Dynamics and restoration of Pilgerodendron uviferum forests on Chiloé Island, North Patagonia, Chile

Transcription

1 Dynamics and restoration of Pilgerodendron uviferum forests on Chiloé Island, North Patagonia, Chile Thesis submitted in partial fulfilment of the requirements of the PhD degree of the Faculty of Forest and Environmental Sciences, Albert-Ludwigs Universität, Freiburg im Breisgau, Germany by Jan Richard Bannister Hepp Freiburg im Breisgau, 2012

2 Dean: Prof. Dr. Jürgen Bauhus First examiner: Prof. Dr. Jürgen Bauhus Second examiner: Prof. Dr. Michael Scherer-Lorenzen Third examiner: Pablo J. Donoso PhD. Date of thesis defense: 16 November 2012 Gedruckt mit Unterstützung des Deutschen Akademischen Austauschdienstes

3 Statement of Originality I hereby declare that this thesis has never been submitted to another examination commission in Germany or in another country for a degree in the same or similar form. The material in this thesis, to the best of my knowledge, contains no material previously published or written by another person except where due acknowledgement is made in the proper manner. Jan R. Bannister Hepp 27 th July 2012

4 Statement of contributions to this PhD Thesis This doctoral thesis includes seven chapters. Among them, three chapters were developed to be published as peer-reviewed papers, including: Chapter 2: Bannister JR, PJ Donoso, J Bauhus Persistence of the slow growing conifer Pilgerodendron uviferum in old-growth and fire-disturbed southern bog forests. Published in the journal Ecosystems 15: Chapter 3: Bannister JR, S Wagner, PJ Donoso, J Bauhus The importance of seedtrees for passive restoration of Pilgerodendron uviferum in fire disturbed southern bog forests. To be submitted. Chapter 4: Bannister JR, RE Coopman, PJ Donoso, J Bauhus The importance of micro-topography and nurse canopy for successful restoration planting of the slow-growing conifer Pilgerodendron uviferum. Published in the journal Forests 4: This research was carried out between October 2008 and July All results of this thesis are the result of my own analyses. I conceived, designed and performed the research, analyzed the data and wrote all the manuscripts, general introduction and synthesis of this thesis. This work was partially supported through field assistants during the three fieldworks in Chiloé Island (2009, 2010 and 2011), laboratory technicians and two master students. Prof. Dr. Jürgen Bauhus, was my principal supervisor, conceived and designed the research and was involved in all the stages of this study, including the development of the experimental design, structuring of the thesis and improvement of manuscripts. Pablo Donoso PhD was my second supervisor from the beginning of the research and helped with the design of the research, supported the fieldwork logistics and was involved in the improvement of the manuscripts. Prof. Prof. Dr. Sven Wagner and Dr. Rafael Coopman participated as coauthors in the third and fourth chapters, respectively.

5 For my parents Valeska and William, who are a model of courage for me, and who have constantly supported and believed in my dreams. For my grandfather Ricardo Hepp Dubiau, who accompanies me from heaven. He taught me to love nature and the south. For my Natalia, who taught me to see the beauty in the little living beings of nature and life.

6 The Road goes ever on and on down from the door where it began. Now far ahead the road has gone, and I must follow, if I can, pursuing it with eager feet, until it joins some larger way where many paths and errands meet. And whither then? I cannot say (Bilbo Baggins) The writing phase of this project was inspired by the music of the Icelandic composers Jón Birgisson, Kjartan Sveinsson, Orri Dýrason and Georg Holm of Sigur rós.

7 Acknowledgments First of all, I would like to thank God for permitting me to live my dreams the last 32 years and for sending me to be born and raised in an incredible family. Thanks to the trees, mosses, small animals, spirits and strange beings of the Chiloé forests, who permitted me to work in their territories, under their attentive look in relative peace and harmony. During the last four years, many people have helped me with my doctoral studies in Freiburg. First, I would like to thank my Doktorvater Prof. Jürgen Bauhus, who believed in the idea of this project four years ago. He has trusted and supported me from the beginning of this doctoral journey, encouraging, stimulating and inspiring me throughout the whole process. He has also opened and expanded my scientific curiosity. I will never forget the good moments with you especially the rainy days in Tantauco Park. I also would like to thank my second supervisor Prof. Pablo Donoso of the Universidad Austral de Chile. He was always there when I needed him, especially during my fieldwork in Chile. His practical way of thinking and our conversations about Chilean forestry were a permanent inspiration and motivation during my doctoral studies in Germany. I am especially grateful to Prof. Antonio Lara at the Universidad Austral de Chile for introducing me long ago to the world of Pilgerodendron uviferum forests and for all our inspirational conversations over the last eleven years. I wish to express my gratitude to Prof. Albert Reif and Prof. Benno Pokorny from the Institute of Silviculture for their generous and useful suggestions and constant help during my stay in the institute. Additionally, I would like to thank Prof. Michael Scherer-Lorenzen from the Institute of Biology for being my second reviewer and Prof. Sven Wagner from the Technische Universität Dresden for introducing me to the single tree approaches. Also I thank Patrick Pyttel for translating the summary of this work in German and Nathan Briggs, David Forrester and Katie Pratt for improving the English in early drafts and sections of this thesis. This work was financed by a DAAD-CONICYT scholarship that supported my doctoral studies at the University of Freiburg. I am especially grateful to Lucía Montero from the CONICYT office and Britta Bauch from the DAAD office, who were always ready to help and answer questions. The three field visits in Chile were also financially supported through grants by the Georg-Ludwig-Hartig and Futuro foundations. I am especially grateful to the administration and staff of the Tantauco Park for constant help and support in the field. Especially, I want to thank Alan Bannister and Andrés Caracciolo, who were always there when I needed them, and for their faith in this project from the

8 beginning. They have believed in all the crazy experiments, measurements and activities that I had to do in the park. Importantly, I want to thank my faithful partners who helped and supported me under very difficult conditions including the long and rainy days in the field when they were completely wet and cold: Natalia Carrasco-Farias, Andrés Caracciolo, Rodrigo Ramirez, Georg Löffler, Jonas Flade, Daniel Rieck, Guillermo Fullá and Álvaro Gutiérrez, and many forest rangers of the Tantauco Park. Without your help this study could not have been done and I think that in the end the sacrifice was worth it. For assistance in the laboratory I would like to thank Germar Csapek, Nathan Briggs, Georg Löffler, Matthias Schmidt, Tiemo Kahl, Osvaldo Vidal, and especially, Renate Nitschke for her help and patience in the lab. I am also grateful to David Forrester, Tiemo Kahl, Hendrik Stark, Stefanie Gärtner, Rodrigo Vargas, Daniel Uteau and Daniel Soto who helped me with statistical analyzes during the last years and Andrés Holz for interesting discussions. During my studies in Freiburg I have participated in the graduate school Environment, Society and Global Change of the Faculty of Forest and Environmental Sciences. I would like to thank those involved in this programme, especially the programme coordinator Esther Muschelknautz for all her help during the last years organizing seminars, courses and other activities. One of the greatest and most motivating moments of this PhD was the trip to the IUFRO Conference in South Korea. Thanks to all the members of the GS who were involved in that magical trip that I will never forget. To spend all this time in the Waldbau-Institut in Freiburg was an incredible and enjoyable experience; it is difficult to imagine a better place to study than in this institute. I am grateful to all of you. Especially, I am grateful to Cristabel, Hendrik, Somidh, Tamalika, Patrick, Katja, Pifeng, Tiemo and Thomas for nice moments, conversations and discussions during my stay in the institute. I would also like to thank the secretaries of this Institute, Ursula Eggert and Bernardette Trautwein, for their patience because without them many daily problems would not have been so easy to resolve. Thanks for always having such lovely smiles. Also I thank Mathias Frowein for liven up the nice Kaffeepausen. An important piece of my gratitude belongs to my dear roommate Chunling Dai. She was like a second mother for me, she was in our office each day with a smile, especially in bad and desperate moments when there was no motivation to work. One day I will show you and your family the dolphins and blue whales of Chiloé. In the future, when I remember the last months of the fight for submitting this thesis, I will always remember the support of Somidh and Osvaldo, my fight partners: Yes we could!

9 Life is not only work and Freiburg would be very different without all the good friends that I met during the last 4 years. Thanks for sharing a little piece of your lives with me: Bárbara, María Paz and Iván. My good kayakers friends Berti, Matze and Georg, showed me the unique beauty of kayaking during the cold winter in the snow-covered Schwarzwaldbäche or during the hot summer in the Alpine rivers. The time will come when we will paddle in the Andes! I want to thank my good Chilean friends Juanita, Rodrigo and Osvaldo, and also my good Greek friend Dimitrios. We have shared a wonderful life here in Freiburg, I do not have any doubt that we will continue our friendship in Chile where we could do a lot together! You were my second family in Freiburg, you were my baseline, and that is may be the biggest treasure that I discovered in Germany and that I will take with me to Chile. Without the support and love of my parents Valeska and William and my brothers Alan and Mark, my Freiburg trip would not have functioned. Especially, I want to thank my small godsons Vicente, Felipe and Anton. These years in Germany I could not give you all the love and time that you deserved. The times will change and soon I will be with you Finally, I want to thank the most important person in this whole trip, my beloved Natalia, who accompanied and shared with me each day of this European adventure, full of magical moments between Icelandic volcanoes, geysers and trolls, North Scandinavian landscapes and Vikings, and medieval cities and temperate forests in Central Europe. We have a lot of new horizons still to discover and innumerable histories to tell.

10 Preface We were exhausted; the last hours my brother Mark and I were fighting with our machetes against a dense understory of Chusquea quila, one of the Chilean bamboo species. Our dream as young explorers was to build a small path between our family s small hut, in the northeast side of the Tepuhueico Lake peninsula in Chiloé Island, and the southwest side. After some hours of ascending the undisturbed densely forested peninsula in southwest direction, we came to an old forest dominated by Saxegothaea conspicua. We saw the canopies and forms of the trees and we were impressed, they were so similar to Treebeard, the eldest of the Ents of the Fangorn forest in Tolkien`s book Lord of The Rings. We rested only briefly, because our objective was much more to the southwest, we had to continue! After a few hundred meters we finally came to a flat area that we knew was the top of the peninsula. We thought the most difficult part of the expedition was finished, but then we noticed that in this flat area, the forest was not so high but much more dense. We had to begin to fight not against bamboos but against the so called tepuales. Many old chilotes had told us about these tepuales, but there are no words to describe this intricate three meter high net of trunks. These tepuales are so intricate that the ground under the legs is up to 2 meters over the real ground. In this type of forests the machetes do not function at all, we had to creep over the trunks, step by step, sometimes going through dark tunnels between the trunks. We were exhausted; finally we had to stop in order to come back before it got dark. While we were taking a short break, before to continue back to the hut, I saw a small seedling that was not like the others near my hand. It was a conifer and after some seconds I knew that it was a Guaitecas Cypress. Suddenly I lifted my head and realized that we were in the middle of an old undisturbed Pilgerodendron uviferum dominated forest. It was a magic moment; we were the first humans to be in that place, the paradise of mosses and lichens, of small birds and leeches. It was strange, while I was seeing the old trees, the mosses, and the lichens, I felt something rare inside me. But we had to go back, the night was coming That year was 1999 and thirteen years have passed since then. It was the first time that I saw a real and unburned forest of the species. After that I decided to research Pilgerodendron uviferum forests, and I have been committed to this endeavor ever since. This doctoral thesis was inspired by that magic moment in the forest.

11 Contents SUMMARY... 1 ZUSAMMENFASSUNG... 4 RESUMEN... 7 CHAPTER ONE Introduction Peatlands and bog forests Pilgerodendron uviferum bog forests Resilience of Pilgerodendron uviferum forests to fire Restoration approaches Objectives and hypotheses Study area CHAPTER TWO Persistence of the slow growing conifer Pilgerodendron uviferum in old growth and firedisturbed southern bog forests Abstract Link to the article in journal Ecosystems CHAPTER THREE The importance of seed-trees for passive restoration of Pilgerodendron uviferum in firedisturbed southern bog forests Abstract Introduction Methods Study area Seed and seedling dispersion around P. uviferum seed trees Germination experiment Spatial distribution of seed trees at the landscape level Results Dispersion of seeds and seedlings of P. uviferum around seed trees Seed Germination Spatial distribution of seed trees at the landscape level Discussion Capacity of Pilgerodendron uviferum to recover burned areas Is passive restoration a suitable approach for disturbed P. uviferum forests? Acknowledgments CHAPTER FOUR The importance of micro-topography and nurse canopy for successful restoration planting of the slow-growing conifer Pilgerodendron uviferum Abstract Link to the article in journal Forests

12 CHAPTER FIVE Synthesis Stand dynamics in Pilgerodendron uviferum old-growth forests and the effect of fire-disturbances The natural rate of recovery of Pilgerodendron uviferum forests after firedisturbances Towards a restoration strategy for Pilgerodendron uviferum bog forests in North Patagonia Recommendations for further research Undisturbed P. uviferum bog forests Ecological restoration in disturbed P. uviferum dominated bog forests CHAPTER SIX References CHAPTER SEVEN Appendices... 73

13 SUMMARY ZUSSAMMENFASSUNG - RESUMEN SUMMARY While the majority of peatlands and wetlands are concentrated in the boreal zone, millions of ha of these ecosystems are also found in subantarctic areas of the southern hemisphere. In spite of the extensive area of bogs in the southern cone of South America, there have been very few studies on structure and dynamics of conifer bog forests in this region, although many of them have been greatly changed by humans. Southern bog forests dominated by the conifer Pilgerodendron uviferum in Northern Patagonia are a typical case of an ecosystem with low resilience to disturbance by fire, which kills most trees and seeds. These forests, which have been burned to a large extent, cover an area of almost 1 million ha (6.8% of Chilean native forests). The aim of this doctoral thesis was to study fire-disturbed and undisturbed old-growth P. uviferum bog forests in North Patagonia in order to develop the scientific ecological basis for conservation and restoration strategies. Therefore the work focussed on three important aspects of restoration: a) understanding the processes occurring in undisturbed old-growth forests, b) analysing the natural rate of recovery of fire-disturbed forests, and c) exploring some options for restoration. Using a multi-scaled approach, this study started analysing diameter and age structure, foliar and soil nutrient levels and the light environment in old-growth and fire-disturbed P. uviferum stands on Chiloé Island (43ºS) in North Patagonia. Secondly, the seed dissemination potential and effective seedling recruitment distance from seed trees of the species were quantified, the suitability of substrates for the germination of P. uviferum seedlings was assessed, and the spatial distribution of seed trees of the species at the landscape level was analyzed. Finally, restoration plantings were established in disturbed upland and bog forests, to evaluate growth and mortality as well as performance (foliar nutrients and photosynthesis) of P. uviferum seedlings in relation to different site conditions characterized by variations in microtopography and canopy cover. According to the findings of this doctoral thesis, longevity (>880 years), extremely slow growth (<1 mm diameter per year) and tolerance to shade and stress are the main mechanisms of P. uviferum to persist in an extreme environment shaped 1

14 SUMMARY ZUSSAMMENFASSUNG - RESUMEN by low nutrient availability and frequently waterlogged conditions. Therefore, in contrast to propositions made in previous studies, old-growth P. uviferum forests are not a transitional phase in forest succession and may be maintained in the landscape for many centuries or millennia. However, although extremely stress-tolerant, the species is very sensitive to fire and recovery from this type of disturbance is very slow. Seventy years after high-severity fires in the study area, seed trees were extremely infrequent at the landscape level (0.3 trees/ha) and were aggregated at scales up to 30 m. Seed dissemination potential and effective seedling recruitment distance from individual P. uviferum seed trees was limited to 20 m, and seeds germinated best on moist Sphagnum or mineral soil substrate. This finding showed that high-severity fires can practically eliminate the species from parts of the landscape, where neither propagules nor seed trees survive. At the same time it points to the importance of biological legacies such as seed trees for the recovery of disturbed sites. Natural regeneration from seed trees can assist the recovery of P. uviferum populations following fire disturbance, but their effect is spatially limited at a landscape level. Therefore, restoration planting to complement existing seed trees may assist natural recovery of P. uviferum in disturbed bog forests and add to genetic diversity. To examine conditions for successful restoration planting, trials were established in bogs and upland forest to assess whether manipulation of potentially limiting conditions may assist establishment of seedlings. Seedlings were planted on mounds or in depressions on bog sites to examine the effect drainage or beneath a canopy or in the open on upland sites to examine the effect of light. In bogs, there was no significant effect of micro-topography on growth and survival of P. uviferum plantings. However, fluorescence measurements indicated lower stress in seedlings established on mounds. Seedlings in upland areas established beneath a nurse canopy had lower mortality and higher relative shoot growth, higher foliar nutrients, higher photosynthetic light use efficiency and higher fluorescence values than those planted in the open. This indicates that seedlings of the slow growing P. uviferum may tolerate extremely wet conditions, yet suffer from stress when grown in the open. These results may improve restoration plantings of the species in disturbed bog forests. The mixed passive-active restoration approach, relying on naturally regenerated parent trees complemented through low-density planting, which is suggested here, 2

