. Botkin, 132–153. New York: Springer Verlag. With kind permission of Springer Science+Business Media.) Be a Mindful Agent of Plant Dispersal Throughout this chapter we have discussed the importance of matching the adaptations of selected plants to their intended environment. As Davis’s research showed, however, the natural distribution of plants does not represent a perfect balance between climate and adaptations. Instead, it is the result of interacting factors (availability of propagules, chance dispersal, ability to establish, competition) taking place in a changing space. Before an adapted plant can grow and reproduce somewhere, it has to be brought to that location, whether by the forces of gravity, wind, or water or with the help of animals. Animals of all sorts, from birds with berries in their guts to bears with burrs on their legs, are responsible for dispersing plants (see chap. 7). Every time we humans select plants for a landscape or garden and decide where to get these plants and where to plant them out, we are also acting as agents of plant dispersal.  Right Plant, Right Place: Biogeography and Plant Selection 31 In some cases, we can use the built landscape to support and direct the diffusion of desired plant populations. In restoration plantings designers often reintroduce native plant species that were histori- cally present on a site or species that are currently present in a reference site (see chap. 10). This type of dispersal can knit together disconnected populations and help create continuous corridors of habitat (see chap. 9). In these types of plantings, where it is anticipated that new plants will actively interbreed with surrounding populations, attention to plant genetics is especially important. In addition to filling gaps in species’ ranges, one also might attempt to expand the edges of a species’ current range, to see whether it could be done for horticultural or design purposes or in anticipation of climate change (see chap. 10). Planting out nonnative garden and landscape plants with a long and innocu- ous history in a region is another example of diffusion dispersal, one that simply further establishes populations of these plants in the built environment. In other cases, planting a site involves the jump dispersal of exotic plants, whether these are conti- nentally native plants that are not locally present or, as we discussed in the last section, adapted plants from other regions of the globe. Ecologically, this is a more uncertain form of dispersal, both because it is less clear how well an exotic plant will integrate with the flora and fauna of a local area (see chap. 7) and because of the threat of invasiveness (see chap. 5). For newer introductions, it pays to do one’s homework and to consider limited introductions at first into a controlled setting. Dispersing plants is an ecological act, one performed by humans for millennia with various conse- quences, from spreading useful plants across the globe to introducing diseases that wipe out native species. It is an act worth reflecting on fully when selecting plants for an ecological landscape. We should neither disperse plants blithely nor be afraid to try anything new. Conclusion Biogeography gives us a first lens through which to view landscapes ecologically. Plants in nature grow where they do for a variety of reasons, yet they all grow where they do without us needing to care for them. They accomplish this because, through various adaptations, they are suited to the environmental conditions of the regional climate and the local microclimate. They grow and senesce, flower and set seed, tough out winter and drought with durable leaves, or go dormant and reemerge in concert with the cycles of their area. When we design in harmony with the environment, we take advantage of these adaptations in our plant selection, creating landscapes that are in keeping with the character of an area not from specific design intent but because we are following the same rules as nature. Following nature’s rules can only make our job as caretakers easier. On one hand, we should maximize our use of local adaptations by planting local ecotypes where possible. On the other hand, we should recognize that not all the plants that could grow well in our particular environment currently do, that what is native to a given locale changes over time, and that dispersal is a natural process. Using adapted exotic plants in a thoughtful manner can be compatible with an ecological approach to design. 2. Beyond Massing: Working with Plant Populations and Communities So far we have discussed how factors such as climate, microclimate, and place of origin influence why individual plants grow where they do in nature and how those factors should influence our selection of plants for the landscape. Of course, in nature and in the built environment plants rarely grow alone. They grow with other plants, both from their own species and from others. Traditionally, in landscape and garden design, we call a substantial group of the same species a mass and the selection of plant species and varieties brought together for a project a plant palette. These terms reveal the aesthetic basis of how plants are conventionally grouped: to occupy a space, as a mass, and to work together harmoniously, as the colors chosen and mixed on a painter’s palette. There is more to grouping plants than aesthetics, however. The demographic lens through which ecolo- gists view assemblages of plants offers practical insights into how plants grow together and how we can group them in our designs. In ecology, a group of individuals of a species living together in an area is called a population (Silvertown and Charlesworth 