Levee Vegetation Research Symposium 2012 August 28, 2012

Transcription

1 Levee Vegetation Research Symposium 2012 August 28, 2012 (Introduction to Caroline Zanetti) Ben Carter: Dr. Caroline Zanetti, and she is joined here on the podium, in particular, to assist with the Q&A by Michel -- Vineshay? Vennetier. Ben Carter: Vennetier. So, please welcome our colleagues from France. Hello, everybody. So, before to speak to you about roots, I just present quickly [unintelligible] to be [unintelligible], as you know, becomes easier since February. So, at the National Research Institute of Science and Technology for Environment and Agriculture, it s a public institute and joins the provision of the Ministry of Research and the Ministry of Agriculture. There are neuf -- nine, sorry -- original center in the two branches on the -- in France. And, the work force in [Ilsa] is about 1, people, including and 50 scientists, and 2,040 [PIAG] and post-doctoral students. So, IRSTEA environmental research, that is widely recognized in the fields of continental surface waters, environmental technologies, and land management. You see from the picture, which are the three main topics: land, water and environmental technologies field directed right to the need of society, scientific and technology and different [unintelligible] for public policy in the form of research, science [unintelligible], models, and operational tools. And, then, an engineering approach that includes multidisciplinary components. You see in Aix-en-Provence, where we work with Michel, for cases of activity on natural results and the vulnerability of ecosystems. In my research unique, unique -- so sorry -- hydraulic and generic -- we are we have 16 patent research engineers. Four patent technician, one [product] engineer, that s me. And, eight PLG students this year. And, the main topic, first, geomechanics, erosion, and hydraulic and hydromechanics and stability, like you see on the first Levee Vegetation Research Symposium 2012 August 28, 2012 Page 1 of 16

2 picture right here on the dike. And, to stay on topic, it s performance and safety of hydraulic structure. So, on dikes and dams. And, the [petition] in back is study in this -- in this topic. So, now, I going to speak to you about development and the composition of woody root stems in French [dikes]. So, we have in front, a considerable number of embankments, dikes and dams, all of them old and rooted. More than 10,000 kilometers of root protection dikes, like you see on the picture on the left. The left, okay. And, is [unintelligible] Grenoble in France. And, more than 5,000 kilometer of canal abutment. Navigate our waterways and I draw electric plant feeder canals on the one river. A large river, in France with large lake and also, thousands of little dams, like you see on the right, with one [unintelligible] or leisure activities and fishes on this little dams are often present. And, all this structures protect human and economic stakes. So, as you know, dikes and riverbanks are favorite sites for vegetation growth. Vegetation and all rivers and canals provide many ecological and social services in terms of terrestrial and aquatic biodiversity, as Alison said. Erosion, cultural and slope stabilization, water and landscape quality, as well as leisure activities. But woody vegetation and does several programs which are not compatible with dikes safety. First, low visibility of dike slopes, attraction of burrowing animals, and risk due to root systems embankments. And, use of the picture, there is a river flood protection dikes, where it s a lot of big big trees. So, the objectives of our research -- to study tree wood system characteristics with tree topics, stem structure, root architecture, and root decomposition. Then, to assess the risk to woody vegetation on dike and dams. And, finally, to propose recommendation for helping decision makers to choose the best compromise between risk control and all of the issues. So, I present to you, three parts in the presentation. First methods, results, and recommendation. So, we began by -- we begin by extraction, results, and experiments -- are three parts of the methods. So, first, extraction of woody stems. We have excavated 350 trees of their various species. They were excavated cautiously on 11 sites. So, three steps of excavation are, first, to -- to extract shallow woody stems with hand tools, like in the picture. Then, to extract deeper roots with -- with a shovel, and when all is moving, we can slinging up the stump. Yes, three, as I said to you. So, we did nine species: poplar, willow, black locust, alder, maple, ash, oak, larch, and pine. The second steps was to measure woody stem structure and root architecture. So, manual measurements on the stumps. So, we measure plants with depth, root number, distribution, and the rotation on the root stump. And, measures on the -- of the root s architecture, we measure the middle and around direction and inclination of roots. Levee Vegetation Research Symposium 2012 August 28, 2012 Page 2 of 16