15 SUMMARY ZUSSAMMENFASSUNG - RESUMEN could be the most effective and efficient option to restore disturbed southern bog forests dominated by P. uviferum in North Patagonia. This low-cost approach may help managers of forest land to restore the species in the future. Furthermore, the multiscale approach used in the present doctoral thesis studied the underlying ecological and physiological processes occurring in disturbed and undisturbed sites prior to planning a restoration program. In this context, it may be adopted for other ecosystems with low resilience and high degradation, where restoration is likely to be extremely expensive and the outcome uncertain. At the same time this study, which was inevitably limited in scope, showed that there are still very large knowledge gaps that should be filled to base management and restoration of these unique forests on a sound scientific basis. 3

16 SUMMARY ZUSSAMMENFASSUNG - RESUMEN ZUSAMMENFASSUNG Während die Mehrheit der Moore und Feuchtgebiete in der borealen Zone konzentriert sind, befinden sich große Flächen dieser Ökosysteme auch in subantarktischen Regionen der südlichen Hemisphäre. An der Südspitze Südamerikas gibt es ausgedehnte Nadelbaum-Moorwälder. Obwohl sich diese Wälder über viele Tausend Hektar erstrecken, ist über ihre Struktur und Dynamik nur wenig bekannt. Die Süd-Moorwälder Nordpatagoniens umfassen eine Fläche von fast 1 Mio. ha (6,8% der chilenischen Waldfläche). Kennzeichnend für diese Wälder ist die Baumart Pilgerodendron uviferum. Ein weiteres Charakteristikum dieser Waldökosysteme ist ihre geringe Resilienz gegenüber Störungen durch Feuer. Insbesondere von Menschen gelegte Brände sind die Hauptursache für ihre flächige Zerstörung. Das Ziel der vorliegenden Dissertation war es den Einfluss von Feuer, auf das Wachstum Nord- Patagonischer P. uviferum Moorwälder zu untersuchen. Auf dieser Grundlage sollten Strategien entwickelt werden, die zum Erhalt und zur Restaurierung dieser Wälder beitragen. Die vorliegende Arbeit ist auf drei wichtige Aspekte der Wiederherstellung ausgerichtet: a) Verständnis der Vorgänge in ungestörten Urwäldern, b) Analyse der natürlichen Rate der Erholung von Feuer-gestörten Wäldern, und c) Entwicklung von Optionen für die Restaurierung. Die vorliegende Arbeit basiert auf einem multiskalierten Ansatz. Zunächst wurden Durchmesser und Altersstruktur, Blätter- und Bodennährstoffgehalte sowie die Lichtverhältnisse in alten, durch Feuer gestörten P. uviferum Wäldern auf der Insel Chiloé (43 S, Nord-Patagonien) gemessen. Zudem wurde die Samen Verbreitung sowie die potenzielle und effektive Keimlingsetablierung in Relation zur Entfernung von Samenbäumen quantifiziert und die räumliche Verteilung von Samenbäumen auf Landschaftsebene analysiert. Die Eignung von Substraten für die Keimung von P. uviferum Sämlingen wurde experimentell untersucht. Abschließend wurden P. uviferum Sämlinge in unterschiedlichen Situationen in durch Feuer gestörten Hochland- und Moorwäldern gepflanzt. Auf Grundlage dieses Experiments wurde Wachstum und Mortalität sowie die Leistung (Blattnährstoffe und Photosynthese) von P. uviferum Sämlingen in Abhängigkeit von variierendem Überschirmungsgrad und unterschiedlichen mikrotopografischen Bedingungen bewertet. 4

17 SUMMARY ZUSSAMMENFASSUNG - RESUMEN Im Rahmen der vorliegenden Arbeit wurde festgestellt, dass P. uviferum eine sehr langlebige Baumart (Alter z.t. > 880 Jahre) mit einem extrem langsamen Wachstum (<1 mm Durchmesser pro Jahr) ist. Wachstumshemmend wirken Überschirmung und durch Staunässe und Nährstoffarmut geprägte Standort. Aus diesen Beobachtungen und im Gegensatz zu früheren Untersuchungen lässt sich folgern, dass P. uviferum Wälder keine Übergangsphase in der Waldentwicklung darstellen, sondern eine über viele Jahrhunderte bzw. Jahrtausende überdauernder Waldgesellschaft. Trotz ihrer hohen Toleranz gegenüber Überschattung und Nährstoffmangel reagiert die Baumart sehr empfindlich auf Feuerereignisse sodass die Folgen von Brandereignissen noch sehr lange messbar bleiben. Im Untersuchungsgebiet waren selbst 70 Jahre nach einem heftigen Waldbrand Samenbäume extrem selten (0,3 Individuen/ha) und höchstens in kleinen Gruppen mit einer flächigen Ausdehnung von bis zu 30 m vorhanden. Die Verbreitung von Samen sowie die potenzielle und effektive Keimlingsetablierung wurde lediglich in bis zu 20 m Entfernung vom Samenbaum beobachtet. Die meisten Keimlinge wurden auf feuchtem Sphagnum oder Mineralboden beobachtet. Dieses Ergebnis zeigt, dass heftige Feuer die Baumart aus Teilen der Landschaft völlig verschwinden lassen können. Gleichzeitig verdeutlichen sie die große Bedeutung von Samenbäumen für die Erholung gestörter Wälder. Nach Feuerereignissen wird die generative Verjüngung der P. uviferum Population durch Samenbäume gewährleistet, jedoch bleibt die Wirkung von Samenbäumen auf die Landschaftsebene beschränkt. Restaurative Pflanzungen können die Wirkung existierender Samenbäume ergänzen indem sie die natürliche Verbreitung von P. uviferum in gestörten Moorwäldern unterstützen und gleichzeitig die genetische Vielfalt erhöhen. Um optimale Bedingungen für erfolgreiche restaurative Pflanzungen zu identifizieren, wurden Pflanzversuche angelegt. Dabei wurden Keimlinge auf Hügeln oder in Vertiefungen auf Moorstandorten bzw. unter Schirm und im Freiland gepflanzt, um darzustellen welchen Einfluss die Wasserverfügbarkeit bzw. Licht auf das Keimlingswachstum ausüben. Auf Moorstandorten, konnte kein signifikanter Effekt der Mikrotopographie auf Wachstum und Überleben von P. uviferum Pflanzungen nachgewiesen werden. Jedoch deuten die Fluoreszenz Messungen daraufhin, dass 5

18 SUMMARY ZUSSAMMENFASSUNG - RESUMEN Sämlinge auf Hügeln weniger unter Stress geraten. In Hochlagen und unter Vorwald gepflanzte Sämlinge wiesen neben einer geringeren Sterblichkeitsrate auch ein relative höheres Sprosswachstum, mehr Blattnährstoffe, eine höhere photosynthetische Lichtnutzungseffizienz und höhere Fluoreszenz-Werte auf als die auf der Freifläche gepflanzten Sämlinge. Dies deutet darauf hin, dass langsam wachsende P. uviferum Sämlinge extrem nasse Bedingungen tolerieren können, doch auf der Freifläche unter erheblichem Stress leiden. Diese Ergebnisse verbessern die Ausgangssituation künftiger restaurativer P. uviferum Anpflanzungen in gestörten Moorwäldern. Die Mischung aus aktiven (Pflanzung) und passiven (Samenbäume) Maßnahmen stellt möglicherweise den effektivsten und effizientesten Weg zur Restauration gestörter Nord-Patagonischer P. uviferum Moorwälder dar. Dieser Low-Cost-Ansatz kann Praktikern dabei helfen P. uviferum Wälder wieder herzustellen. Darüber hinaus kann der in der vorliegenden Dissertation verwendete Multi-Skalen-Ansatz für andere stark degradierte Ökosysteme mit geringer Widerstandsfähigkeit übernommen werden, wo Wiederherstellungsmaßnahmen sehr kostenintensiv und ihr Erfolg unsicher sind. Gleichzeitig zeigt die vorliegende, in ihrem Umfang zwangsläufig begrenzte, Untersuchung auf, dass noch sehr große Wissenslücken bestehen, die dringend geschlossen werden müssen, damit die Wiederherstellung der einzigartigen Moorwälder über eine solide wissenschaftliche Grundlage verfügt. 6

19 SUMMARY ZUSSAMMENFASSUNG - RESUMEN RESUMEN Mientras la mayoría de los humedales y turberas a nivel mundial están concentrados en la zona boreal, millones de hectáreas de estos ecosistemas se pueden encontrar también en zonas subantárticas del hemisferio sur. Pese a la gran superficie de turberas existente en el cono sur de Sudamérica, existen escasos estudios sobre la estructura y dinámica de los bosques turbosos allí ubicados los cuales son dominados generalmente por coníferas. Los bosques turbosos australes, dominados por la conífera Pilgerodendron uviferum en Patagonia Norte, representan un típico caso de ecosistema con baja resiliencia a incendios, en los cuales se queman gran parte de los árboles y semillas. Estos bosques, que han sido quemados extensivamente, cubren una superficie de casi 1 millón de hectáreas (6,8% de los bosques nativos chilenos). La presente tesis doctoral tuvo como principal objetivo el estudiar bosques turbosos inalterados y quemados de P. uviferum en Patagonia Norte de tal forma de desarrollar la base ecológica-científica necesaria para futuras estrategias de conservación y restauración. De esta forma, el trabajo fue enfocado en tres importantes aspectos relacionados a la restauración: a) entender los procesos ecológicos que ocurren en bosques turbosos inalterados, b) analizar el grado de recuperación natural de los bosques quemados, y c) explorar algunas opciones para su restauración. Usando un enfoque multi-escala, este trabajo empieza analizando estructuras diamétricas y de edades, valores nutritivos foliares y del suelo, y el ambiente lumínico dentro de bosques inalterados y alterados dominados por P. uviferum en la Isla de Chiloé (43ºS) en Patagonia Norte. En bosques alterados se cuantificó el potencial de diseminación de semillas y la distancia efectiva de reclutamiento desde los árboles semilleros de P. uviferum, además se evaluó la idoneidad de diferentes sustratos para la germinación de semillas de la especie, y se analizó la distribución espacial de árboles semilleros a nivel de paisaje. Por último, en zonas alteradas, se establecieron plantaciones con la especie en sitios turbosos y de mejor drenaje (bosques de monte), para evaluar el crecimiento y mortalidad como también el rendimiento (nutrición foliar y fotosíntesis) de plantas de P. uviferum, en relación a diferentes condiciones de sitio (micro-topografía y cobertura de copas). 7

20 SUMMARY ZUSSAMMENFASSUNG - RESUMEN De acuerdo a los resultados de esta tesis doctoral, la longevidad (>880 años), el crecimiento extremadamente lento (<1mm en diámetro el año), y la tolerancia a la sombra y al estrés son los principales mecanismos de P. uviferum para persistir en ambientes extremos, los que se caracterizan por poseer baja disponibilidad de nutrientes y frecuentes inundaciones. De esta forma, en contraste a lo postulado en estudios previos, rodales adultos dominados por esta especie no son una fase transicional en la sucesión de estos bosques, pudiendo la especie mantenerse por varios siglos o milenios en el paisaje. Sin embargo, pese a ser una especie tolerante al estrés, es muy sensitiva al fuego y su recuperación luego de incendios es extremadamente lenta. Es asi como setenta años después de incendios catastróficos en el área de estudio, existe una muy baja frecuencia de árboles semilleros a escala de paisaje (0,3 árboles/ha), y éstos están agregados en escalas de hasta 30 m. El potencial de diseminación de semillas y la distancia efectiva de reclutamiento estuvieron limitados a 20 m, y las semillas germinaron mejor en sustratos húmedos compuestos por Sphagnum y suelo mineral. Estos resultados revelan que incendios severos pueden prácticamente eliminar la especie de extensas zonas del paisaje, donde no sobreviven ni propágalos ni árboles semilleros. Además estos resultados destacan la importancia de la persistencia de legados biológicos, como son los árboles semilleros, para la recuperación de sitios alterados. La regeneración natural post-incendio puede asistir la recuperación de poblaciones alteradas de P. uviferum, pero su efecto espacial a nivel de paisaje es extremadamente limitado. Por este motivo, plantaciones con la especie como complemento a los árboles semilleros podrían ayudar a asistir la recuperación natural de P. uviferum en bosques turbosos alterados y sumar diversidad genética. Para examinar las condiciones necesarias para una restauración activa exitosa con P. uviferum, se establecieron plantaciones de la especie en turberas y bosques de monte para evaluar si la manipulación de las condiciones potencialmente limitantes puede asistir el establecimiento de las plantas. En turberas, las plantas fueron establecidas en camellones y depresiones para examinar el efecto del drenaje del suelo. En bosques de monte (con mejor drenaje), las plantas se establecieron bajo cobertura de copas y en áreas descubiertas, para de esta forma examinar el efecto de la luminosidad en el establecimiento. En turberas no hubo un efecto significativo de la micro-topografía en el crecimiento y sobrevivencia de la especie. Sin embargo, 8

21 SUMMARY ZUSSAMMENFASSUNG - RESUMEN mediciones de fluorescencia indicaron menor estrés en plantas establecidas en camellones. En bosques de monte, plantas establecidas bajo cierta cobertura de copas presentaron menor mortalidad, mayor crecimiento relativo apical, mayor nutrición foliar, mayor eficiencia fotosintética en el uso de la luz y mayores valores de fluorescencia que las plantas establecidas en zonas descubiertas de vegetación. Esto demuestra que P. uviferum puede tolerar condiciones de humedad extrema, pero sufre por estrés cuando crece en zonas descubiertas. Estos resultados podrían ayudar de forma importante a mejorar futuras plantaciones de P. uviferum y actividades de restauración en bosques alterados. El enfoque mixto pasivo-activo de restauración sugerido en esta tesis, basado en regeneración natural proveniente de árboles semilleros y la plantación complementaria a baja densidad, sería la opción más eficiente y efectiva para restaurar los bosques turbosos alterados dominados por P. uviferum en Patagonia Norte. Este enfoque de bajo costo ayudará a profesionales relacionados con el manejo de bosques, en la planificación futura de actividades de restauración con la especie. Finalmente, el enfoque multi-escala usado en esta tesis doctoral, en el cual se estudian los procesos ecológicos y fisiológicos esenciales que ocurren en sitios alterados e inalterados de forma previa a la planificación de las actividades de restauración, puede ser adoptado para otros ecosistemas con baja resiliencia y alta degradación, donde la restauración es extremadamente costosa y sus resultados inciertos. Al mismo tiempo este estudio, el cual fue acotado en su ámbito, mostró que aún existen grandes vacíos de conocimiento que debieran ser llenados, para basar en fundamentos científicos el manejo y restauración de estos bosques únicos. 9

22 Chapter One: Introduction 10

23 Chapter One: Introduction Chapter One Introduction Southern bog forests of North Patagonia. Río Zorra Valley, Chiloé Island 11