2001). An area is usually arbitrarily defined by the ecologist studying the population; it could be a measured research plot, an identifiable patch of habitat such as a meadow or copse of trees, or an entire watershed. However an area is defined, the individuals of a plant popu- lation within it are likely to interact—to interbreed, to compete for localized resources, in some cases to share resources via root grafting, even to communicate when attacked by pests. Plant populations, ecologists tell us, have a variety of characteristic structures: spatial structure, genetic structure, size structure, and age structure (Hitchens 1997). These structures are not stable but shift over time. In an ecological landscape design, we seek to bring ecologists’ understanding of how plant populations are structured, and how they change, to the selection and placement of plants in the built landscape in order to set up landscapes with internal dynamics that don’t require regular high levels of maintenance in order to remain full and healthy. A community, in ecology, consists of populations of different species coexisting in a common envi- T. Beck, Principles of Ecological Landscape Design, 33 DOI 10.5822/978-1-61091-199-3_2, © 2013 Travis Beck 34 Principles of Ecological Landscape Design ronment. We can all think of recognizable groupings of plants in recognizable habitats: Atlantic coastal sand dune communities, for instance, with their beach grass (Ammophila breviligulata), beach pea (Lathyrus japonicas maritimus), beach plum (Prunus maritima), and seaside goldenrods (Solidago sempervirens). Even the names suggest these plants grow together in the same place. The concept of plant communities is intuitive yet hard to pin down scientifically. Still, it is one of the most fruitful areas for landscape designers to explore, whether we seek inspiration from communities observed in nature or attempt to make novel use of the principles that govern the commingling of plants. We can be guided in these efforts by the findings of ecologists regarding the number of species that occur in communities and the numbers of individuals within each species’ population, and more importantly, by the mechanisms behind these numbers. For our designs to function as ecological landscapes, we have to look beyond the sculptural and visual qualities of plants to the ways in which they grow together. When we design, we have the oppor- tunity to set up self-perpetuating groups of plants that respond to a site, to each other, and to changes that take place over time. To accomplish this, however, we have to understand the factors that drive the locations, sizes, genetics, numbers, and proportions of different plants. In other words, we have to see plant groupings as populations and communities. Plant Populations Have a Spatial Structure Naturally occurring plant populations are often arranged in striking ways: the dense blanket of color of a field of bluebonnets (Lupinus texensis), for instance, or the regular trunks of a stand of pitch pine (Pinus rigida), or the sweeps of grasses in a meadow (fig. 2.1). What explains these spatial patterns? As we have seen, small-scale environmental differences play a significant role in determining what plants grow where. Patches of certain plants may grow on dry mounds, in wet depressions, or in areas of greater soil depth, creating a spatial structure that reflects environmental differences. Scattered plants may establish and mature only where safe sites exist, creating a dotted pattern (see chap. 8). Layered on top of this environmental control of plant distribution are other factors, such as the location of parent plants and their means of spread (Greig-Smith 1979; Kershaw and Looney 1985). Seedlings from parent plants occur most frequently near the parent. When a plant dies, therefore, the most likely plant to take its spot is one of its own offspring, or offspring from a neighbor. Many plants also spread vegetatively. For instance, stoloniferous redtwig dogwoods (Cornus sericea) grow to create dense patches. Suckering plants such as beeches and redwoods create groves. Plants such as clover may spread more loosely, popping up between other plants. Taken together, these phenomena help explain the clustered patterns of many plant populations. Other factors that affect the spatial structure of plant populations include competition (see chap. 3), the activities of animals (see chap. 7), and patterns of disturbance (see chap. 8). In all cases, the spatial pattern of a plant population is a reflection of underlying ecological processes. Develop Planting Patterns in Concert with Ecological Processes In the Meadow at Longwood Gardens in Pennsylvania, plants are managed as populations, and their spatial patterning reflects the myriad ecological processes that act on them, as well as human intention.  Beyond Massing: Working with Plant Populations and Communities 35 Figure 2.1 Populations of grasses, rushes, and other herbaceous plants form evident spatial patterns in the foreground of this wet meadow in Roxborough State Park, Colorado. (Photo by Travis Beck.) First of all, the populations of the various plant species reflect the environmental heterogeneity of the site. The hillside on which the Meadow grows creates a range of conditions, from drier, less fertile areas at the top of the south-facing slopes to wetter, richer conditions at the bottom. In the richer areas, plants such as joe-pye weed (Eutrochium maculatum) have established strong masses. In the drier areas grow plants such as butterfly weed (Asclepias tuberosa) and the locally rare Elliott’s broomsedge (Andropogon gyrans). Plants’ means of spread is another important factor. Joe-pye weed