3 We have -- we have a picture presenting some measurement on the -- on the roots, and we -- we have calculated two parameters, principle [unintelligible] -- they have a decreased rate with a proximal limiter and distal limiter, and total root length. And, the branching rates, with the number of branches on a root. The last part of the methods presented here was experimentation for analyzing root decomposition. So, first steps was to collect root samples on a live stump extraction. The second was to measure samples or land [unintelligible] and wait -- fresh wait, and to be resumed at 50 centimeter depth. The next step was to measure the sample at predetermined interval of times -- every two years, in order to assess the decomposition speed and to observe changes in wood mass and the root structure. So, you see the free [unintelligible]. The sample measurement is from Sample 20 in length and the different place of diameter. To be rid of, for example, after the extraction of roots inside some -- in oaks basket. So, actually, we have not samples of different diameter [unintelligible], so, 2, 3, 5, 8, and 10 centimeters in diameter, of different species: poplar, oak, ash, locust, pine, and larch. Forty samples of four species were scanned by X-ray tomography for study of wood density. So, we going to see the results of all this topic. First, root system structure. So, stump structure, volume, and root distribution depend mainly on two parameters: first, material texture. We distinguish coarse and fine material. And, water of ability. So, we have four -- we have defined four types of root structure -- root system structure: hard, shallow, [mixed] and tap root systems. And, we have observed that the -- on front of coarse materials, we can see some different structure. And, on fine material, we observe often, outer stems. While on coarse material, we ve seen [mixed] and tap root system. Shallow rooted stems are present on the two types of materials -- on the fine and coarse materials. And, it s -- it s a depend on the water access on. On fine materials, shallow systems are at the toe of the foot, because they are limited by the water table. While, on the coarse material, shallow rooted stem are present on the top of the dike. And, because the water table is too -- too depth -- too deep to -- for the roots to grow, and so, they -- the stay very shallow to -- to kept moisture of the -- on the surface of the soil. And, the species impact very weak. So, I -- I present you some picture because it s the best mean to show you. On the locust tree, you have on the left dike with coarse materials and on the right, dikes with fine materials. And, we see, for the same species -- locust -- that, on coarse materials, we have low density, shallow, and mixed root systems. Shallow on the top, and middle slope, mixed root systems. Levee Vegetation Research Symposium 2012 August 28, 2012 Page 3 of 16

4 While, on fine materials, we have earth system -- earth root systems, with thick and dense roots. It s the same case where, on the poplar, on coarse material on the left, we have tap and mixed root systems. While on fine material, we have shallow and earth root systems. So, shallow at the dike, so -- so, we see the two parameters: material first, and the water access influence. The oak tree, here, is on two different sides, both with the same materials, fine materials or sandy materials with [unintelligible]. And, we see the same structure for these two -- for these two oak. So, the parameters influence single root system structure, are different. And, we see that the environmental context is more important than tree genetics. And, I present you this graph. It s a graph using France to determine what sort of root system structure we have on -- on dikes when we know materials, types, and water access. And, so, we can find what type of root systems we -- we may have on -- on the dike. Well, I don t speak about factors of dissymmetry. The winds of some of [unintelligible], rock or [unintelligible], the proximity of the river of the canal. We have seen that roots go in direction of the rivers -- of the waters accessible and the -- the local concentration of nutrient organic, I will show you some picture. And, the water tables -- deep of the water, but water table is another important point. It s influencing the depth of the root systems. So, now, I want to talk to you about architecture of roots. We have defined a root typology. On all the study species, we have observed this typology. So, three types of roots: vertical tap roots, running roots, and short roots. You see on the pictures the running roots and long roots and short roots on the same stump we have these two types of roots. So, these three types of roots have different morphology, lengths, and [condition] and branching frequency. Here this was an ash tree. It was on pine tree, so, running roots not more than four meters long, short roots, and tap roots. We have calculated to quantify the difference in the morphology to parameters. First, the diameter decreased right. So, we have a tap root, a running and short root in green, running roots in red, and the -- inside free also types, depending on the root diameters. So, large, medium, and small. Large, up to five centimeters and small, less than one centimeter. So, we have high diameter decrease rates on tap roots and short roots. And, the diameter -- the diameter of decreased rates decreases with root diameter. So, small roots of lower diameter decrease rates. We have the same result for branching rates. And, we think -- and, with the same graph with branching rates, that we have a high branching rate on tap roots and short roots. We see on the picture, very much -- very many branching and the branching rates increases when root diameter decrease. So, small roots have high branching rates. So, we see clearly that the running roots are very different than short and tap roots. We calculate also parameters, segment deeper, main access branches, diameter, and pipe model parameters that Michel is calculate. Levee Vegetation Research Symposium 2012 August 28, 2012 Page 4 of 16