24 Chapter One: Introduction The present doctoral thesis about the dynamics and restoration of Pilgerodendron uviferum forests on Chiloé Island in North Patagonia is the product of almost four years of research, between 2008 and The following chapter introduces this research and aims to explain the importance of P. uviferum bog forests in a global context. These forests may serve as an example of ecosystems with low resilience to fire disturbances that have been extensively burned and severely degraded, and for which there is an urgent need for restoration strategies. 1.1 Peatlands and bog forests Boreal and artic wetlands cover almost 3.5 million km² of terrestrial land surface (Shaw et al. 2003). Owing to the disequilibrium between rates of carbon fixation and decay processes in these ecosystems, they accumulate vast quantities of peat. It is estimated that one-third of the global carbon pool is stored in northern peatlands, which currently function as net sink for atmospheric carbon (Gorham 1991). While the majority of peatlands and wetlands is concentrated in the boreal zone (3.5 million km²), millions of ha of these ecosystems are found in subantarctic areas in parts of Australia, New Zealand, and the southern cone of South America (Chile and Argentina) (Shaw et al. 2003). Peatlands, which are wetland ecosystems characterized by accumulation of peat (NWWG 1997), develop through either of two main pathways: paludification and terrestrialization (Klinger 1996). Paludification refers to peat accumulation (principally Sphagnum sp.) that starts directly over a formerly dry mineral soil and which involves the formation of waterlogged conditions, whereas terrestrialization is the process by which a shallow water body is gradually filled in with organic and inorganic material (Lavoie et al. 2005). Paludification leads to the formation of marshes, fens, swamps and bogs (NWWG 1997). These four terms, and the ecosystems classes that they represent, are the most common used in the wetland literature, and it is important to clarify that there are no sharp boundaries between them (Rydin & Jeglum 2006). Marshes are characterized by standing or slowly moving water with submergent, floating or emergent plant cover. However, many marsh habitats are not peatlands since they have only little peat, so that most vascular species are rooted in the underlying mineral soil from which they 12

25 Chapter One: Introduction can take up nutrients (Rydin & Jeglum 2006). Swamps is a term that has been used in reference to forested peatlands. Swamps are influenced by minerotrophic groundwater, either on mineral or organic soils, and are dominated by trees or tall shrubs. They are characterised by the dominance (over 30% of cover) of tall woody vegetation and the wood-rich peat laid down by this vegetation (NWWG 1997; Lavoie et al. 2005). Fens are minerotrophic peatlands (rich in dissolved minerals) with a water table slightly below, at, or just above the surface (Rydin & Jeglum 2006). The vegetation cover is mostly composed by graminoid species and brown mosses, but also moderately decomposed Sphagnum (NWWG 1997; Lavoie et al. 2005). Trees and shrubs occur mostly in the driest fen sites, where microtopographic features such as moss hummocks provide elevated habitats above the water table (NWWG 1997). Bogs are ombrotrophic peatlands with the surface above the surrounding terrain, or otherwise isolated from laterally moving mineral-rich soil waters (Rydin & Jeglum 2006). Usually, the water table is at or slightly below the bog surface and precipitation, fog or snowmelt are the primary water sources. Since precipitation in remote regions where bogs occur contains few dissolved minerals and is mildly acidic, the surface bog waters are nutrient-poor and acidic. Bogs can be covered by trees or be treeless and are usually covered with Sphagnum spp. and ericaceous shrubs (NWWG 1997). The general key features of the four classes of peatlands based on the Canadian classification of the National Wetlands Working Group (NWWG 1997) and Rydin & Jeglum (2006) are summarized in Table 1.1. Peatlands dominated by Sphagnum sp. are often associated with forests. However, relatively little is known about the ecology, structure and dynamics of forests growing on these peatlands, and they are poorly defined. Until now they have been referred to as forested peatlands (Lavoie et al. 2005), swamp forests (Jennings et al. 2000), and bog forests (Holz & Veblen 2009). For the purpose of this thesis the term bog forest will be used to describe forest varying in cover from sparse to dense areas and affected by paludification (principally Sphagnum). 13

26 Chapter One: Introduction Table 1.1. Key features of marshes, swamps, fen and bogs (based on NWWG 1997 and Rydin & Jeglum 2006). Peatland attribute Marsh Swamp Fen Bog General Mineral wetlands Peatland and mineral wetland Accumulation of peat Accumulation of peat Shape Flat, concave, or sloping Flat, concave, or sloping Flat, concave, or sloping Flat, convex, or sloping Nutrient regime Minerophic, euto-mesotrophic Minerophic, euto-oligootrophic Minerophic, euto-oligootrophic Ombrotrophic, oligotrophic Origin of water Limno-, topo-, soligenous Limno-, topo-, soligenous Topo-, soligenous or limnogenous Ombrogenous Soil / peats Little accumulation of organic material and peat of aquatic plants Highly decomposed woody peat and organic material Decomposed sedge or brown moss peat Moderately decomposed Sphagnum peat with remains of shrubs Ground surface microtopography Level or tussocky Irregular, with high hummocks, logs and tree bases provide support for hummocks Level, or with low hummocks, or patterned with low or high ridges alternating with depressions (flarks) Level, or with low hummocks and hollows, or patterned with hummocks or ridges alternating with hollows Moisture regime Shalow surface water which fluctuates dramatically Water table at or below the surface Fluctuating water table which may be at, or a few cm above or below, the surface Water table at or slightly below the surface and raised above the surrounding terrain Vegetation Submergents, floating-leaved, reeds, tall sedges Forests, tall shrub thickets, herbs, graminoids, bryphytes Open or sparse cover of low trees, low shrubs, herbs, graminoids, bryophytes Open or with low trees, dwarf shrubs, low cyperaceous plants, bryophytes Paludification increases soil moisture, reduces soil temperature and nutrient availability, and alters the vertical distribution of roots to an extent that the productivity of bog forests may decline by 50-80% after 100 years of succession 14

27 Chapter One: Introduction (Simard et al. 2007). However, some tree species, in particular some conifers can grow, albeit very slowly, and persist on these unfavourable sites (Bond 1989; Coomes et al. 2005). Most information for these types of ecosystems relates to bog forests dominated by conifers of the genera Picea and Pinus in the boreal zones of Canada and Scandinavia (e. g. Hörnberg et al. 1995; Gunnarsson & Rydin 1998; Roy et al. 1999; Kuuluvainen et al. 2002; Hanssen 2003, Fenton et al. 2005; Simard et al. 2007; Moroni et al. 2010). Despite their large extent of more than 12,000 km 2, there have been very few studies that have investigated structure and dynamics of the bog forests in peatlands in the southern cone of South America (Rydin & Jeglum 2006),. 1.2 Pilgerodendron uviferum bog forests In the south of Chile (North Patagonia), there are over islands of different sizes which are covered by millions of hectares of temperate rain forests. One of the most important species of these Patagonian forests is Pilgerodendron uviferum (Ciprés de las Guaitecas), one of the 9 conifers of the southern cone of South America. In contrast to other conifers in that region which grow on terrestrial sites (e. g. Araucaria araucana, Austrocedrus chilensis, Fitzroya cupressoides), P. uviferum forests are found in landscapes and on substrates shaped by glaciation, on wet sites with high annual rainfall and acidic and poorly-drained soils ( mm). Often the species is associated with Sphagnum bogs (Lara et al. 2006) (Figure 1.1). These southernmost conifer forests of the world (Martinez 1981) cover ha from 40ºS to 55ºS, representing 6.8% of native forests area in Chile (CONAF 2011). However, if all vegetation communities with the species would be taken into account, the covered area was significantly higher, since many forested areas are not considered as forests (< 10 trees/ha and <2 m height). Pilgerodendron uviferum (D.Don) Florin (Ciprés de las Guaitecas) is an endemic and dioecious conifer of Patagonia (Figure 1.2). It is considered a long-lived species of slow growth that may live for more than 850 years (Aravena 2007). Its well recognisable tree rings are good indicators of the past climate (Roig & Boninsegna 1991) and disturbances (Holz & Veblen 2009). Also P. uviferum wood and bark contain secondary metabolites which inhibit bacteria and fungi, rendering the wood 15

28 Chapter One: Introduction remarkably decay resistant (Solís et al. 2004) and valuable for construction (Bannister et al. 2008). The historical utilization of these forests for timber extraction (fence posts, railroad sleepers, boat and building construction) has caused the destruction of extensive areas of cipresales through burning to facilitate access to the trees (Lara et al. 2006; Bannister et al. 2008). The importance of the timber is indicated by the industry that was established in Chiloé island (North Patagonia) in the XIX century on the basis of the P. uviferum resource. From there more than railroad sleepers per year were exploited and sold for the railway system of northern Chile and Perú (Otero 2006). The historical utilization of this species accompanied by illegal harvesting in natural reserves are continuous problems for the conservation of this species in Patagonia (Szeicz et al. 2000; Bannister 2004). Against this background, the species was classified by the IUCN as vulnerable (Walter & Gillet 1998) and was included in Appendix I of CITES (Hechenleitner et al. 2005; Soto et al. 2010). Pilgerodendron uviferum bog forests are some of the least-studied coastal temperate rainforests in southern South America (Holz & Veblen 2009) and there is practically no information about their dynamics, in particular not in the old-growth stage (Bannister et al. 2008). Previous studies about the structure and dynamics of P. uviferum in old-growth swamp forests, suggested that the succession of these forests is related to a gradient in soil drainage improvement (Cruz & Lara 1981). It has been hypothesized that establishment pulses of the species occur in open sites with waterlogged soils and when seed is available. Further it has been proposed that in the absence of disturbances, as the soil drainage improves, establishment of P. uviferum will gradually decline and it will be replaced by more shade-tolerant angiosperm species (Cruz & Lara 1981; Bannister 2004; Bannister et al. 2008). In accordance with this commonly held view, old-growth stands of this species would represent only a transitional phase in forest succession. 16

29 Chapter One: Introduction Figure 1.1 Pilgerodendron uviferum bog forests in the Río Zorra Valley on Chiloé Island, North Patagonia. These forests cover flat areas (a) but also small hills formed by glacial till (b), generally associated with Sphagnum bogs. Figure 1.2 Pilgerodendron uviferum tree (a) and branches of a female tree with immature and mature, seed-bearing cones (b). 1.3 Resilience of Pilgerodendron uviferum forests to fire Disturbances are an integral part of ecosystem dynamics. Humans alter disturbance regimes in fundamental ways increasing scale and severity and reducing the resilience of ecosystems (Walker 2011). The resilience of an ecosystem has been defined in terms of the magnitude of disturbance that a system can absorb before changing state ( ecological resilience ), and the time required following a perturbation 17

30 Chapter One: Introduction for a system to return to a steady-state ( engineering resilience ). Hence a possible measure of resilience is how far the system has moved from that equilibrium (in time) and how quickly it returns (Holling 1973; Gunderson 2000). On the other hand, forest systems can be resistant, that is the capacity of the ecosystem to absorb disturbances and remain largely unchanged (Holling 1973). Finally, resistance is related to the concept of stability, which is the capacity of an ecosystem to persist or remain within a range of variation around a specified ecosystem state, or the capacity to maintain a dynamic equilibrium in time while resisting change (Holling 1973; Thompson et al. 2009). According to the panarchy concept of Holling (2001), during the slow sequence from the exploitation to the conservation phase in an adaptive cycle, ecosystems accumulate biomass to a point where the resilience is minimal and the ecosystems becomes an accident waiting to happen. In this context, forests that have rarely experienced stand level disturbance may be especially sensitive to intensive disturbances, in particular if these represent types of disturbances the systems have not evolved with (Burns 1993; Parish & Antos 2006). This appears to apply to P. uviferum bog forests, which can accumulate high biomass with dense, complex and highly combustible understoreys of Tepualia stipularis and other Myrtaceae during long periods without stand replacing disturbances. In fact, a pollen sequence on the Taitao peninsula (46º S) showed a period of approx. 10,000 years after the last glaciation with a relative continuous presence of P. uviferum (Lumley & Switsur 1993). In the high rainfall environment of coastal Patagonia, fires are most likely restricted to human activity, even though individual trees may be struck by lightning (Holz & Veblen 2009). Since P. uviferum trees are commonly of low stature and deep rooted, they are not easily uprooted by wind storms. Locally, changes in water table due to tectonic movements can also be important (Szeicz et al. 2003). However, following the arrival of Euro-Chilean settlers in the 18 th and 19 th centuries, there has been a considerable increase in fire frequency in North-Patagonia (Holz 2010). Since there has been no coevolution of the species with fire disturbances, it lacks obvious traits to withstand fires (e. g. thick bark) like other conifers of Chile (e. g. Araucaria araucana, Burns 1993), and therefore the historic fires have caused dramatic changes in the structure and composition of these forests (Cruz & Lara 1981; Bannister et al. 2008). 18

31 Chapter One: Introduction Natural disturbances rarely eliminate all structural elements from the preceding stand. Many structures and organisms survive, including sexually mature trees or tree regeneration, which are called biological legacies (Franklin 1990; Franklin et al. 2002). Quantity and types of biological legacies differ greatly among disturbances leading to widely varying starting points for stand structural development (Franklin et al. 2002). The importance of these legacies in the process of ecosystem recovery has been widely studied in the Northern Hemisphere (e. g. Franklin 1990; Franklin & MacMahon 2000; Hanssen 2003; Keeton & Franklin 2005; Manning et al. 2006; Herrera & García 2009; Kashian et al. 2012), but little work has been done in South-America. Some studies in Northern Patagonia have indicated the importance of P. uviferum seed trees, as biological legacies, for the recovery of the species in disturbed bog forests ( Cruz & Lara 1981; Bannister et al. 2008), but the effect of these seed trees for the recovery of P. uviferum populations at the landscape level has not been quantified. 1.4 Restoration approaches Ecological restoration is the process of assisting the recovery of an ecosystem that has been degraded, damaged, or destroyed (SER 2004). This definition shows that restoration is not something theoretical but has to do with active management and intervention. The fundamental science upon which the actions of ecological restoration should be based is called restoration ecology (Van Andel & Aronson 2006). The current paradigm in restoration ecology focuses on returning a degraded system to some desired state and directing system development along a desired trajectory (Hobbs & Norton 1996). However, sometimes it is not possible to achieve the desired state and other alternative states have to be defined and searched. These new states should have complexity function levels as high as possible, in order to provide the same or similar ecosystem services as the original state (Figure 1.3). The value of a model comprising alternative states is that it recognizes that systems can shift abruptly between two or more states, that the dynamics of the degraded state are different from those in the pristine or target state, and that therefore the trajectory of recovery will probably be different from that of degradation (Suding et al. 2004). 19

32 Chapter One: Introduction Figure 1.3 Traditional view of restoration options for a degraded system (adapted from Hobbs & Norton 1996). Figure 1.4 A conceptual model of system transitions between states of varying levels of function with two types of restoration thresholds, one controlled by biotic interactions and the other by abiotic limitations (adapted from Whisenant 1999). Restoration of degraded systems depends on the removal of the influences leading to degradation but often this will be not sufficient to promote restoration 20

33 Chapter One: Introduction (Hobbs & Norton 1996). Whisenant (1999) has suggested that there are two types of restoration thresholds that prevent the system from returning to a less-degraded state: one that is caused by biotic interactions, and the other that is caused by abiotic limitations (Figure 1.4). Embodied in the idea of restoration thresholds are the abiotic, biotic and socioeconomic filters suggested by Hobbs & Norton (2004). These filters vary in their importance along relatively easily defined gradients. The type of restoration response needed depends on which thresholds have been crossed and the degree of effort needed to restore the system to a particular state will vary along these gradients (Whisenant 1999; Hobbs & Harris 2001; Hobbs & Norton 2004). An important starting point to define a desired ecosystem state, set goals and develop a forest restoration strategy is to have a reference (see Pickett & Parker 1994; Aronson et al. 1995; Landres et al. 1999). This reference should be characterized by the structure, dynamics and species composition of the forest in conditions prior to man-made degradation. However, natural forests may be so variable over time that they do not provide any static targets for restoration. To overcome this problem, a characterization of the natural range of variability of the forests and the principal processes that influence them can be a useful option for managers that need to set restoration goals (Landres et al. 1999; Kuuluvainen et al. 2002). After setting the goals (desired state) for a restoration strategy, it is also important to know the processes that are occurring in the current or degraded state of the forest. With a clear view of what is happening in the ecosystem to be restored, it is possible to develop the most promising restoration activities. Until now, forest restoration approaches or activities have been divided in passive restoration, the removal of environmental stressors and the use of natural colonization of tree species and secondary succession to restore a site and active restoration, the use of artificial regeneration, burning and/or thinning to achieve a desired structure (Rey Benayas et al. 2008; Morrison & Lindell 2011). To identify the best restoration approach for a specific ecosystem, the natural rate of recovery has to be studied. According to Holl & Aide (2011), the natural rate of recovery of an ecosystem is affected by the intrinsic ecosystem resilience to a disturbance, the land-use history and the characteristics of the landscape. In this context, Holl & Aide (2011) suggested that three questions have to be answered by managers to identify the goals of a restoration project: (1) If we 21