forms solid masses by means of aggressively spreading underground rhizomes. Seeds from little bluestem grass (Schizachyrium scoparium) can spread short distances on the wind, thereby pushing out the edges of existing masses, if conditions allow. The sawtooth sunflower (Helianthus grosseserratus) spreads so aggressively from seed that Longwood’s land managers have had to use herbicide to knock it back. Longwood’s management of the site also affects spatial patterning. Various species have been introduced to the Meadow over time, both from seeds and from plugs (small rooted plants). Where they established well, their presence reflects human intention. Rotational burning of the Meadow favors certain species, such as Canada goldenrod (Solidago canadensis), in recently burned areas. Where land managers remove invasive species, they replant, so the latest plantings reflect the spatial patterns of invasive species establishment but with more appropriate plants. 36 Principles of Ecological Landscape Design The combination of all these factors has resulted in a diverse community in which clear bands and masses of plant populations are discernible. The spatial patterns that have emerged reflect natural processes layered with a history of human interventions. They are both highly ecological and aestheti- cally striking (fig. 2.2). Figure 2.2 Populations of switchgrass (Panicum virgatum) and little bluestem form distinct spatial patterns in the Meadow at Longwood Gardens. (Photo by Travis Beck.) Plant Populations Are Genetically Diverse Plant populations are made up of individual plants of the same species, from dozens to thousands of them, each with its own characteristics. Every individual in a population differs from the others in any number of ways—age, size, timing of flowering, tolerance of heat and cold, resistance to disease, just to name a few. Just as local ecotypes can be found within a species, within a population individual variation can be seen. To the extent that this variation has a genetic basis, it is a fundamental basis of plant biodiversity (see chap. 5). Genetic diversity is maintained most apparently in populations through sexual outcrossing. When one plant in a population exchanges pollen with another plant to create a fertilized embryo within a seed, the genes the two plants carry are shuffled together. The offspring that result from this and similar outcrossings are then slightly genetically different from either of their parents. Interestingly, however, diversity within plant species persists despite the fact that many plants  Beyond Massing: Working with Plant Populations and Communities 37 produce genetically identical or closely related offspring and many others spread clonally (fig. 2.3). Dandelions (Taraxacum officinale) often use apomixis (in which embryos develop without fertilization having occurred) to produce seeds that are genetically identical to the parent. Self-fertilization, which produces seeds containing genes from only one parent, is most common in annuals such as the weedy common groundsel (Senecio vulgaris). Aspen (Populus tremuloides) suckers from its roots to create large stands of trees, as is evident in the fall when all the trees of a single clone turn color simultane- ously. In all these cases genetically identical or closely related plants can spread quickly through an open area to which the parent plant has already proved to be adapted (Harper 1977). Figure 2.3 Aspens sucker to create stands of genetically identical trees. Several such stands together form an aspen population of moderate diversity. (Photo by Mark Muir, US Forest Service.) Ecologists have found that even in cases of asexually reproducing plants such as aspen and dan- delion, local populations contain intermediate levels of genetic diversity (Ellstrand and Roose 1987). In these populations all the individuals that share the same genes are called a genet. Each physically separate individual of a genet is a ramet. Populations of asexually reproducing plants achieve interme- diate levels of diversity by consisting of several co-occuring genets. Thus, a population of strawberries, for instance, might consist of seven hundred separate plants, one hundred ramets of each of seven genets. Whether working on sexually or asexually reproducing plants, then, natural selection seems to maintain genetic diversity in local populations. 38 Principles of Ecological Landscape Design There are several advantages to genetic diversity (Falk et al. 2006). One is that heterozygous plants (those with different alleles for a given trait) tend to perform better. Diverse populations avoid the effects of inbreeding depression, where negative mutations are more commonly expressed. Also, in di- verse populations there is a subtle differentiation of adaptations down to the level of individual plants. This allows the population to take advantage of the smallest environmental variations on the site and slight differences in the ability to coexist with other species. This diversity can be especially important if environmental variables fluctuate or change directionally on a site over the lifetime of a generation of plants, which is perhaps why most longer-lived woody plant species reproduce through outcross- ing. Finally, diverse populations are able to better withstand pressure from pests and pathogens. The population presents a more varied target, and resistant individuals may be able to spread their genes more widely in future generations (see chap. 7). In sum, genetic diversity helps a plant population on a given site survive, adapt, and persist through time. Create Diverse Populations For long-term stability in the face of environmental fluctuations, broad-scale