5 So, the effects of species on architecture. So, I [unintelligible] that architectural parameters are similar on all species to date. We find there s three types of roots. The tap roots -- not always the tap roots, but short and running roots find on all species. But, some species present some very large running roots. You have the poplars, the locust, and the willow. Here is a poplar. You see very big, long roots when we extract the dike. On the locusts, you see the big, horizontal roots that go through the dikes and willow roots. So, you see the longer running roots. So, this three species: locust, poplar, and willow -- have intact the highest speed growth in diameter and length. We have just, for example, calculate for poplar growth and diameter, seven millimeter per year and with the lens, 70 centimeter per years. It s a mean. So, we could see that these tree species on French dikes are not so -- so favorable. Then, decomposition. So, we have seen that true decay depends on two parameters. Three species, and wood diameter. So, the difference is due to wood property of roots. We have said during root extraction -- stump extraction, some decaying roots, and we see that the wood inside the roots is decayed, is rotten. And, the bark is often conserved. So, we have a [unintelligible] inside the roots. And, when we extract our samples, get the samples, be rated 50 centimeters. In death, we observe some live roots crossing our decomposed roots, insects, and fungus action on the wood being decomposed. And, after two years of decomposition of these six species -- poplar, locust, oak, ash, pine, and larch -- we have the conifers difference on another site and one thing. We have so -- we have classified the -- the root decomposition level for all the species and we see that poplar and locust are the less resistant. Oak and ash, medium, and pin and larch, more resistant, but, because they are in mountain sites where biological activities is reduced in time compared to than in plain. For the density, [unintelligible] tomography resists, so you see the -- the scanner. We have obtained some slides of root. So, on every root is here, you -- you have an ash root, we can see a structure into the structures of wood so it rings -- annual rings. And, after two years of decomposition, on the same species, on ash, we can t observe the -- the tree rings and we -- we see the losing of -- of structure in the wood. We have the density colors: White is a high density, so the bark. And, the low density is -- is in gray or black. So, if we compare the four species studied in plain -- so, poplar, ash, locust, and oak -- poplar in red, ash in green, locust in blue, and oak in orange -- we see that there is a different rate of decomposition between each species. And, the poplar is a species -- of the three we have seen, visually, the less resistance. While, the locust, finally, visually, is very decomposed but in density the wood -- not lose so [unintelligible]. Levee Vegetation Research Symposium 2012 August 28, 2012 Page 5 of 16

6 So, we have classified poplar, ash, locust, and oak in the -- on level of decomposition. So, we [unintelligible] result after four years at healthy states, after two years, and after four years. And, the next steps will be to -- to extract the samples in in 14? And, we ll observe the decomposition in action. And, we consider that it will be the end of the experiment when we can t extract the root samples when all -- where it s crumbling when it can t get -- so, the next result are suspected risks. And, so, you know -- everybody know, that tree [roots] on [unintelligible] dikes may generate 50 programs and reduce [the durability]. Roots in L dikes may generate two types of [break]: externally erosion, which is related to tree uprooting, like you see on the picture on the left and, during some end roots. And, internal erosion, which is related to galleries created by rotten roots, as you see in [unintelligible]. And, also, to the mechanical action of live -- of live roots, which can decompact dikes materials or destroy maximally protection wall or joints on a rigid concrete wall of protection for example. So, the materials in France at risk for dikes, yeah -- yeah, you have a picture -- a photo -- take by Michel, where? In Arizona. In -- in Arizona. So, [unintelligible]. And, we can see that if any favorable matter yours is accessible to roots inside a dike, roots do find it and penetrate. Favorable means more water and a chance less compacted soil. So, you can see on the picture, that there is a layer -- a lake -- A layer. A layer with here all the level soil with organic materials and less compacted soil where the roots growth. And, in all the material, multi-layered, dikes on either, for examples, with patchwork of materials, root selectivity develop in the most favorable ones, including all topsoil remaining below the dike, not removed before first building it. So, the roots are under the -- the -- under the dike level. And, [unintelligible] if we are some good materials. And, then, materials in France Swiss because in [unintelligible] materials, so we ve here on the pictures silty matrix after root decomposition. We have galleries conserved and so in -- in [unintelligible] matter, it s a problem, which in grander materials, so sandy and gravel matters, gallery -- galleries or sinkholes are scrambling. The form of the roots is not conserved. Then the types and dimension of dikes and [influence also] due to vegetation. In France, we have two types of dikes: temporary water fill, and permanent water fills. So, temporary are flood protection dikes, often narrow and old, with embankment control only during floods. We can t see Levee Vegetation Research Symposium 2012 August 28, 2012 Page 6 of 16