34 Chapter One: Introduction take a passive restoration approach, what results do we expect? (2) If intervention is necessary, how and when should we intervene to achieve the project goals? and (3) How can restoration resources be used most efficiently at landscape or regional scale? In the last two decades, there was a noticeable increase in the number of studies and projects concerning ecological restoration in Chile. However, most of them were concentrated in the Mediterranean region of central and central-south Chile (e.g. Zamorano et al. 2008; Armesto et al. 2009; Schiappacasse et al. 2012), probably because these forests contain the highest biodiversity and they are the most endangered forests of the country (Bannister et al. 2012). In contrast, there are only few studies that have addressed restoration of North Patagonian forests, and they are focused on Fitzroya cupressoides or Valdivian forests (Lara et al. 2008; Armesto et al. 2009). Despite the extensive area of burned P. uviferum bog forests in North Patagonia, the extremely low natural rate of recovery of these ecosystems, and the precarious conservation problems of P. uviferum, there has been only one trial for active restoration of these forests (Armesto et al. 2009; Carmona et al. 2010). In this context, one important research area for the future is in the restoration ecology of southern bog forests. The lack of natural regeneration of this species after fires, the disappearance of intact stands to a state where only few remnants are left in the landscape, and a high demand for timber from the local inhabitants point to the challenges ahead (Cruz & Lara 1981; Bannister et al. 2008). Furthermore, there is an urgent need to gather relevant information to develop ecologically-based restoration strategies for P. uviferum bog forests. 1.5 Objectives and hypotheses The main purpose of this doctoral thesis was to study fire-disturbed and undisturbed old-growth P. uviferum bog forests in North Patagonia in order to develop the scientific basis for future conservation and restoration strategies for these disturbed forests. This work was focussed on three important aspects of restoration: first, understanding the processes occurring in undisturbed old-growth forests, analysing the natural rate of recovery of the fire-disturbed forests, and exploring some 22

35 Chapter One: Introduction options for restoration. These aspects are included independently in the following three chapters (2, 3 and 4). Accordingly, in chapter two it was hypothesized that the conifer P. uviferum can persist in bog forests in the absence of intensive, large-scale disturbances. This is the pre-condition for the development of old-growth forests in which this species maintains its dominance through continuing regeneration. A second hypothesis was that large scale fire disturbances threaten the local persistence of the species. To answer these two hypotheses, forest structure, the light environment and foliar nutrients of P. uviferum and its companion angiosperm species in old-growth and firedisturbed P. uviferum stands were analyzed. In chapter three, it was hypothesized that the natural recovery of the species populations in burned areas is limited by seed availability. Here, the extent to which the recovery of P. uviferum populations may be facilitated through natural regeneration from seeds of remnant trees was assessed. Additionally, the availability of suitable substrates for germination (microhabitats) or the competition with bryophytes at establishment may be limiting the recovery of P. uviferum populations in these areas. To answer these hypotheses, the seed dissemination potential and effective seedling recruitment distance from seed trees of the species was quantified, the suitability of substrates for the germination of P. uviferum seedlings assessed, and the spatial distribution of seed trees at the landscape level analyzed. Based on this information, the potential for a passive restoration approach to recover disturbed P. uviferum bog forests in North Patagonia was discussed. In chapter four, the aim was to assess how different site conditions that may be manipulated to facilitate seedling establishment affect the performance of P. uviferum restoration plantings. Therefore, it was hypothesized that growth and survival of planted P. uviferum seedlings will be better in conditions of relatively better drainage that occur on elevated micro-topography. Additionally, it was hypothesized that in upland sites, where impeded drainage is not limiting tree growth, growth and survival of planted seedlings will be higher under canopy protection. To answer these hypotheses, restoration plantings in disturbed upland and bog forests were established, growth and mortality of P. uviferum seedlings at sites of different microtopography and light conditions was evaluated, and the performance of seedlings 23

36 Chapter One: Introduction (foliar nutrients, photosynthesis) were analyzed. Based on this information, the potential of some active restoration options to recover disturbed P. uviferum bog forests in North Patagonia is discussed. 1.6 Study area The Tantauco Park (~ 43º10`S and 74º05`W) on Chiloé Island, North Patagonia, was selected as the study region because it contains the majority of ecological site types in which P. uviferum grows and has both old-growth and fire-disturbed forests on comparable sites (Figure 1.5). Chiloé Island ( ha) is the second biggest island in South America after Tierra del Fuego. A total of 21.3% of the surface is covered by P. uviferum dominated forests (CONAF et al. 1999). Since the XVI century, this island was the centre of the Spanish colonization of the Patagonian islands because of its strategic location in the middle of the maritime routes from Magallanes district to Valparaiso and Callao ports in the north. The early colonization of these islands caused an evident decline in the area of the islands forests with the consequent change of the land use to agriculture and urban areas. Southern Chiloé, where Tantauco Park is located, has a cool-temperate climate with strong oceanic influence with a mean annual temperature of 10 C, and high annual precipitation that can reach up to 6000 mm in some places (di Castri & Hajek 1976; Pérez et al. 2008). The landscape in the study area has been shaped by the last glaciation (glacial-postglacial transition ca. 13,000 yr. BP), which created a mix of hills from glacial till and shallow valleys, over Pre-Cambrian and Tertiary Metamorphic rocks with extremely acidic, poorly-drained soils with Gley horizons (Villagrán 1988). Altitudes range between 50 and 280 m above sea level. In depressions, sites are mostly covered by Sphagnum bogs and cushion plants, and where drainage is better, coniferous species like P. uviferum and Podocarpus nubigena or hardwood species like Tepualia stipularis, Nothofagus nitida, Nothofagus betuloides, Drimys winteri and Weinmannia trichosperma dominate the vegetation. 24

37 Chapter One: Introduction Figure. 1.5 Study area in the Tantauco Park on Chiloé Island, North Patagonia. Colours correspond to different ecosystems. Inset shows a detail of the study area with the names and locations of the sectors where this research was focussed. Generally, the disturbed forests of the study area can be divided in two topographical types: bog areas located in flat areas on raised peat bogs (Sphagnum and cushion plants); and upland areas on hills shaped by till with better drainage, but also some presence of Sphagnum or other bryophytes (Figure 1.2). For the purpose of this doctoral thesis, P. uviferum forests in this area were classified into undisturbed and disturbed bog and upland forests. While undisturbed bog forests corresponded to old-growth stands, disturbed bog forests were characterized by large quantities of standing dead trees of P. uviferum with some regrowth (post disturbance cohorts) of mostly other tree species. Undisturbed upland forests corresponded to old-growth stands and were dominated by P. uviferum and T. stipularis, where the latter species creates very complex structures owing to its horizontal growth habit. Finally, disturbed upland forests were characterised by dead trees of P. uviferum and T. stipularis with some regrowth of other tree species. Field sites were located in old-growth and disturbed forests of the Río Zorra valley and in disturbed areas around Lake 25

38 Chapter One: Introduction Chaiguata and Chaiguaco (Figure 1.5). Southern Chiloé, including the area of Chaiguata lake, was burned in an extensive fire between 1942 and 1944 (Bannister et al. 2008; Holz 2010). In this remote area, disturbed sites that have not been salvage-logged can be found. This allowed investigating the influence of fire-disturbance separately from additional effects that subsequent logging might have had. Study sites varied according to the different hypothesis to be tested. Therefore more information about the study area and the specific study sites is provided in each of the following chapters. 26

39 Chapter Two: Persistence of Pilgerodendron uviferum Chapter Two Persistence of the slow growing conifer Pilgerodendron uviferum in old growth and fire-disturbed southern bog forests Jan R. Bannister, Pablo J. Donoso, Jürgen Bauhus Published in the journal Ecosystems * Interior of an old-growth Pilgerodendron uviferum upland forest. Río Zorra Valley, Chiloé Island. *Bannister JR, PJ Donoso, J Bauhus Persistence of the slow growing conifer Pilgerodendron uviferum in old-growth and fire-disturbed southern bog forests. Ecosystems 15: Available in: 27

40 Chapter Two: Persistence of Pilgerodendron uviferum 2.1 Abstract In spite of the extensive area of bogs in the southern cone of South America, there have been very few studies on structure and dynamics of conifer bog forests in this region. Previously, it has been assumed that in the absence of intensive disturbance, the dominant conifer Pilgerodendron uviferum (D.Don) Florin would be replaced through other angiosperm species. Here we hypothesized a) that this conifer can persist without intensive disturbances and develop into old-growth forests with continuing regeneration and b) that high-severity disturbances through fire threaten its local persistence. To test this hypotheses, we analyzed diameter and age structure, foliar and soil nutrient levels and the light environment of old-growth and firedisturbed P. uviferum stands on Chiloé Island (43ºS) in North Patagonia. Longevity (>880 years), extremely slow growth (<1 mm diameter per year) and tolerance to shade and stress are the main mechanisms of P. uviferum to persist in nutrient-poor and waterlogged conditions. Hence, old-growth P. uviferum forests are not a transitional phase in forest succession and may be maintained in the landscape for many centuries or millennia. However, in fire-disturbed stands, live trees of the species were rare and regeneration negligible, showing that high-severity fires can eliminate the species from parts of the landscape, where neither propagules nor seed trees survive. This underpins the importance of biological legacies such as seed trees for the recovery of disturbed sites, and points to the need for active restoration approaches to restore fire-degraded P. uviferum forests. Keywords: bog forests, Chiloé Island, forest dynamics, light availability, N/P ratio, persistence mechanisms, Sphagnum. Article available in: 28

41 Chapter Three: Potential for passive restoration of P. uviferum forests Chapter Three The importance of seed-trees for passive restoration of Pilgerodendron uviferum in fire-disturbed southern bog forests Jan R. Bannister, Sven Wagner, Pablo J. Donoso, Jürgen Bauhus Pilgerodendron uviferum seed trees assisting the recovery of fire-disturbed bog forests Chaiguata Lake, Chiloé Island 29

42 Chapter Three: Potential for passive restoration of P. uviferum forests 3.1 Abstract Southern bog forests dominated by the conifer Pilgerodendron uviferum in Northern Patagonia are a typical case of an ecosystem with low resilience to disturbance by fire, which kills most trees and seeds. These forests, which are mostly burned, cover an area of almost 1 million ha (6.8% of Chilean native forests). So far there have been no studies that quantified the importance of surviving trees for the recovery of populations of this endangered species in fire-disturbed areas. In this study, we hypothesize that the natural recovery of P. uviferum populations in burned areas is limited by seed availability. Alternatively, availability of suitable substrates for germination or the competition with bryophytes at establishment may be limiting the recovery of P. uviferum populations in these areas. Using a multi-scaled approach, we quantified the seed dissemination potential and effective seedling recruitment distance from P. uviferum seed trees, assessed the suitability of substrates for the germination of P. uviferum seedlings, and finally analyzed the spatial distribution of seed trees of the species at the landscape level. Here, we show that 70 years after a fire on Chiloé Island (43 S), seed trees were extremely infrequent at a landscape level (0.3 trees/ha) and were aggregated at scales up to 30 m. Seed dissemination potential and effective seedling recruitment distance from individual P. uviferum seed trees was limited to 20 m. Seeds germinated best on moist Sphagnum or mineral soil substrate. Our results indicate that natural regeneration from seed trees can assist the recovery of P. uviferum populations following fire disturbance, but their effect is limited at a landscape level. Supplementary planting of seed trees may assist natural recovery of P. uviferum in disturbed bog forests and add to genetic diversity. This mixed passive-active restoration approach could be the most effective and efficient option to restore P. uviferum forests in North Patagonia. Keywords: biological legacies, bog forests, forest restoration, North Patagonia, seed dispersal, spatial population patterns. 30

43 Chapter Three: Potential for passive restoration of P. uviferum forests 3.2 Introduction Disturbances are an integral part of ecosystems. Humans alter disturbance regimes in fundamental ways increasing scale and severity and reducing the resilience of ecosystems (Walker 2011). The resilience of an ecosystem has been defined in terms of the magnitude of disturbance that a system can absorb before changing state ( ecological resilience ), and the time required for a system to return to a steady-state following a perturbation ( engineering resilience ). Hence a possible measure of resilience is how far the system has moved from that equilibrium (in time) and how quickly it returns (Holling 1973; Gunderson 2000). Forests that have rarely experienced stand level disturbance may be especially sensitive to intensive disturbances, in particular if these represent types of disturbances the systems have not evolved with (Burns 1993; Parish & Antos 2006). This appears to apply to bog forests dominated by the endemic, slow-growing and long-lived conifer Pilgerodendron uviferum (D. Don) Florin. (Ciprés de las Guaitecas) in North Patagonia. These southernmost conifer forests of the world (Martinez 1981) cover almost 1 million ha from 40 S to 55 S (6.8% of Chilean native forests), and they are often found in high rainfall environments ( mm per year) and on acidic and poorly drained soils covered by Sphagnum bogs (Lara et al. 2006). In this environment, these forests develop for long periods without significant disturbances, accumulating high amounts of biomass with dense and complex understoreys (Bannister et al. 2012). In the last 200 years, humans have caused the destruction of extensive areas of these forests through broad-scale burning to facilitate access to the trees (Bannister et al. 2008). Aboriginals and Euro-Chilean settlers have amplified fire activity to such an extent (Holz & Veblen 2011), that widespread anthropogenic fires in the last two centuries have caused dramatic changes in the structure and composition of these forests (Cruz & Lara 1981; Bannister et al. 2008). Pilgerodendron uviferum does not show any specific adaptations to fire disturbance such as thick bark, resprouting ability, protected seeds or a soil seed bank. In many areas, almost 70 years after the last fire, forest regrowth has been extremely slow and P. uviferum is still absent in large tracts of the landscape (Bannister et al. 2008). Owing to the widespread disturbance of these forests through fire, the 31

44 Chapter Three: Potential for passive restoration of P. uviferum forests conservation status of the species considered by the IUCN is vulnerable (Walter & Gillet 1998), and it has been included in Appendix I of CITES (Lara et al. 2006). Natural disturbances rarely eliminate all structural elements from the preceding stand. Many organisms survive, including sexually mature trees or tree regeneration. These are called biological legacies (Franklin 1990; Franklin et al. 2002) and their importance for ecosystem recovery has been documented for a range of ecosystems, especially in the northern Hemisphere (e.g. Franklin 1990; Franklin & MacMahon 2000; Hanssen 2003; Keeton & Franklin 2005; Manning et al. 2006; Herrera & García 2009; Kashian et al. 2012). Some studies in North Patagonia have indicated the importance of P. uviferum seed trees, as biological legacies, for the recovery of the species in disturbed bog forests (Cruz & Lara 1981; Bannister et al. 2008), but the effect of these seed trees for the recovery of P. uviferum populations at the landscape level has not been quantified. However, the recovery of populations may not only be limited by the availability of seeds but also by the availability of suitable sites for regeneration. Safe sites for natural regeneration are determined by seedbed type, moisture retention and microhabitat (Hörnberg et al. 1997; Cornett et al. 2000). For example for Scandinavian Picea abies bog forests, the main factor affecting seed emergence at a local scale was the microhabitat, and the most important factor for seedling mortality was overgrowth by bryophytes (Hörnberg et al. 1997; Hanssen 2003). This is also valid for Picea mariana and Pinus sylvestris bog forests in Canada (Roy et al. 1999) and Scandinavia (Gunnarsson & Rydin 1998), respectively. In this landscape context, long-term seedling survival is lower in depressions owing to vigorous Sphagnum growth or mortality caused by inundation (Hanssen 2003). In the case of P. uviferum forests, it is not known whether availability of suitable microhabitats may limit regeneration of the species in burned areas. Restoration of degraded systems depends on the removal of the influences leading to degradation but often this will be not sufficient to promote restoration (Hobbs & Norton 1996). There are two types of restoration thresholds that prevent the system from returning to a less-degraded state: one that is caused by biotic interactions, and the other that is caused by abiotic limitations (Whisenant 1999). The type of restoration response needed depends on which thresholds have been crossed and the effort needed to restore the system to a particular state (Whisenant 1999; 32