resistance to pests and pathogens, and ability to continue to evolve, genetic diversity in a plant population is essential. This is as true in the constructed landscape as it is in nature. To benefit from these advantages of intraspecific diversity, designers should use cultivated varieties selectively, consider developing regional or site- specific landraces, and insist on broad-based collection of seed for the production of straight species for landscape use. Plant breeders have selected, developed, and named a wide array of cultivated varieties of many popular landscape plants. Consider the coneflowers. In addition to the standard North American native Echinacea purpurea we now have E. purpurea ‘Magnus,’ ‘Little Magnus,’ ‘Kim’s Knee High,’ ‘Kim’s Mop- head,’ ‘Ruby Giant,’ ‘Vintage Wine,’ ‘Double Decker,’ ‘Fatal Attraction,’ ‘Pink Shuttles,’ and ‘Pink Poodle,’ to name just a few. The goal of the cultivated variety is to reduce variability and produce uniformity of certain traits. Take E. purpurea ‘Magnus,’ for example. Whether propagated sexually by seed or asexually by root cuttings, the result is a plant with a strong, deep pink flower color and wide petals in a flat disk. This artificial selection greatly reduces the natural genetic variation found in a sexually reproducing population of the straight species. Some of the more colorful coneflower cultivars are produced by hybridizing purple coneflower with the yellow Echinacea paradoxa. Offspring of these crosses are then themselves crossed, to produce plants such as Echinacea ‘Sunrise,’ ‘Sun Down,’ ‘Summer Sky,’ ‘Twilight,’ and ‘Harvest Moon.’ Crossing plants in this way can create genotypes with so-called hybrid vigor, resulting from a new combination of beneficial dominant genes. Once a promising offspring of the cross is identified and selected for release, however, it is reproduced asexually through divisions, cuttings, or micropropagation, which means each plant is genetically identical to the original parent. Certainly there are advantages to the use of cultivars. Uniformity of size, shape, and color can help achieve design intent. Many cultivars are selected for their disease resistance, which can help plantings avoid or survive infection. Asexual annual weeds such as dandelion and suckering early-successional plants such as aspen suggest that in the right situation there are ecological advantages to using clones.  Beyond Massing: Working with Plant Populations and Communities 39 In a uniform site where a particular plant is known to do well, a fast-growing cultivar can provide quick, reliable coverage. Even in these situations, however, ecological designers would do well to create mul- ticlonal populations of several related cultivars, to achieve the intermediate level of diversity found in natural populations of clonal plants. Intermediate levels of diversity in a population can also be created by applying the agricultural concept of a landrace. Landraces are less formalized breeds of animals or varieties of plants that are adapted to local environmental and cultural conditions. They are distinctive in some manner, such as being early or late fruiting, but because selection is less rigorous than in the development of a modern named variety, landraces contain a higher level of genetic variability. Landraces of ornamental plants are not readily available but can be created by any land manager in one of two ways. One can select from an existing population of a species on the basis of flower size, disease resistance, or whatever characteristics one chooses and only allow certain plants to reproduce. Alternatively, one can plant multiple cultivars, or a cultivar and the straight species from which it is derived, and allow them to interbreed and again select the preferred plants to continue the experiment. The landrace approach allows the manager of a population of plants over time to combine the benefits of local adaptation and genetic variability with selection for particular characteristics. To achieve high levels of genetic diversity within a population, however, it is necessary to use straight species rather than cultivars or landraces. It is also important that these plants be propagated by seed, rather than vegetative methods, to allow for sexual outcrossing and the accompanying genetic shuffling to occur. Beyond that, it is important to know how the seeds from which these plants are produced were collected (fig. 2.4). Growers often take many seeds from a single notable parent plant Figure 2.4 Collecting seeds from brownfoot (Acourtia wrightii) in Arizona. Randomized collection over a period of time increases the genetic diversity of propagated plant material. (Courtesy of BLM/UAH, Seeds of Success.) 40 Principles of Ecological Landscape Design and then increase their stock from this limited initial collection. Where a small number of specimen plants are needed, selection from notable individuals can be appropriate. To achieve diversity, however, broader collection is necessary. Geneticists now recommend collection techniques such as randomized sampling over a period of time in order to capture 95 percent of the genetic diversity of local popula- tions (Guarino, Ramanatha Rao, and Reid 1995). Only when all these methods are applied can the plant populations of our constructed landscapes approach the beneficial levels of diversity found in natural plant populations. Populations Include Individuals of Different Sizes and Ages In designed landscapes, plants of the same species are often the same size. In some cases this is sim- ply a byproduct of redoing an entire area at one time, as when all the trees in a new park are planted the same year. In other cases it is a