7 some water go through some -- you see water escape from the dikes. You can see only during flood. And, the other type is a permanent water fill, so canal dikes. And, they are recent -- more recent, and large -- very large. And, with continuous control of the embankment, with, you know, [unintelligible] instruments. So, we -- we -- we -- we could observe these two dikes separately because there is not the same impact of roots. So, you see on the picture, a little dike with a larch. And, this larch take all the place in the moment, while on the runner canal dikes, so large dike -- the root systems present a few part of the dikes. So, the root architecture and the species parameters, locust, poplar, and willow, have a some big, long horizontal roots, as you see, which can penetrate the dike. So, poplar willow have also big and long tap roots which -- which can create sinkholes. Okay. You have seen this one? So, I say that s a poplar, locust, and willow, as I said, presents some long roots and poplar, willow also have big and long tap roots, which can create sinkholes in coarse materials, when -- so what the table is -- is -- in depth, is deep, they form a big tap root to -- to go access to go to the access water. So, short roots are less than the roots than for dikes and running roots, [unintelligible]. So, in some places, we have here all the parameter influencing risk for dikes. So, we see materials at first, dike dimension, and dike types, and root architecture and species on the water site. And, now, we go to the recommendations and everybody knows it s a controversial situation. Some structure architectural types and wood decomposition influence risk for dikes. It s not me. [French language]. We have to -- We said that the stem structure architectural and wood decomposition influence risk for dikes. Ideally, no trees should grow on dikes and all new trees growth was this -- when I -- ideally, no trees should grow on dikes and all new trees growth should be stopped, ideally. Short, green grass is a better solution to protect slope and control life. But issues concerning trees are always particularly sensitive and dikes are often part of the greenbelt in town, spaces for recreation, like jogging, biking, fishing, and place with high velocity. So, when we arrived to cut trees on dikes for safety reasons, habitants -- habitants living behind the levees are first to concerned by the safety, by the integrity of the -- of the dike, don t want that we get trees because they -- they want to keep trees and dikes. So, it s a controversial situation. The French general recommendations are: that the vegetation on dikes should be controlled for safety reasons, okay. By killing existing trees, the structure of dike materials change due to rotting roots. Those creating galleries are heterogeneous leading to internal [rot] erosion. So, stump extraction is [unintelligible] method for trees and dikes, but it s better than any recent program. So, if a tree is -- is cut down, the stump should be removed and dikes material recompact and dike repaired. Levee Vegetation Research Symposium 2012 August 28, 2012 Page 7 of 16