45 Chapter Three: Potential for passive restoration of P. uviferum forests Hobbs & Harris 2001). Generally, forest restoration approaches have been divided in passive and active, where the former comprises removal of environmental stressors and use of successional processes to restore a site, and the latter refers to artificial regeneration through planting or sowing, burning and/or thinning to achieve or promote a desired structure (Rey Benayas et al. 2008; Morrison & Lindell 2011). To identify a suitable restoration approach for a specific disturbed ecosystem, information about the natural rate of recovery is required (Holl & Aide 2011). In this study, we aimed to assess the extent to which the recovery of P. uviferum populations may be facilitated through natural regeneration from seeds of remnant trees. Our principal hypothesis was that the natural recovery of P. uviferum populations in burned areas is limited by seed availability. Also, the availability of suitable substrates for germination (microhabitats) or the competition with bryophytes at establishment may be limiting the recovery of P. uviferum populations in these areas. For this we (a) quantified the seed dissemination potential and effective seedling recruitment distance from P. uviferum seed trees, (b) assessed the suitability of substrates for the germination of P. uviferum seedlings, and (c) analyzed the spatial distribution of seed trees of the species at the landscape level. Based on this information, we discussed the potential for a passive restoration approach to recover disturbed P. uviferum bog forests in North Patagonia. 3.3 Methods Study area The study area was located in Tantauco Park (~ 43º10`S and 74º05`W) on Chiloé Island, North Patagonia (Figure 3.1). This conservation area contains the majority of ecological types in which P. uviferum grows, as well as extensive disturbed areas that were burned mainly in an extensive fire in 1943 (Bannister et al. 2008; Holz & Veblen 2011). Southern Chiloé has a cool-temperate climate with strong oceanic influence, a mean annual temperature of 10º C, and high annual precipitation that can reach up to 6,000 mm in some places (di Castri & Hajek 1976; Pérez et al. 2008). The landscape in 33

46 Chapter Three: Potential for passive restoration of P. uviferum forests the study area has been shaped by the last glaciation (ca. 13,000 yr. BP), which created a mix of hills from glacial till and shallow valleys, over Pre-Cambrian and Tertiary Metamorphic rocks with extremely acidic, poorly drained soils with Gley horizons (Villagrán 1988). Altitudes range between 150 and 280 m a.s.l. The disturbed forests of the study area had been burned but subsequently not salvage-logged. This allowed us to investigate the influence of fire-disturbance separately from additional effects that subsequent logging might have had. These areas were mostly covered by Sphagnum bogs and cushion plants and characterised by dead trees of P. uviferum and Tepualia stipularis with some regrowth of broadleaf species like T. stipularis, Nothofagus nitida, Drimys winteri and Weinmannia trichosperma. The study area can be divided in two topographical types: bog areas located in flat areas on raised peat bogs (Sphagnum and cushion plants); and upland areas on hills shaped by till with better drainage (Figure 3.2a). Figure 3.1 Study area in the Tantauco Park on Chiloé Island, North Patagonia. Gray colour corresponds to Pilgerodendron uviferum dominated forests. 34

47 Chapter Three: Potential for passive restoration of P. uviferum forests Seed and seedling dispersion around P. uviferum seed trees To quantify the seed dissemination potential of P. uviferum seed trees in disturbed forests, one representative 76 year old isolated seed tree of 15.5 cm dbh and 4.8 m in height was selected to establish a dissemination assay (Figure 3.2b). A total of 28 seed traps of 0.25 m² were established at 2 m distance from the seed tree and then every 5 m up to 30 m in each cardinal direction, at 30 cm height from soil. Seeds in traps were protected from predation through nets. Once every year (May) for 3 years, seed traps were emptied and the seeds counted. Figure 3.2 a) Sampled bogs in the foreground and upland areas in the background; b) Seed traps near seed tree.; c) Seed tree with some clustered regeneration; d) Germination of P. uviferum seedlings in a Sphagnum substrate. To calculate the effective seedling recruitment distance from P. uviferum seed trees, effective regeneration was quantified in the surroundings of 20 seed trees, which were isolated (>150 m of distance to other seed tree) and had already 35

48 Chapter Three: Potential for passive restoration of P. uviferum forests regenerated (Figure 3.2c). The diameter at breast height (dbh), or root collar diameter (RCD), sex, and position in a coordinate system (X-Y) were measured for each seed tree and seedling. To determine tree age, two increment cores were obtained at the lowest possible stem height from each seed tree and one increment core for up to three individuals of their regeneration > 3 cm of dbh. Tree cores were mounted and sanded following standard procedures (Stokes & Smiley 1968). To develop an individual tree-based model to predict the spatial distribution of seedlings, 1 m wide concentric rings were laid around each of the 20 selected trees. Regeneration frequency was counted within each ring and then transformed to density. Regeneration density around the 20 selected trees was then fitted with a lognormal model (Stoyan & Wagner 2001). This model was combined with an allometric component of the function used by Ribbens et al. (1994) so that the final model we used corresponded to [1]:, [1] where μ and σ are the logarithms of the median and the standard deviation of the distribution, N (standard total recruitment) is the number of recruits produced for a tree of a standarized dbh dbhref, β is a factor that modifies N and dbh is the diameter of the tree. We chose to scale N relative to a tree of 13.5 cm dbh, representing the mean of the 20 selected seed trees. Model parameters were calculated with the R software package (R Development Core Team 2010) and fitted using the least squares method for non-linear models Germination experiment To evaluate the effect of different substrates found in disturbed sites on the germination of P. uviferum seeds, a germination experiment with locally collected seeds (weight: 523,560 seeds/kg) was performed in the nursery of Tantauco Park (Figure 3.2d). This experiment comprised 4 different substrates: moss, mineral soil, 36

49 Chapter Three: Potential for passive restoration of P. uviferum forests cushion plants, and sand, which served as a control. To capture the interactions that may occur between substrate and moisture content, these substrates were kept either wet (watering each day) or were allowed to fall dry (watering once a week), yielding 8 treatments in total. In each of 4 replicates per treatment 100 seeds were sown at 3 cm distance between them. Following the suggestion of Donoso et al. (1980), the seeds were stratified for 60 days at 4 ºC in moist sand prior to sowing. Seed germination rates were monitored for 6.5 months between July 2010 and February 2011, and at the end the height of germinants were sampled in each treatment (ca. of 122 germinants per treatment). Analyses of normality and homogeneity of variance for all variables were done with Shapiro-Wilk W, Kolmogorov Smirnov and Levene tests. Data for germination rates were normally distributed. Thus, based on substrate and moisture content as factors, we performed a Two-way ANOVA for comparisons among treatments and significant differences (P<0.05) were further analyzed with post-hoc comparisons based on the Bonferroni method. Heights of germinants, which were, even after various transformations, not normally distributed, were compared with Kruskal-Wallis tests. Where significant differences were found (P<0.05), differences between means were tested with post-hoc comparisons based on Mann Whitney U tests Spatial distribution of seed trees at the landscape level Using a satellite image of the study area in the disturbed forests near Chaiguata Lake (Figure 3.1), we selected a representative area of 100 ha (1km x 1km) with both upland and bog sites to map the location of every living P. uviferum tree > 1m in height. In areas with clusters of living trees we mapped only trees 5cm dbh or 2 m height. For each tree, we recorded the dbh (cm), rcd (cm), height (m), sex, and site condition (bog or upland). For every female tree with regeneration, we recorded seedling abundance in 3 categories: low (< 9 seedlings; 1st quartile), intermediate (between 9 and 45 seedlings; 2nd and 3rd quartile), and high (>45 seedlings; 4th quartile). The division between quartiles was based on the analysis of regeneration around the 20 seed trees. The age of the trees was estimated from diameters based on 37

50 Chapter Three: Potential for passive restoration of P. uviferum forests a regression model (r²=0.843; P<0.001) developed for the study area by Bannister et al. (2012). To characterize the spatial patterns of living P. uviferum trees at the landscape level, we examined the univariate and bivariate spatial point patterns of P. uviferum trees within the mapped area, using the Wiegand-Moloney`s O-ring statistic (Wiegand et al. 1999). Compared to Ripley`s K(t) function (Ripley 1977), the use of the O-ring statistic has the advantage that it is not cumulative, and therefore does not confound effects at larger scales with effects at shorter distances (Wiegand & Moloney 2004). In this context, the mark-correlation function g 12 (r) is the analogue of Ripley`s K 12 (r) when replacing circles by rings of radius r. The O-ring statistic O 12 (r) = λ 2 g 12 (r) gives the expected number of points of pattern 2 at distance r from an arbitrary point of pattern 1, were λ is the first-order intensity of the pattern (Wiegand & Moloney 2004). In the case of bivariate spatial patterns, O 12 (r) = λ 2 indicates independent patterns, O 12 (r) < λ 2 indicates repulsion between seed trees, and O 12 (r) > λ 2 points to attraction between pattern 1 and 2. For univariate patterns, O(r) is calculated by setting pattern 2 equal to pattern 1. In this case O(r) = λ indicates randomness, O(r) < λ indicates regularity and O(r) > λ is a sign of aggregation (Wiegand & Moloney 2004). To assess the significance of the O-ring statistic under a given null model, we generated 95% confidence envelopes by calculating for each distance r the 5 th lowest and highest values of the summary statistic from 99 Monte Carlo simulations of the null model. In the mapped area, there were signs of heterogeneity in the first-order intensity λ and clusters in the patterns, which may be related to topographical conditions. Therefore we could not use complete spatial randomness or independence as null models for univariate and bivariate second-order analyses, respectively. Alternatively, null models based on a heterogeneous Poisson process using a moving window estimate of radius R = 100 m were used for approximation of the heterogeneous firstorder intensity λ. In the case of the bivariate analysis, we kept the locations of pattern 1 fixed and randomizing pattern 2. All spatial point pattern analyses were performed with the software Programita (Wiegand & Moloney 2004) based on a grid of 100 x 100 cells of 10 x 10 m size. The analyses were done with a ring width of 1 cell and a maximum distance r of 50 cells (500 m, half of the area side). 38

51 Chapter Three: Potential for passive restoration of P. uviferum forests 3.4 Results Dispersion of seeds and seedlings of P. uviferum around seed trees Over a period of 3 years, seed production of P. uviferum was irregular presenting mast years (2010) and years with hardly any seeds (2011). Therefore the analysis of the pattern of seed dissemination was focused on the year We found a negative exponential seed-rain curve with distance from seed tree, where most seeds were trapped within a 20 m radius (Figure 3.3a). However, throughout the whole observation period, the majority (93-100%) of seeds fell in the first 5 m from the seed tree. Figure 3.3 Dispersion seeds and seedlings around P. uviferum seed trees. a) Seed dissemination over a 3 years period; b) Log-normal density based model for the effective seedling recruitment distance from 20 trees of different dbh. B = μ; C = σ and Beta = factor of the allometric component; N = standard total recruitment; dbhref = mean dbh of the 20 selected seed trees. MDD = mean distance; MAX = distance at which the maximum seedling density occurs. The effective seedling recruitment distance from P. uviferum seed trees fitted well a log-normal distribution (Figure 3.3b). The maximum density of seedlings occurred at 1.6 m and the mean distance of seedlings to the seed tree was 4.7 m. Seedling recruitment was negligible beyond a radius of 20 m from seed trees and it was concentrated in easterly directions (70% of seedlings between 0º and 180º). There 39

52 Chapter Three: Potential for passive restoration of P. uviferum forests were no significant correlations between the dbh of seed trees and the maximum distance of the regeneration and seedling frequency. However, there was a positive correlation between seed tree age and maximum distance of the regeneration (r=0.553; r 2 =0.306; P=0.011) and seedling frequency (r=0.631; r 2 =0.398; P=0.003). There was also a positive correlation between seed tree height and seedling frequency (r=0.534; r 2 =0.285; P=0.015). From a total of 772 seedlings recorded in the vicinity of the 20 seed trees, 49% had reproductive structures and the observed male/female ratio of seedlings was 1.74 (Table 3.1). The smallest female and male seedlings with cones were about 40 to 50 cm tall with a root collar diameter of ca 0.5 cm. Figure 3.4 a) Mean cumulative seed germination rate for each substrate. Different letters indicate significant differences (P 0.05) between treatments. b) Final mean height of germinants in each treatment. No significant differences between treatments (P>0.05). Treatments: SphDry = dry Sphagnum; SphMoist = moist Sphagnum; MinDry = dry mineral; MinMoi = moist mineral; CusDry = dry cushion; CusMoi = moist cushion; SanDry = dry sand; SanMoi = moist sand Seed Germination The first germinants emerged 3 months after sowing and their numbers increased constantly until month 5, after which germination ceased. Substrate and moisture content, as independent factors had significant effects on seed germination (P 0.001). Germination rates were significantly higher in moist Sphagnum and moist mineral soil than in all the other substrates (Figure 3.4a). Mortality of germinants in 40

53 Chapter Three: Potential for passive restoration of P. uviferum forests these substrates was only 2% and 3%, respectively, whereas it was much higher in the substrates undergoing drying-rewetting cycles. Seeds germinated also under dry conditions but substantially less successful as indicated by extremely low germination rates in dry cushion plants and sand (Figure 3.4a). The interaction between substrate and moisture content did not have a significant effect on seed germination. Final heights of germinants were not significantly different between treatments with successful germination (Figure 3.4b) Spatial distribution of seed trees at the landscape level At a landscape level, most individuals of P. uviferum were found in smaller diameter classes and the estimated ages of P. uviferum trees < 81 years (Table 3.1). Of the mapped area, 72% was in upland sites and 28% was in bogs (Figure 3.5). However, 46.7% of the P. uviferum trees mapped were in upland areas and 53.3% in bogs (Table 3.1). Male and female trees contributed 48 and 45% of the total number, respectively. Only 28% of the trees (0.3 trees/ha) had any regeneration in their vicinity. From the female trees with regeneration, 50% and 30% had intermediate or high seedling abundance, respectively (Table 3.1). The univariate O-ring statistic revealed that the distribution of P. uviferum trees was not uniform but clustered for all trees together, as well as for males and female trees with regeneration at scales 30 m (Figure 3.6). Although there was some weak regularity or aggregation at some scales ( and near 180 m); beyond that scale distribution patterns were almost always random up to 500 m. Female trees without regeneration had a mostly random distribution at all scales (Figure 3.6d). The bivariate O-ring statistic revealed attraction between male and female trees with regeneration at scales 30 m, independency at scales between 40 and 50 m followed by repulsion at scales of 60 to 130 m. At larger scales, the distribution of these two groups was largely independent from each other (Figure 3.6e). In contrast, the bivariate spatial pattern between males and female trees without regeneration was mostly independent at all scales, with some repulsion at scales between 40 to 90 m, (Figure 3.6f). 41

54 Chapter Three: Potential for passive restoration of P. uviferum forests Table 3.1 Principal attributes of disturbed P. uviferum forests at the landscape level. P. uviferum trees in the mapped area (100 ha) N (trees/100 ha) % Total Site Bogs Upland Gender undefined males females M:F sex ratio 1.07 Structure Age a (years) 29 to 81 dbh (cm) 2 to 20.5 height (m) 1 to 9 Associated regeneration N (trees/100 ha) % Females with regeneration Abundance of seedlings low, < intermediate, >9 to < high, > Sex of young trees b N (seedl/20 seed trees) % Total number 772 Undefined 394 (51.0) Males 240 (31.1) Females 138 (17.9) M:F sex ratio 1.74 a Ages estimated from dbh-model from Bannister et al. (2012); b Regeneration data from the analysis of 20 seed trees with regeneration (see method) 42