8 So, vegetation management plans, let funding to compromise at tomorrow s narrowed cost than with severe architectural solution, like you see on the picture with the construction of a jagged fragment wall to create a sill in the body of the dike for [unintelligible] in Poland. So, we think that vegetation management is a better tool. In conclusion, in [unintelligible] vegetation management on dikes could be ecologically and sociologically acceptable. We have to choose appropriate maintenance dates allowing the preservations of fauna biodiversity, and flora. And, so, the vegetation management have to be in winter. The preservation of shattered spaces with bushes, little trees of certain species -- avoid poplar, willow, and locust, you have [unintelligible]. It s possible. And, the contribution of trees on middle and isle slope of large dikes under controlled condition is possible if we respect visual observation. We limit tree eye and diameter, and change area position every 10 years, and extract big stumps. So, to conclude, in France, we have many dikes and I think [unintelligible] dikes are not so ugly landscapes. Thank you. Ben Carter: The technology reminds me of my cell phone. Every time I get to an important point, it cuts out. So -- That s a problem. Questions: Ben Carter: Thank you very much, Dr. Zanetti. Are there questions? And, could we bring up the lights so that everyone wakes up again? Yes, Mark. Mark List: Mark List with DWR. Question on material type. You said coarse versus fine. Please can you repeat slowly. Mark List: Sure. Yes. For material type, you mentioned coarse versus fine. What was your size fraction between the two? Between two types of materials? Yes, coarse and fine. Yes, yes. To simplify, we have a class of two types of matters in coarse and fine, because in France, we have often these -- often, these two types of dykes with canals dikes, which are construct with gravel and sandy materials to be draining. To allow to drain and drainage of the embankment. While, on narrow dikes of flood protection dikes, there is -- they are Levee Vegetation Research Symposium 2012 August 28, 2012 Page 8 of 16

9 [unintelligible] with taking materials inside the river and put on after each flood event. And, so, [unintelligible] is a mixed with fine and few coarse elements so the -- the -- the main part of the materials are fine. Between fine and coarse, it means that the coarse is one more than 50% of the content of the material with the gravel, stones and rocks. Mark List: And, are these large pieces of gravel or -- or are we talking about fist size or we re talking larger or smaller? More than two centimeters is coarse. Mark List: More than two centimeters is coarse. Laura Kaplan: Gravel. Michel, could you move your mike closer? Michel, closer to the -- Ben Carter: Next question? Chris Kees: I m Chris Kees from -- from ERDC. Do -- do your x-ray tomography scans of the roots -- are they precise enough to extract the microstructure of the roots at the different phases, or it mainly just kind of a bulk measurement of density? We don t want to study microstructure of roots, just woody roots. So, evidently with mechanical [shovel], we can take old roots entire and often, there is some break. So, we proceed to the cuts of first roots, shallow roots, and we point with the bump marker appurtenance of the roots, and we keep out to -- to preserve the -- the root. And, after, we got to make measurements on soil, we have cut some roots -- principal roots we want to -- to study. And, when we extract the stems, when we see that there is a root broke -- broken roots, we go to -- to -- to dig and to extract them. So, to study woody roots -- coarse roots -- it s -- and the -- and a large number of trees and of species on different condition, we -- we -- we think it s best compromise, because we don t have time to study all temp -- to study a big number of trees, like this, to excavate it very, very cautiously with one tool. Levee Vegetation Research Symposium 2012 August 28, 2012 Page 9 of 16

10 Basically, we try to extract as tuber as small as possible geometer. We then, consider, after two millimeters, you know. But, of course, extracting a wall of big root systems of 2 millimeters is impossible, particularly, on coarse material with rocks. So, we did our best. Ben Carter: Next question? Female Voice: Does France have data on failures at dikes because of trees? We have not quantified data because the problem is that, when we have trees responsible to the failure of the dam, after the failures, we have no proof -- proof -- proof. Because all is -- is good. But, we have some places where obviously the failure was due to -- due to previous trees that were cut at this places. But, we don t have studies to call numbers to -- we have numbers when they are -- they are -- they re concrete gallery, going through the dikes to capped water. Because, when there is a breach, we see -- we can see the gallery inside, so if they are, [Remy Touman] has -- you maybe know? Have some -- some number statement of this part of the responsibility of galleries inside -- a wall inside embankments. But, on tree roots, we don t have statistical numbers. Ben Carter: In the back. Donald Gray: Donald Gray. In your introductory remarks, you cited some negative impacts of vegetation. I think you said that it attracted rodents. Is this based on the study of European rodents? Rodents are -- Rodents. Burrowing animals. It s quite difficult to me to cut trees, so, it s impossible to -- to study burrows animal and to say that it would be destroyed. But, we have, in fact, some problems with burrowing animals. I -- I have -- I can -- I don t -- I haven t seen, before yesterday, with Peter and the -- the [unintelligible] models -- it s impact of burrowing mammals in dikes. There is an impressive that the dimension and the length of this burrow are very impressing. But, in France, we -- we put on the -- on dikes where -- where are burrowing animals, some [French language] -- some, you know, in metal? Some metallic nets. Levee Vegetation Research Symposium 2012 August 28, 2012 Page 10 of 16

11 Yes. Metallics -- [French language]. Donald Gray: Like, scurrying. Okay, metallic, yes. To prevent burrowing -- Animals. To dig or to prevent burrowing animals. In fact, what we observe that, as soon as the -- the undergrove where the dense cover with shrubs, burrowing animals are concentrated on these places because they like to quiet, and not to be seen. And, so, removing the -- the -- this -- this vegetation at low level is a way to prevent burrowing animals, for sure. Ben Carter: Right here. Jon Bray: Jon Bray. You mentioned that fully decomposed roots created galleries -- opened galleries -- that were conserved in cohesive, or fine grain soils. What was the rate at which you saw that? And, when you did find it, how long were these galleries? Okay. I can t give you some numbers because we have extracted 350 trees, and we have observed that these types of -- of fruit decomposition in bags. Maybe Ten times. Ten times. So, it s -- maybe just a -- on another tree, there is also proof. So, I can t give you some numbers. The last part of the question? Jon Bray: How long? How long. So, it s quite difficult because we work with the mechanical shovel. So, we cut a slide of materials, when we see this, we stopped, and we extract. And, we -- we -- we go to analyze, we -- we can break all the dike because, when we work often on canals dikes, we keep water inside the canal. So, the manager of the dikes said, Okay. There is a -- a hole, but I don t want you to go on all the site of the dike and break all the dikes to -- to see where is the hole. So, we -- we can extract from one two meter, but no -- no more. Levee Vegetation Research Symposium 2012 August 28, 2012 Page 11 of 16

12 In some cases, I can answer for the easier dike because we had the opportunity to break an old dike so to get through it completely. And, there were some galleries going through the wall dike and maybe five, six meters in length. When we see the running roots, if a root like this is decomposed, we think that it s possible that go through the under dike -- Ben Carter: Right -- right here? George Sills: Just kind of a follow-up to this lady s question. I m George Sills. I understand that once the dike has failed, you can t always do a forensic study to prove exactly what the failure is. But, if you re -- if you have a statement where you know that these trees have caused some failures, do you have documented incidents where you were called to a dike that said, Hey, we ve got a problem. You all come try to save the problem. And, the tree s the problem. Have you understood all the question? I don t have all understand. Ben Carter: Maybe Mr. Sills, if you could rephrase your question, please? More slowly and -- George Sills: Okay. What I m -- what I m wondering, have you been called -- do you have documented cases where people were fighting the flood and the they had to come out to where there was a tree. And, that tree was the problem? Yes. We have some documented cases where the decaying stumps were at the very beginning of a -- of a breach. And, also, some galleries were clearly associated even if there was not a breach, but some opening, some seepage, and some gallery inside. But, the dike was clearly associated with some stumps. It s a report from -- manager report. Yes. Documents. Ben Carter: Next question? In the back. Dave Zezulak: Dave Zezulak, Fish and Game. You showed some cross sections of some of the levees that showed levee materials that were stratified, some of which were -- showed topsoil, Levee Vegetation Research Symposium 2012 August 28, 2012 Page 12 of 16

13 you stated, laid upon by other materials to -- to build the levee. Do you have thoughts on whether the lateral roots that you saw going through levees, sought out those organic layers, or topsoil areas, to traverse the levee, as opposed to more random locations within the levee? It seems to me that, from our observations, the roots are very, very opportunistic, and we have some idea -- idea that when even a few centimeters layer of organic material -- of simply of less compacted material and rich material is enough for a root to cross the wall levee. And, we had some dikes we can cut completely and there was some patchwork of material because we are made in -- from the seventeenth century, rebuilt several times, so it s a patchwork of coarse and fine material, rich and poor. And, the roots are always were where can they be. Okay. And, we have another example, which is very clear. We ve had the compacted core of a dike and, when there was, probably, a bad compaction, and there was some very small parts of the dike that were not compacted. Just a few centimeters, but the engine probably doesn t cover the different passages. And, we had some -- some roots that were five centimeter in one direction, and 70 centimeters with in the other direction -- growing like a wall inside the dike, just following the compaction crack inside. So, where the compaction or organic matter layer, or rich layer, is enough for a root to cross a dike. Dave Zezulak: Thank you. Ben Carter: Over here. Doug Shields: You have done some great work on root architecture in levees. But, you ended up with a conclusion that tree roots are dangerous to levees. And, the best vegetative cover on -- on levees is -- you call it green grass. I m wondering, at a national scale, throughout the country of France, have you compared the performance of grass levees versus those that -- that support trees? [French language] In fact, one of the conclusion is that, generally, we should avoid trees on dikes, but we had lots of contrary examples and a lot of dikes, we recommend not to cut all the trees, all the vegetation, and just manage it in the best way -- to find the best compromise between ecology, sociology, and security. And, on large levees, we have written also, Hellen and me, who have management plans for vegetation. Levee Vegetation Research Symposium 2012 August 28, 2012 Page 13 of 16