55 Chapter Three: Potential for passive restoration of P. uviferum forests Figure 3.5 Diagram showing the mapped area of 100 ha (1 km x 1km) near Chaiguata Lake in Tantauco Park and the distribution of male and female P. uviferum trees. 43

56 Chapter Three: Potential for passive restoration of P. uviferum forests Figure 3.6 Univariate (a-d) and bivariate (e-f) second-order analyses of the P. uviferum trees located in the mapped area (100ha). The mapped area was divided in a grid of 100 x 100 cells, each of 10 m width. The O-ring (a-d) and O 12 -ring (e-f) statistics are shown as solid lines and the 95% confidence envelopes of 99 randomizations of the pattern over the study area are shown as dotted lines. The solid horizontal grey line indicates the mean intensity λ of the general pattern (a-d) and the mean intensity λ 2 of pattern 2 (e-f). 44

57 Chapter Three: Potential for passive restoration of P. uviferum forests 3.5 Discussion Capacity of Pilgerodendron uviferum to recover burned areas The widespread fire of the early 1940s that burned the study area caused dramatic changes in the structure of the P. uviferum bog forests. Almost 70 years after the large fire, recovery of P. uviferum populations on bogs has been extremely slow, indicating a low resilience of these bog forests in relation to fire disturbance. Seed trees were extremely infrequent and their contribution to the regeneration of the disturbed area was confined to their immediate surroundings (< 20 m). Moist substrates comprising Sphagnum or mineral soil were the best substrates for the germination of P. uviferum seedlings (near 60% germination rate, 2-3% mortality), and these are precisely the most frequent substrates in the study area. Thus, it appears unlikely that availability of germination substrate may be limiting seedling establishment. However, in northern bog forests dominated by Picea abies, Picea mariana or Pinus sylvestris, the risk of waterlogging is lower on hummocks and elevated microsites, which are therefore safe sites for development of seedlings (Hörnberg et al. 1997; Gunnarsson & Rydin 1998; Roy et al. 1999; Hanssen 2003). Safe sites for germination are not necessarily safe sites for seedling survival, and to be overgrown by bryophytes is one of the most important mortality factors for slowgrowing tree seedlings in bogs (Hörnberg et al. 1997). Hence only seedlings that have rapid height growth are able to stay ahead of the upward growth of Sphagnum (Ohlson 1995). Although we did not analyze the competition between P. uviferum seedlings and Sphagnum, there was close agreement between seed dissemination and effective seedling recruitment patterns around seed trees, indicating that although this competition exists, there must be enough safe sites for seedling survival. This also indicates that the post-disturbance recovery of P. uviferum populations is clearly limited by the low survival of seed trees or propagules during fire and subsequently the limited seed dispersal. The short-distances of seed dissemination and effective seedling recruitment patterns around P. uviferum seed trees (<20 m) indicate their limited capacity to provide regeneration at the landscape scale. However, in addition to the dominating 45

58 Chapter Three: Potential for passive restoration of P. uviferum forests short-distance dispersal, an infrequent long-distance dispersal pattern is indicated by the presence of isolated trees in the landscape (Figure 3.5). All isolated P. uviferum trees, except for one isolated male tree, were younger than the period since the last fire (1943). Hence, these isolated trees must have originated from long-distance seed dispersal. Additionally, more than half of the living trees were located in bogs (53 %), although they only contributed to 30 % of the area. Thus bogs may be either more suitable for regeneration, or the probability of seeds to survive the fire was higher in bogs than in upland forests. However, the mean estimated age of living trees was similar on bogs (49 years) and upland sites (47 years), and the number of remnant trees was equal (2 trees each). Moister substrates are better for the germination of P. uviferum seedlings (Figure 3.4), therefore it is likely that bogs offer better conditions for seedling establishment than upland sites. This is consistent with results from Filion & Morin (1996), who found that in P. mariana bog forests of Canada, seedlings were mostly associated with concave microsites, colonized by Sphagnum. However, more research is needed to assess how microtopography influences the survival of P. uviferum seedlings. In boreal conifer bog forests a close connection between crown fire severity and ecosystem recovery has been observed (Arseneault 2001; Johnstone et al. 2009). This is consistent with the high-severity crown fire pattern described in P. uviferum forests in upland areas by Holz & Veblen (2009). In our case, even the oldest remnant tree (81 years) could not have participated in the crown of the previous forests and must have survived as small seedling. Sexual maturity in P. uviferum occurs already in very small trees. Almost 50% of the observed regeneration (including small seedlings < 50 cm height) and 90% of the trees had visible reproductive structures. Despite the early fecundity, we have observed an extremely low density of P. uviferum trees (1.2 trees/ha) and females with regeneration (0.3 trees/ha). This can be partially explained by the clustered distribution of the trees at the landscape level. There was aggregation of P. uviferum trees at scales up to 30 m, which is in agreement with patterns of effective regeneration around seed trees (Figure 3.3). Female trees without regeneration showed no aggregation, because they were isolated from male trees (Figure 3.6), contrasting with male and female trees with regeneration that showed attraction at 46

59 Chapter Three: Potential for passive restoration of P. uviferum forests scales up to 30 m (Figure 3.6). This indicates that in the current condition of the disturbed landscape, a substantial proportion of female trees is too far away from males, also pointing to a limited effective dispersion of pollen. In addition, our results show that P. uviferum, like many other conifer species (Donoso et al. 2006; Gärtner et al. 2011), is a masting species with highly variable seed production between years (Figure 3.3). Thus a coincidence of poor seed production and fire-disturbance in one year may inhibit the recovery of the species` population after disturbance. In this context, the presence or absence of P. uviferum seed trees in the disturbed landscape is the determining biotic factor that limits the natural recovery of the species population in these disturbed sites. This is in agreement with our hypothesis, that the natural recovery of P. uviferum populations in burned areas is limited by seed availability, and highlights the importance of biological legacies in the process of ecosystem recovery (e. g. Franklin 1990; Franklin & MacMahon 2000; Keeton & Franklin 2005) Is passive restoration a suitable approach for disturbed P. uviferum forests? Natural regeneration from seeds of remnant trees can effectively assist the recovery of P. uviferum populations. However, the extremely low density and aggregated pattern of seed trees in disturbed landscapes would result in very long times for the restoration of P. uviferum populations. If we consider a circular area with 20 m radius around seed trees as effectively regenerated, almost 70 years after the fire only a 2.2% of the landscape has recovered passively towards P. uviferum forests. However, the low correlation between seed tree size and maximum distance or frequency of the regeneration, and the small minimum size of P. uviferum individuals with reproductive structures, highlight the importance also of small trees to assist the regrowth of these forests. As suggested by Holl & Aide (2011), the natural rate of recovery of an ecosystem is affected by the intrinsic ecosystem resilience to a disturbance, the level of ongoing human degradation, and the characteristics of the landscape. In these forests characterised by low resilience in relation to fire disturbance and low current human degradation, increasing the number of P. uviferum seed trees in the landscape could 47

60 Chapter Three: Potential for passive restoration of P. uviferum forests remove the principal biotic filter (sensu Hobbs & Norton 2004), which is currently retarding the natural recovery. Here, a mixed passive-active restoration approach appears a promising option for the restoration of disturbed P. uviferum forest. This may also increase genetic diversity of post-disturbance populations, which may otherwise originate from very few individuals. Planting male P. uviferum seedlings near female trees without regeneration, planting dispersed small groups comprising male and female trees, or the sowing of pre-treated seeds in disturbed forests would be inexpensive measures to accelerate the restoration of P. uviferum forests. The area inside a radius of 20 to 30 m from seed trees with regeneration can be left for natural recovery (passive restoration). 3.6 Acknowledgments We are especially grateful to the administration and staff of the Tantauco Park for constant support in the field. A. Caracciolo, N. Carrasco-Farias, G. Löffler, G. Fullá, J. Flade and D. Rieck assisted under difficult conditions in the field. Without their help this study could not have been done. For assistance in the laboratory we thank G. Löffler, T. Kahl, N. Briggs and O.J. Vidal. Jan Bannister received a DAAD-CONICYT scholarship to support his PhD studies at the University of Freiburg, where he participated in the graduate school Environment, Society and Global Change. This research was also financially supported through grants by the Georg-Ludwig-Hartig and Futuro foundations. 48

61 Chapter Four: Active restoration in P. uviferum forests Chapter Four The importance of micro-topography and nurse canopy for successful restoration planting of the slow-growing conifer Pilgerodendron uviferum Jan R. Bannister, Rafael E. Coopman, Pablo J. Donoso, Jürgen Bauhus Published in the journal Forests * Seedlings of Pilgerodendron uviferum established over mounds in fire-disturbed bog forests. Chaiguata Lake, Chiloé Island *Bannister JR, RE Coopman, PJ Donoso, J Bauhus The importance of microtopography and nurse canopy for successful restoration planting of the slow-growing conifer Pilgerodendron uviferum. Forests 4: Available in: 49

62 Chapter Four: Active restoration in P. uviferum forests 4.1 Abstract Recent studies have shown that, owing to a lack of seed trees, the natural rate of recovery of fire-disturbed bog forests previously dominated by the endemic and endangered conifer Pilgerodendron uviferum (D. Don) Florin is extremely slow. Hence, increasing the number of seed trees in the landscape through restoration planting could remove the principal biotic filter limiting recovery of these forests. Here, we analyzed how the success of restoration plantings may be improved through the choice or manipulation of microsites in P. uviferum forests on Chiloé Island in North Patagonia. For this purpose, we manipulated microtopography in water-logged sites in bogs (mounds, flat terrain, mineral soil) and changed canopy conditions (gaps, semiopen, closed canopy) in upland sites with better drainage. In bogs, there was no significant effect of microtopography on growth and survival of P. uviferum plantings. However, fluorescence measurements indicated lower stress in seedlings established on mounds. Seedlings in upland areas established beneath a nurse canopy had lower mortality and higher relative shoot growth, foliar nutrients, photosynthetic light use efficiency and chlorophyll fluorescence values than those planted in the open. This indicates that seedlings of the slow growing P. uviferum may tolerate extremely wet conditions, yet suffer from stress when grown in the open. Here, the removal of canopy appeared to have also removed or reduced mycorrhizal networks for seedlings, leading to poorer nutrition and growth. Based on these results, recommendations for restoration plantings in highly degraded P. uviferum forests are presented. Keywords: Active restoration, conifer bog forests, Chiloé Island, North Patagonia, seedling growth, Sphagnum. Article available in: 50

63 Chapter Five: Synthesis Chapter Five Synthesis Competition between Pilgerodendron uviferum seedlings and bryophytes in undisturbed bog forests. Río Zorra Valley, Chiloé Island. 51

64 Chapter Five: Synthesis The main purpose of this doctoral thesis was to study fire-disturbed and undisturbed old-growth Pilgerodendron uviferum bog forests in North Patagonia to develop the scientific basis for future conservation and restoration strategies. This doctoral thesis present data on structure of old-growth P. uviferum forests and present measurements for the oldest trees recorded for this species. Additionally, this is the first work that provides data on nutrition and light environment of the species and demonstrates that the species does not require intensive disturbances to persist in bog forests (Chapter two). Furthermore, this is also the first study of processes for passive restoration in disturbed P. uviferum forests, provides the first quantification of seed dispersal and germination preferences of P. uviferum, and analyzes for first time the passive restoration potential in disturbed southern bog forests (Chapter three). This study provides the first data of photosynthetic traits for the species (Chapter four). The results of the present doctoral thesis suggest that a mixed passive-active restoration approach appears to be a promising option for the restoration of these forests. The following synthesis aims to focus the discussion on three important aspects of restoration that were identified in the previous chapters: first, understanding the processes occurring in undisturbed old-growth forests, analysing the natural rate of recovery of fire-disturbed forests, and exploring options for restoration. This doctoral thesis, probably like many others, started with a few research questions. The initial steps undertaken here to understand these ecosystems have led to many more questions. Therefore, I aim to develop here some recommendations for further research in these ecosystems. 5.1 Stand dynamics in Pilgerodendron uviferum old-growth forests and the effect of fire-disturbances The findings of this doctoral thesis contradict the common concept of succession in P. uviferum dominated forests, which postulates an improvement of drainage conditions under P. uviferum and subsequent colonisation of these sites by other species and replacement of P. uviferum (Cruz & Lara 1981; Bannister et al. 2008). For this improvement in drainage conditions, Cruz & Lara (1981) described four different forest types which differ in the importance of P. uviferum as a dominant species: 52

65 Chapter Five: Synthesis turbera con ciprés (bog with P. uviferum); bosque abierto de ciprés (open P. uviferum forest); bosque de ciprés-tepú (mixed forest dominated by P. uviferum and T. stipularis); and bosque de tepú con ciprés muy ralo (T. stipularis forest with sparse P. uviferum trees). In accordance with this commonly held view, old-growth stands of this species would represent only a transitional phase in forest succession. In chapter two, it was shown that P. uviferum is a highly stress-tolerant conifer that can grow under adverse conditions of low light, low nutrient availability and water logging. In the absence of intensive large-scale disturbances, longevity, capacity for extremely slow growth, tolerance to shade and low nutrient availability and also decay-resistance may be the primary mechanisms that lead to the persistence of P. uviferum in these unproductive sites. This is the pre-condition for the development of old-growth forests in which this species maintains its dominance through continuing regeneration. Furthermore, in chapter four it was shown that P. uviferum seedlings are very efficient in the use of light at conditions of lower light availability beneath the canopy of associated broadleaved species. The tolerance to conditions of low nutrient availability may be facilitated by mycorrhizal networks, which would be most important for young, establishing trees. In addition, such mycorrhizal networks may provide seedlings growing in shaded conditions also with carbohydrates (Simard 2009; Simard et al. 2012). This would further support stability of P. uviferum communities. Possible mechanisms for nutrient uptake and below-ground interactions in this species, which are important in this extremely nutrient poor environment, obviously deserve more intensive research than could be done in this study. In any case, it could be demonstrated that old-growth P. uviferum forests are not a transitional phase in forest succession and may be maintained in the landscape for many centuries or millennia. Moreover, paludification in these forests contradicts the idea of drainage improvement of Cruz & Lara (1981). In contrast, in the absence of major catastrophic disturbances, ongoing paludification in bog forests will lead to decreasing forest productivity with time (Fenton et al. 2005; Lavoie et al. 2005). Under these conditions, peat accumulation is primarily under strong autogenic control (succession), whereas peat reduction is mostly under allogenic control such as through fires (Simard et al. 2007). 53

66 Chapter Five: Synthesis Therefore, a better alternative to explain the dynamics of P. uviferum forests and to better understand the processes driving dynamics under undisturbed conditions may be based on a division of P. uviferum forests into different types, in bog and upland forests, that follow different successional pathways (Figure 5.1). In the absence of catastrophic disturbances (such as stand replacing fires), the dynamics of P. uviferum differ between both forest types, although sometimes it is difficult to delineate them in the field, especially in the ecotone between flat and sloping areas. In bog forests, P. uviferum colonizes bogs especially hummocks with better drainage conditions and lower competition from bryophytes. As the succession proceeds, the species maintains continuous and abundant recruitment of regeneration, in constant competition with Sphagnum, which is typically growing several centimetres per year as a consequence of ongoing paludification. As Cruz & Lara (1981) suggested, many times the site is so poor that the maximal expression of this type of forests are trees of low stature but advanced ages. Bog forests include the types bog with P. uviferum and open P. uviferum forest described by Cruz & Lara (1981) (Figure 5.1). On the other hand, in upland forests the species colonizes extremely moist slopes or hills (e.g. left by glaciers retirement) with better drainage and less Sphagnum competition than in bog forests. Consequently, these forests accumulate higher basal areas and dense, complex and eventually highly combustible understoreys of Tepualia stipularis and other Myrtaceae. Additionally, the canopy layer is dominated by a higher variety of tree species with heights up to 20 m. Despite the increasing abundance of other species, P. uviferum can maintain a continuous, albeit not abundant recruitment of regeneration and therefore its dominance in the landscape for many centuries or millennia, possibly under a disturbance regime characterised by gap dynamics. Upland forests include the types mixed forest dominated by P. uviferum and T. stipularis and T. stipularis forest with sparse P. uviferum trees described by Cruz & Lara (1981). Obviously, there are also upland sites in these landscapes with very different forest vegetation without P. uviferum. Perhaps these are characterised by a more nutrientrich substrate, at which other species are more competitive. However, it is evident that more research is needed in undisturbed P. uviferum forests in order to develop a complete model for the dynamics of the species. 54