14 While we say it s not necessary to remove vegetation and trees and -- on all parts of the levee. So, the general rule is to prevent people to plant or to keep trees while it is not necessary. It s the best way, probably. But, there are lots of example why is possible to keep tree and vegetation, provided that it is correctly managed. Because the problem is, when there is vegetation -- woody vegetation on the levee -- you have absolutely to manage it, to prevent the trees to become too big, to prevent the roots and assume to become too big and things like that. But, you can afford some vegetation in levees, provided that you manage it and you have real assessment of safety problems. And, we have seen some reports -- Doug Shields: Well, if you have so many trees on your levees in France, and trees are dangerous to levees, you must have levee failures there every year. They must fail all the time. They fail regularly. Yes. Ben Carter: One last question. Stefan Lorenzato: This is Stefan Lorenzato with the Riparian Habitat Joint Venture. I m curious about your root decay work. When you cut the samples for the root decay time trial, those were all healthy, live roots that you buried. Is that correct? Ben Carter: As opposed to? Stefan Lorenzato: Roots of a tree that had fallen over and was in partial decay when it started. So, we re looking at the decay of a healthy root to a point where it s -- it s no longer recognizable as wood, right? That s the goal of that part of your study? [French language]. Well, I mean -- you -- you -- is the question whether our samples can be compared with real roots? Stefan Lorenzato: correct? No. Whether -- it s a healthy, new, green wood that you bury in the ground, Yes. Levee Vegetation Research Symposium 2012 August 28, 2012 Page 14 of 16

15 Stefan Lorenzato: And, then, the stump examples where you ve had leakage, those were also green trees that were cut -- Yes. Stefan Lorenzato: Do you -- in your excavations, did you have anywhere you found that -- that the channels and the -- Yes. Stefan Lorenzato: - and the cohesive sediments -- were those channels in -- of -- in healthy trees, or were those channels you found from rotting, sickly trees? Okay. We have found some decay roots on all stumps up to our cut regularly. And, when we cut some trees, some roots may die and so we may have decaying roots on living trees. On experimental sites, after four years of decomposition, we -- we don t have observed some channel inside the root. It s too early, probably. Yes. But, we have extracted also around, maybe, 30 or 40 decaying stumps, you know? Not on the living trees it supports. And, in such cases, we have quite often observed galleries inside the root, close to the stump. But, after that, it s quite difficult because it s very fragile and, when we try to dig, it s very difficult to follow the hole of a decaying root inside the dike. Stefan Lorenzato: Okay. And, then -- part of what I m trying to get at is, if the tree dies naturally, is it the same decay -- Stefan Lorenzato: -- Stefan Lorenzato: Yes. - channel making process as if the tree is cut down and you have a healthy tree Yes. - and, then the root decays. Levee Vegetation Research Symposium 2012 August 28, 2012 Page 15 of 16

16 As the root decays, maybe it will be a bit longer, but not so much. As soon as the stump is dead, the root start to decay. And, maybe faster or so because there is some propagation of fungus and insects along the roots. It cannot happen in our samples, which are chopped. Stefan Lorenzato: Okay. Thank you. Ben Carter: Thank you very much, ladies and gentlemen. Caroline, Michel, thank you very much for sharing results. Okay, ladies and gentlemen, our next presenter is one of those experts that Tony Wright referred to as one that has numerous letters behind her name, and is actually pursuing more! Levee Vegetation Research Symposium 2012 August 28, 2012 Page 16 of 16