67 Chapter Five: Synthesis Figure 5.1 General conceptual diagram showing a hypothetical model of succession in Pilgerodendron uviferum forests on Chiloé Island. The P. uviferum forest types defined by Cruz & Lara (1981) are in grey font colour. In the high rainfall environment of coastal Patagonia, fires are most likely restricted to human activity, even though individual trees may be struck by lightning (Holz & Veblen 2009). Pilgerodendron uviferum does not show any specific adaptations to fire disturbance such as thick bark, resprouting ability, protected seeds or a soil seed bank. In this context, large scale and human made fire disturbances threaten the local persistence of the species (Chapter two and three). In these disturbed forests, the regrowth is often dominated by species such as D. winteri, W. trichosperma, N. betuloides, N. nitida and T. stipularis, which corresponds to a novel successional pathway. These forests are equivalent to the type mixed second-growth forest with P. uviferum dead standing trees (Cruz & Lara 1981). In these post-disturbance regrowth stands, live trees of P. uviferum are rare and regeneration is negligible, showing that 55

68 Chapter Five: Synthesis high-severity fires can eliminate the species from parts of the landscape, where neither propagules nor seed trees survive. 5.2 The natural rate of recovery of Pilgerodendron uviferum forests after firedisturbances The natural rate of recovery of an ecosystem is affected by the intrinsic ecosystem resilience to a particular type of disturbance, the level of ongoing human degradation, and the characteristics of the landscape (Holl & Aide 2011). Since there may be long periods of time without significant disturbances in the coastal areas of North Patagonia (Lumley & Switsur 1993), P. uviferum forests can accumulate high biomass over time, especially in upland areas (Chapter 2). During the slow sequence from the exploitation to the conservation phase in an adaptive cycle (Holling 2001), ecosystems accumulate biomass to a point where the resilience is minimal and the ecosystems becomes an accident waiting to happen. Forests that have rarely experienced stand level disturbance may be especially sensitive to intensive disturbances, in particular if these represent types of disturbances the systems have not evolved with (Burns 1993; Parish & Antos 2006). In chapter three it was shown that the natural rate of recovery of P. uviferum populations in burned areas is limited by seed availability. It was also shown that the availability of suitable substrates for germination at establishment was not limiting the recovery of P. uviferum populations in the burned areas. These findings indicate that natural regeneration from seed trees can assist the recovery of P. uviferum populations following fire disturbance, but their effect is limited at a landscape level (< 30 m). Furthermore, large scale fires not only kill existing P. uviferum trees but also destroy propagules and thus threaten the local persistence of the species. The findings of chapter two and three confirm the low resilience of P. uviferum undisturbed forests in relation to fire disturbance. At present, 72.5% of the 930,074 ha of Chilean P. uviferum dominated forests are inside protected state areas (CONAF 2011). This situation suggests that the current human degradation in these forests is low. However, the historical use of this species accompanied with illegal harvesting even in reserves and on southern islands are 56

69 Chapter Five: Synthesis ongoing problems for the conservation of this species in Chile (Szeicz et al. 2000; Bannister 2004). There are no studies that have quantified the percentage of burned P. uviferum forests, but there is a general agreement between researchers that it is extremely high (Lara et al. 2006). Particularly during the 20 th and 21 st centuries there has been a considerable increase in fire frequency in North-Patagonia (Holz & Veblen 2011). Furthermore, under the warming and drying climate scenarios forecast for south-western South America, humans are likely to continue to have an amplifying effect on wildfires in Patagonian rain forests (Holz & Veblen 2011). The increase in fire frequency, the low resilience to fire disturbances of these ecosystems and the slow recovery after burning (Chapters two and three) raise important questions about the viability and conservation problems that these ecosystems will face in the future. At the landscape level, P. uviferum forests are divided in two topographical areas, bog areas located in flat areas on raised peat bogs, and upland areas on hills shaped by till with better drainage. Holz & Veblen (2009) documented one pattern of fire occurrence for each of these two topographical areas in disturbed P. uviferum forests: a) infrequent high-severity crown fires in association with severe, extensive drought in upland forest and b) more frequent low-severity surface fires in association with less severe droughts in bog forests. In the ecotone between upland and bog areas there is a fire regime of mixed severity in which both surface and crown fires occur. The close connection between crown fire severity and ecosystem recovery that has been observed in boreal conifer forests (Arseneault 2001; Johnstone et al. 2009) is analogue to the high-severity crown fire pattern described in upland areas by Holz & Veblen (2009). Apparently, the intensive crown and surface fires in the study area 70 years ago exceeded the resistance of P. uviferum to fire, especially in upland forests, which are now characterised by lower rates of recovery (Chapters two and three). Thus bogs, were the fires have lower severity, may be either more suitable for regeneration, or the probability of seeds to survive the fire is higher in bogs than in upland forests (Chapter three). This underpins the importance of biological legacies such as seed trees for the recovery of disturbed sites (Cruz & Lara 1981; Franklin 1990; Bannister et al. 2008; Franklin et al. 2002) (Figure 5.1). 57

70 Chapter Five: Synthesis 5.3 Towards a restoration strategy for Pilgerodendron uviferum bog forests in North Patagonia As it was stated in chapter one, restoration ecology involves returning a degraded system to some desired state and directing system development along a desired trajectory (Hobbs & Norton 1996). Since undisturbed P. uviferum forests have complex structures and extremely slow growth rates (Chapter two), it is not possible to redevelop them in a short or even medium term through restoration of disturbed areas. These ecosystems do not have problems of erosion, because the soils are mostly covered by mosses, shrubs and other vegetation. Therefore, no immediate action is required to halt further degradation. To restore the complex pre-disturbance forest structures and to provide the same ecosystem services than the original state, it is necessary to promote the processes that permit the re-establishment of P. uviferum at the landscape level. To achieve that, it is important to understand the principal processes occurring under undisturbed conditions (Landres et al. 1999; Kuuluvainen et al. 2002). As was mentioned in the introduction, to identify of restoration goals, three questions should be answered (Holl & Aide 2011). Here, I will revisit these questions. (1) If we take a passive restoration approach, what results do we expect? This doctoral thesis shows that 70 years after fire, seed trees were extremely infrequent at the landscape level (Chapter three). While natural regeneration from seed trees may assist the recovery of P. uviferum populations following fire disturbance, this process is very limited at a landscape level. Therefore, a purely passive restoration approach is deemed insufficient and supplementary planting of seed trees is suggested to assist recovery of P. uviferum forests. (2) If intervention is necessary, how and when should we intervene to achieve the project goals? Some ecosystems need minor interventions for initiating or accelerating natural regeneration processes, but others need a major effort in order to achieve their restoration. The type of restoration response needed will depend on restoration thresholds (Figure 1.4) that have been crossed and the degree of effort needed to 58

71 Chapter Five: Synthesis restore the system to a particular state will depend on the ecological filters that are retarding the natural recovery of the forest (Whisenant 1999; Hobbs & Harris 2001; Hobbs & Norton 2004). The disturbed P. uviferum forests in the study area have crossed the restoration threshold caused by biotic interactions (lack of seed trees) but are still on the left side of the threshold caused by abiotic limitations (there are no substrate problems for regeneration). Therefore they require only manipulation of the vegetation in order to recover (Figure 1.4). Here, increasing the number of P. uviferum seed trees in the landscape through low density planting of seedlings to complement existing parent trees could remove the principal biotic filter (sensu Hobbs & Norton, 2004), which is currently retarding the natural recovery of the stands. Hence, a mixed passive-active restoration approach arises as the most promising option to restore P. uviferum forests in North Patagonia. This may also increase genetic diversity of post-disturbance populations, which may otherwise originate from very few individuals. In this context, an area inside a radius of 20 to 30 m from seed trees with regeneration should be left for natural recovery (passive restoration). Then, planting male P. uviferum seedlings near female trees without regeneration, planting dispersed small groups comprising male and female trees, or the sowing of pre-treated seeds in disturbed forests would be inexpensive measures to accelerate the restoration of P. uviferum forests. But how should these trees be planted? One of the most important factors influencing seedling performance are site conditions; therefore, chapter four focused on how different site conditions affect the performance of P. uviferum restoration plantings. In bogs, seedlings established on elevated micro-sites performed slightly better, but this may not justify the effort to create such sites mechanically. In upland areas of better drainage, seedlings established beneath some canopy performed much better and had lower mortality (Chapter four). Hence, clearing of vegetation to establish seedlings should be avoided or done only partially to maintain important ecological interactions with existing vegetation. (3) How can restoration resources be used most efficiently at a landscape or regional scale? What to do in ecosystems with low resilience and high degradation, where restoration is likely to be extremely expensive and the outcome uncertain is a 59

72 Chapter Five: Synthesis difficult problem (Hobbs et al. 2009; Holl & Aide 2011), especially when the ecosystems are in remote areas (Higgs & Roush 2011). For this type of cases, the multiscale approach used in the present doctoral thesis, in which the processes occurring in disturbed and undisturbed sites are studied prior to plan a restoration program, presents a useful option to reduce costs and improve restoration effectiveness. In this doctoral thesis, a mixed passive-active restoration approach is suggested, where active restoration efforts such as planting and sowing should take the distribution patterns of seed trees into account. Chapter four provides useful guidance to plant P. uviferum seedlings in disturbed areas. In bogs, seedlings may be planted in normal flat areas and or natural mounds and small elevated areas be used to avoid waterlogging. In upland areas the creation of gaps or strips is not an effective silvicultural option. Instead, planting seedlings under canopy protection or between canopies of other trees like D. winteri, W. trichosperma, N. nitida or T. stipularis appears to be useful to ensure establishment and faster growth of P. uviferum seedlings. These suggestions are very specific findings that will help managers to improve future afforestation activities to restore the species in disturbed bog forests of North Patagonia. 5.4 Recommendations for further research The present doctoral thesis presents a significant contribution to the knowledge about the structure and dynamics of undisturbed P. uviferum bog forests, and about the restoration approaches for the species in disturbed sites. However, P. uviferum bog forests are still one of the least studied conifer forests of the southern cone of South America. Still much research is needed to improve our understanding of processes in these ecosystems. Below, I propose some areas for future research: Undisturbed P. uviferum bog forests Until now there are no studies that have quantified the proportion of the total area of P. uviferum forests that are actually undisturbed. Identifying the location of undisturbed sites should be a high priority, because it would permit to evaluate the true conservation status of the species at the national scale, and would improve 60

73 Chapter Five: Synthesis the planning of future conservation strategies and a better location of protected areas. The process of paludification, and the complex and dynamic process of competition between bogs and forested areas should receive more attention in future research. This is critical to better understand the dynamics of southern bog forests (Chapter two). This aspect is of high importance in the face of climate change, because these ecosystems store and continue to sequester high amounts of carbon. Apart from P. uviferum, one of the most important tree species in these ecosystems is T. stipularis. This species creates very complex structures owing to its horizontal growth habit. Due to the difficulties to enter undisturbed stands that are dominated in the understorey by T. stipularis, there have been so far no studies that focus on the structure and dynamics of this species. In addition, there is no established inventory method to quantify volume of biomass of this species. It is a priority to fill this knowledge gap, since this is the main tree species used for firewood production in North Patagonia. Until now, there was very limited research on peatland areas and bog forests in the southern hemisphere. It would be important to explore the differences that exist between the dynamics and processes occurring in southern and northern, boreal bog forests to develop a more generic understanding of this ecosystem type In particular the mechanism and below-ground interactions that offer adaptations to these nutrient-poor environments appear to be a rewarding area of research. According to Rozzi et al. (2008), Patagonian islands hosts outstanding non-vascular floristic richness, with > 5% of the world s bryophytes on < 0.01 % of the Earth s land surface. In this context, more research is needed to assess the importance of southern bog forests for sustaining non-vascular biodiversity. 61

74 Chapter Five: Synthesis Ecological restoration in disturbed P. uviferum dominated bog forests One of the main limitations at the time of devising the active restoration experiment of chapter four was the low number of seedlings available in nurseries of southern Chile. To develop a restoration programme for an extensive area, the capacity to raise seedlings of this species would have to be dramatically increased. Therefore, future research should also investigate methods for effective seed collection, storage and treatments as well as efficient nursery production systems for container or bare-root planting stock. Here, the inoculation with suitable types of mycorrhiza may be very important for establishment success. The field trials reported in chapter four present important initial steps to gain some basic information on ways for successful establishment of P. uviferum seedlings in disturbed forests of North Patagonia. However, similar trials should be replicated to cover the heterogeneous conditions across the region to develop more robust results that permit generalization of recommendations. Sowing of pre-treated seeds maybe an inexpensive measure to accelerate the restoration of disturbed P. uviferum forests, if abundant seed is available. Future research may test the effectiveness of direct sowing in the field to evaluate its potential as an alternative to restore disturbed forests. Moist substrates comprising Sphagnum or mineral soil were the best for germination of P. uviferum seeds, and these are precisely the most frequent substrates in the study area. However, safe sites for germination are not necessarily safe sites for seedling survival and therefore future research should investigate effects of micro-sites on seedling survival and development. This appears to be quite important in this environment were height growth of Sphagnum and other mosses may exceed that of P. uviferum seedling. 62

75 Chapter Six: References Chapter Six References Excluding chapters two and four* Fire-disturbed Pilgerodendron uviferum upland forests Cerro Mirador, Chiloé Island. *The references for chapters two and four are included in the original published articles. 63

76 Chapter Six: References Aravena JC Reconstructing Climate Variability using Tree Rings and Glacier Fluctuations in the Southern Chilean Andes. PhD thesis. Ontario: University of Western Ontario. 220p. Armesto JJ, Bustamante-Sánchez MA, Díaz MF, González ME, Holz A, Nuñez-Avila M, Smith-Ramirez C Fire Disturbance Regimes, Ecosystem Recovery and Restoration Strategies in Mediterranean and Temperate Regions of Chile. In: Fire Effects on Soils and Restoration Strategies. Cerda A, Robichaud PR. (Eds.). Science Publishers, United States of America. pp Aronson J, Dhillion S, Floc h E On the Need to Select an Ecosystem of Reference, However Imperfect: A Reply to Pickett and Parker. Restoration Ecology 3:1 3. Arseneault D Impact of Fire Behaviour on Postfire Forest Development in a Homogeneous Boreal Landscape. Canadian Journal of Forest Research 31: Bannister JR Estado de Conservación de Pilgerodendron uviferum (D Don) Florin en el Área Norte de la Cordillera de Pirulil, Isla Grande de Chiloé, X Región. For. Eng. thesis. Valdivia: Universidad Austral de Chile. 79p. Bannister JR, Lara A, Le Quesne C Estructura y Dinámica de Bosques de Pilgerodendron uviferum Afectados por Incendios en la Cordillera de la Costa de la Isla Grande de Chiloé. Bosque 29: Bannister JR, Vidal OJ, Teneb E, Sandoval V Latitudinal Patterns and Regionalization of Plant Diversity Along a 4270-km Gradient in Continental Chile. Austral Ecology 37: Bannister JR, Donoso PJ, Bauhus J Persistence of the Slow Growing Conifer Pilgerodendron uviferum in Old-Growth and Fire-disturbed Southern Bog Forests. Ecosystems (in press). DOI: /s Bond WJ The Tortoise and the Hare: Ecology of Angiosperm Dominance and Gymnosperm Persistence. Biological Journal of the Linnean Society 36: Burns BR Fire-Induced Dynamics of Araucaria araucana-nothofagus antarctica Forest in the Southern Andes. Journal of Biogeography 20: Carmona MR, Aravena JC, Bustamante-Sanchez MA, Celis-Diez JL, Charrier A, Díaz IA, Díaz-Forestier J, Díaz MF, Gaxiola A, Gutiérrez AG, Hernandez-Pellicer C, Ippi S, Jaña-Prado R, Jara-Arancio P, Jiménez J, Manuschevich D, Necochea P, Nuñez-Avila 64

77 Chapter Six: References M, Papic C, Pérez C, Pérez F, Reid S, Rojas L, Salgado B, Smith-Ramírez C, Troncoso A, Vásquez RA, Willson MF, Roíz R, Armesto JJ Estación Biológica Senda Darwin: Investigación Ecológica de Largo Plazo en la Interfase Ciencia-Sociedad. Revista Chilena de Historia Natural 83: CONAF Catastro de los Recursos Vegetacionales Nativos de Chile. Monitoreo de Cambios y Actualizaciones. Período Santiago, Chile. 28p. CONAF, CONAMA, Universidad Austral de Chile, Pontificia Universidad Católica de Chile, Universidad Católica de Temuco Catastro y Evaluación de los Recursos Vegetacionales Nativos de Chile. Informe Nacional con Variables Ambientales. Santiago, Chile. 88p. Coomes DA, Allen RB, Bentley WA, Burrows LE, Canham CD, Fagan L, Forsyth DM, Gaxiola-Alcantar A, Parfitt RL, Ruscoe WA, Wardle DA, Wilson DJ, Wright EF The Hare, the Tortoise and the Crocodile: The Ecology of Angiosperm Dominance, Conifer Persistence and Fern Filtering. Journal of Ecology 93: Cornett MW, Reich PB, Puettmann KJ, Frelich LE Seedbed and Moisture Availability Determine Safe Sites for Early Thuja occidentalis (Cupressaceae) Regeneration. American Journal of Botany 87: Cruz G, Lara A Tipificación, Cambio de Estructura y Normas de Manejo para Ciprés de las Guaitecas (Pilgerodendron uviferum D. Don Florin) en la isla Grande de Chiloé. For. Eng. thesis. Santiago: Universidad de Chile. 215p. di Castri F, Hajek E Bioclimatología de Chile. Vicerrectoría Académica de la Universidad Católica de Chile, Santiago, Chile. 128p. Donoso C, Cortes M, Soto L Antecedentes sobre Semillas y Germinación de Alerce, Ciprés de las Guaitecas, Ciprés de la Cordillera y Tineo. Bosque 3: Donoso C, Lara A, Escobar B, Premoli A, Souto A Fitzroya cupressoides (Molina) I.M. Johnst. In: Las Especies Arbóreas De Los Bosques Templados De Chile y Argentina, Autoecología. Donoso, C. (Ed). Marisa Cuneo, Valdivia, Chile. pp Fenton N, Lecomte N, Legare S, Bergeron Y Paludification in Black Spruce (Picea mariana) Forests of Eastern Canada: Potential Factors and Management Implications. Forest Ecology and Management 213: Filion J, Morin H Distribution Spatiale de la Régénération de L épinette Noire 8 Ans Après un Feu en Forêt Boréale (Québec). Canadian Journal of Forest Research 65

78 Chapter Six: References 26: Franklin JF Biological Legacies: a Critical Management Concept from Mt. St. Helens. Transactions of the Fifty-fifth North American Wildlife and Natural Resources Conference 55: Franklin JF, MacMahon JA Messages from a Mountain. Science 288: Franklin JF, Spies TA, Pelt RV, Carey AB, Thornburgh DA, Berg DR, Lindenmayer DB, Harmon ME, Keeton WS, Shaw DC, Bible K, Chen J Disturbances and Structural Development of Natural Forest Ecosystems with Silvicultural Implications, using Douglas-fir Forests as an Example. Forest Ecology and Management 155: Gärtner SM, Lieffers VJ, Macdonald SE Ecology and Management of Natural Regeneration of White Spruce in the Boreal Forest. Environmental Reviews 19: Gorham E Northern Peatlands: Role in the Carbon Cycle and Probable Responses to Climatic Warming. Ecological Applications 1:182. Gunderson LH Ecological Resilience - in Theory and Application. Annual Review of Ecology Evolution and Systematics 31: Gunnarsson U, Rydin H Demography and Recruitment of Scots pine on Raised Bogs in Eastern Sweden and Relationships to Microhabitat Differentiation. Wetlands 18: Hanssen KH Natural Regeneration of Picea abies on Small Clear-cuts in SE Norway. Forest Ecology and Management 180: Hechenleitner P, Gardner M, Thomas P, Echeverria C, Escobar B, Brownless P, Martínez C Plantas Amenazadas del centro-sur de Chile. Distribución, Conservación y Propagación. Universidad Austral de Chile y Real Jardín Botánico de Edimburgo. 187p. Herrera JM, García D The Role of Remnant Trees in Seed Dispersal Through the Matrix: Being Alone is not Always so Sad. Biological Conservation 142: Higgs ES, Roush WM, Restoring Remote Ecosystems. Restoration Ecology 19: Hobbs R, Norton DA Ecological Filters, Thresholds, and Gradients in Resistance to Ecosystem Reassembly. In: Assembly Rules and Restoration Ecology: Bridging the 66

79 Chapter Six: References Gap Between Theory and Practice. Temperton VM, Hobbs R, Nuttle T, Halle S. (Eds.). Island Press, Washington D.C. pp Hobbs RJ, Harris JA Restoration Ecology: Repairing the Earth s Ecosystems in the New Millennium. Restoration Ecology 9: Hobbs RJ, Higgs E, Harris JA Novel Ecosystems: Implications for Conservation and Restoration. Trends in Ecology & Evolution 24: Hobbs RJ, Norton DA Towards a Conceptual Framework for Restoration Ecology. Restoration Ecology 4: Holl KD, Aide TM When and Where to Actively Restore Ecosystems? Forest Ecology and Management 261: Holling CS Resilience and Stability of Ecological Systems. Annual Review of Ecology Evolution and Systematics 4:1 23. Holling CS Understanding the Complexity of Economic, Ecological, and Social Systems. Ecosystems 4: Holz A Climatic and Human Influences on Fire Regimes and Forest Dynamics in Temperate Rainforests in Southern Chile. PhD. thesis. Boulder: University of Colorado. 276p. Holz A, Veblen TT Pilgerodendron uviferum: The Southernmost Tree-Ring Fire Recorder Species. Ecoscience 16: Holz A, Veblen TT The Amplifying Effects of Humans on Fire Regimes in Temperate Rainforests in Western Patagonia. Palaeogeography, Palaeoclimatology, Palaeoecology 311: Hörnberg G, Ohlson M, Zackrisson O Stand Dynamics, Regeneration Patterns and Long-Term Continuity in Boreal Old-Growth Picea abies Swamp-Forests. Journal of Vegetation Science 6: Hörnberg G, Ohlson M, Zackrisson O Influence of Bryophytes and Microrelief Conditions on Picea abies Seed Regeneration Patterns in Boreal Old-Growth Swamp Forests. Canadian Journal of Forest Research 27: Jennings SM, Hickey JE, Candy SG Comparison of Regeneration Success of Alternative Silvicultural Treatments in Blackwood Swamps. Tasforests 12: Johnstone J, Boby L, Tissier E, Mack M, Verbyla D, Walker X Postfire Seed Rain of Black Spruce, a Semiserotinous Conifer, in Forests of Interior Alaska. Canadian 67

80 Chapter Six: References Journal of Forest Research 39: Kashian DM, Gregory Corace R, Shartell LM, Donner DM, Huber PW Variability and Persistence of Post-Fire Biological Legacies in Jack Pine-Dominated Ecosystems of Northern Lower Michigan. Forest Ecology and Management 263: Keeton WS, Franklin JF Do Remnant Old-Growth Trees Accelerate Rates of Succession in Mature Douglas-fir forests? Ecological Monographs 75: Klinger LF The Myth of the Classic Hydrosere Model of Bog Succession. Arctic and Alpine Research 28:1 9. Kuuluvainen T, Aapala K, Ahlroth P, Kuusinen M, Lindholm T, Sallantaus T, Siitonen J, Tukia H Principles of Ecological Restoration of Boreal Forested Ecosystems: Finland as an Example. Silva fennica 36: Landres PB, Morgan P, Swanson FJ Overview of the Use of Natural Variability Concepts in Managing Ecological Systems. Ecological Applications 9: Lara A, Donoso C, Escobar B, Rovere A, Premoli A, Soto DP, Bannister JR Pilgerodendron uviferum (D. Don) Florin. In: Las Especies Arbóreas De Los Bosques Templados De Chile y Argentina, Autoecología. Donoso C. (Ed.). Marisa Cuneo, Valdivia, Chile. pp Lara A, Echeverría C, Thiers O, Huss E, Escobar B, Tripp K, Zamorano C, Altamirano A Restauración Ecológica de Coníferas Longevas: El Caso del Alerce (Fitzroya cupressoides) en el Sur de Chile. In: Restauración de Bosques en América Latina. González-Espinosa M, Rey-Benayas JM, Ramírez-Marcial N. (Eds.) Mundi-Prensa México, México. pp Lavoie M, Paré D, Fenton N, Groot A, Taylor K Paludification and Management of Forested Peatlands in Canada: a Literature Review. Environmental Reviews 13: Lumley S, Switsur R Late Queaternary Chronology of the Taitao Peninsula, Southern Chile. Journal of Quaternary Science 8: Manning AD, Fischer J, Lindenmayer DB Scattered Trees are Keystone Structures Implications for Conservation. Biological Conservation 132: Martínez O Flora y Fitosociología de un Relicto de Pilgerodendron uvifera (D. Don) Florin en el Fundo San Pablo de Tregua (Valdivia-Chile). Bosque 4:3 11. Moroni MT, Hagemann U, Beilman DW Dead Wood is Buried and Preserved in a 68

81 Chapter Six: References Labrador Boreal Forest. Ecosystems 13: Morrison EB, Lindell CA Active or Passive Forest Restoration? Assessing Restoration Alternatives with Avian Foraging Behavior. Restoration Ecology 19: NWWG The Canadian Wetland Classification System, National Wetlands Working Group (Ed.). University of Waterloo, Canada. 76p. Ohlson M Growth and Nutrient Characteristics in Bog and Fen Populations of Scots pine (Pinus sylvestris). Plant and Soil 172: Otero L La Huella del Fuego. Historia de los Bosques Nativos. Poblamiento y Cambios en el Paisaje del Sur de Chile. Pehuén, Santiago, Chile. 171p. Parish R, Antos JA Slow Growth, Long-Lived Trees, and Minimal Disturbance Characterize the Dynamics of an Ancient, Montane Forest in Coastal British Columbia. Canadian Journal of Forest Research 36: Pérez CA, Armesto JJ, Torrealba C, Carmona MR Litterfall Dynamics and Nitrogen Use Efficiency in Two Evergreen Temperate Rainforests of Southern Chile. Austral Ecology 28: Pickett STA, Parker VT Avoiding the Old Pitfalls: Opportunities in a New Discipline. Restoration Ecology 2: Rey Benayas JMR, Bullock JM, Newton AC Creating Woodland Islets to Reconcile Ecological Restoration, Conservation, and Agricultural Land Use. Frontiers in Ecology and the Environment 6: Ribbens E, Silander JA, Pacala SW Seedling Recruitment in Forests: Calibrating Models to Predict Patterns of Tree Seedling Dispersion. Ecology 75: Ripley BD Modelling Spatial Patterns. Journal of the Royal Statistical Society B 39: Roig F, Boninsegna J Estudios sobre el Crecimiento Radial, Basal, en Altura y de las Condiciones Climáticas que Afectan el Desarrollo de Pilgerodendron uviferum. Revista Chilena de Historia Natural 64: Roy V, Bernier PY, Plamondon AP, Ruel JC Effect of Drainage and Microtopography in Forested Wetlands on the Microenvironment and Growth of Planted Black Spruce Seedlings. Canadian Journal of Forest Research 29: Rozzi R, Armesto JJ, Goffinet B, Buck W, Massardo F, Silander J, Arroyo MT, Russell S, 69

82 Chapter Six: References Anderson CB, Cavieres LA, Callicott JB Changing Lenses to Assess Biodiversity: Patterns of Species Richness in Sub-Antarctic Plants and Implications for Global Conservation. Frontiers in Ecology and the Environment 6: Rydin H, Jeglum J.K The Biology of Peatlands. Oxford University Press, Oxford; New York. 343p. Schiappacasse I, Nahuelhual L, Vásquez F, Echeverría C Assessing the Benefits and Costs of Dryland Forest Restoration in Central Chile. Journal of Environmental Management 97: SER The SER International Primer on Ecological Restoration. Society of Ecological Restoration. 15p. Shaw AJ, Cox CJ, Boles SB Global Patterns in Peatmoss Biodiversity. Molecular Ecology 12: Simard M, Lecomte N, Bergeron Y, Bernier PY, Paré D Forest Productivity Decline Caused by Successional Paludification of Boreal Soils. Ecological Applications 17: Simard SW The Foundational Role of Mycorrhizal Networks in Self-Organization of Interior Douglas-fir Forests. Forest Ecology and Management 258:S95 S107. Simard SW, Beiler KJ, Bingham MA, Deslippe JR, Philip LJ, Teste FP Mycorrhizal Networks: Mechanisms, Ecology and Modelling. Fungal Biology Reviews 26: Solís C, Becerra J, Flores C, Robledo J, Silva M Antibacterial and Antifungal Terpenes from Pilgerodendron uviferum (D. DON) Florin. Journal of the Chilean Chemical Society 49: Soto DP, Bannister JR, Ríos A, Le Quesne C Nuevos Registros de Poblaciones Amenazadas de Pilgerodendron uviferum D.Don (Florin) en su Límite Norte en la Cordillera de la Costa chilena. Gayana Botanica 67: Stokes MA, Smiley TL An Introduction to Tree-Ring Dating. University of Chicago Press, Chicago, Illinois. 73p. Stoyan D, Wagner S Estimating the Fruit Dispersion of Anemochorous Forest Trees. Ecological Modelling 145: Suding K Alternative States and Positive Feedbacks in Restoration Ecology. Trends in Ecology and Evolution 19: Szeicz JM, Haberle SG, Bennett KD Dynamics of North Patagonian Rainforests 70

83 Chapter Six: References from Fine-Resolution Pollen, Charcoal and Tree-Ring Analysis, Chonos Archipelago, Southern Chile. Austral Ecology 28: Szeicz JM, Lara A, Díaz S, Aravena JC Dendrochronological Studies of Pilgerodendron uviferum in Southwestern South America. In: Dendrocronología en América Latina. Roig FA. (Ed.). EDIUNC, Mendoza, Argentina. pp Thompson I, Mackey B, McNulty S, Mosseler A Forest Resilience, Biodiversity, and Climate Change A Synthesis of the Biodiversity/Resilience/Stability Relationship in Forest Ecosystems, Technical Series. Secretariat of the Convention on Biological Diversity, Montreal. 67p. Van Andel J, Aronson J Restoration Ecology. Blackwell Publishing. 319p. Villagrán C Late Quaternary Vegetation of Southern Isla Grande de Chiloé, Chile. Quaternary Research 29: Walker LR Integration of the Study of Natural and Anthropogenic Disturbances Using Severity Gradients. Austral Ecology 36: Walter KS, Gillet HJ IUCN Red List of Threatened Plants. IUCN, Gland; Switzerland and Cambridge, UK. Whisenant SG Repairing Damaged Wildlands: A Process-Oriented, Landscape- Scale approach. Cambridge University, Cambridge. 317p. Wiegand T, Moloney KA Rings, Circles, and Null-models for Point Pattern Analysis in Ecology. Oikos 104, Wiegand T, Moloney KA, Naves J, Knauer F Finding the Missing Link between Landscape Structure and Population Dynamics: A Spatially Explicit Perspective. The American Naturalist 154: Zamorano C, Cortés M, Echeverría C, Hechenleitner P, Lara A Experiencias de restauración con especies forestales amenazadas en Chile. In: Restauración de Bosques en América Latina. González-Espinosa M, Rey-Benayas JM, Ramírez- Marcial N. (Eds.). Mundi-Prensa México, México. pp

84 Chapter Six: References 72

85 Chapter Seven: Appendices Chapter Seven Appendices Canopy of a Pilgerodendron uviferum seed tree in a fire-disturbed upland forest. Chaiguata Lake, Chiloé Island 73