No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Vegetation unit assignments: phytosociology experts and classification programs show similar performance but low convergence
Abstract
Aims
Assigning vegetation plots to vegetation units is a key step in biodiversity management projects. Nevertheless, the process of plot assignment to types is usually non‐standardized, and assignment consistency remains poorly explored. To date, the efficiency of automatic classification programs has been assessed by comparing them with a unique expert judgment. Therefore, we investigated the consistency of five phytosociology expert judgments, and the consistency of these judgements with those of automatic classification programs.
Location
mainland France.
Methods
We used 273 vegetation plots distributed across France and covering the diversity of the temperate and mountainous forest ecosystems of Western Europe. We asked a representative panel of five French organizations with recognized expertise in phytosociology to assign each plot to vegetation units. We provided a phytosociological classification including 228 associations, 43 alliances and eight classes. The assignments were compared among experts using an agreement ratio. We then compared the assignments suggested by three automatic classification programs with the expert judgments.
Results
We observed small differences among the agreement ratios of the expert organizations; a given expert organization agreed with another one on association assignment one time in four on average, and one time in two on alliance assignment. The agreement ratios of the automatic classification programs were globally lower, but close to expert judgments.
Conclusions
The results support the current trend toward unifying the existing classifications and specifying the assignment rules by creating guiding tools, which will decrease inter‐observer variation. As compared to a pool of phytosociology experts, programs perform similarly to individual experts in vegetation unit assignment, especially at the alliance level. Although programs still need to be improved, these results pave the way for the creation of habitat time series crucial for the monitoring and conservation of biodiversity.
... To study whether reducing the number of species and time spent carrying out a vegetation plot had an impact on vegetation unit assignment, we tested the assignment of each incomplete vegetation plot to a vegetation unit. To this end, we used a typology based on the units published at the association and alliance levels (228 and 43 phytosociological units, respectively) (Bardat et al., 2004;Bioret et al., 2014;Gégout et al., 2009;Maciejewski et al., 2020). ...
... To repeat the assignments as many times as necessary, based on a previous study we chose the 'Phi-program' (Gégout & Coudun, 2012;Maciejewski et al., 2020), an automatic classification program which performs similarly as individual experts in vegetation unit assignments (Maciejewski et al., 2020). This automatic classification program uses species fidelity indexes to classify vegetation plots. ...
... To repeat the assignments as many times as necessary, based on a previous study we chose the 'Phi-program' (Gégout & Coudun, 2012;Maciejewski et al., 2020), an automatic classification program which performs similarly as individual experts in vegetation unit assignments (Maciejewski et al., 2020). This automatic classification program uses species fidelity indexes to classify vegetation plots. ...
Aims
Inventorying the habitats composing Natura 2000 sites is mandatory in the European Union and is necessary to implement relevant conservation measures. Vegetation plots, recording the presence or abundance of all plant species co‐occurring within a plot, are currently used to identify terrestrial Natura 2000 habitat types, whose descriptions are mainly based on phytosociological units. However, vegetation plots are time‐consuming and frequently restricted to the growing season. Moreover, no vegetation plots can be regarded as exhaustive, and significant inter‐observer variation has been highlighted. We studied whether reducing the number of recorded species and the time spent carrying out a vegetation plot had an impact on vegetation unit assignment using species presence. We also studied if vegetation plots recorded in winter could be used for vegetation unit assignment.
Location
Mainland France.
Methods
We used 273 vegetation plots covering French temperate and mountainous forests. The time at which species were sighted was recorded. We also estimated whether a species was recognisable in winter. We used a classification program to compare assignments based on complete and incomplete vegetation plots.
Results
Ten species and five minutes were sufficient to assign a plot to an association, and to an alliance, seven species and four minutes. Vegetation unit assignment proved feasible in winter, especially at the alliance level.
Conclusions
We confirmed that a limited number of species is sufficient to assign vegetation plots to vegetation units. However, mapping habitats requires habitat identification and delimitation. This study confirms current field habits, particularly when creating a habitat map, usually based on a limited number of recorded species. Lastly, it confirms that the use of vegetation plots coming from a great variety of sources is relevant to create habitat time series, crucial tools for monitoring habitats at a national scale.
... Cependant, les méthodes actuelles d'inventaires et de cartographies des habitats apparaissent peu performantes pour répondre à la nécessité d'un suivi régulier et harmonisé des habitats à l'échelle européenne. D'une part, de nombreux auteurs ont souligné la difficulté d'établir des identifications des habitats cohérentes et objectives (Oliver et al. 2013;Bouzillé et al. 2017;Meinard & Thébaud 2019;Maciejewski et al. 2020). D'autre part, les méthodes actuelles de cartographie, basées sur des prospections de terrains, sont jugées longues, coûteuses et donc difficiles à mettre en oeuvre sur de vastes territoires (Hearn et al. 2011;Lewis et al. 2013). ...
... La composition floristique de ces groupements est ensuite comparée aux descriptions des syntaxons contenues dans les référentiels de végétation afin de procéder à leur identification. Plusieurs auteurs ont souligné les limites de cette méthode qui laisse place à la subjectivité et peut conduire à une certaine approximation dans l'identification des relevés pouvant entraîner des divergences d'interprétation selon les opérateurs (Kočí et al. 2003;Oliver et al. 2013;Maciejewski et al. 2020). Pourtant, ce processus de rattachement est déterminant et doit être effectué avec rigueur puisqu'il conditionne la mise en place de mesures de gestion, le suivi des dynamiques de la végétation ou l'évaluation de l'état de conservation des habitats (Hearn et al. 2011;Cherrill 2016;Ullerud et al. 2018;Eriksen et al. 2019). ...
... Plusieurs études, accompagnées de relevés phytosociologiques, se sont focalisées sur ce syntaxon endémique des estuaires de la façade atlantique française (Géhu & Géhu 1978;Magnanon et al. 1998;Mesnage 2015;Lafage & Sacre 2016). L'originalité de ce syntaxon réside dans la présence d'Angelica heterocarpa, espèce endémique de ces estuaires (Dupont 1962) qui se développe sur les parties oligohalines estuariennes françaises (Lacroix et al. 2009 Le SE constitue un outil évolutif dont la mise à jour peut être effectuée en fonction de l'évolution des connaissances (Chytrý et al. 2020;Maciejewski et al. 2020 (Fig. 19). aller localement jusqu'à 1m ce qui est possible dans un volume de 1m 3 échantillonné par la mesure de quelques tiges de plantes. ...
L’inventaire et la cartographie des habitats sont des processus essentiels pour la mise en œuvre des politiques de conservation de la nature. Les méthodes actuelles, basées sur des prospections de terrain, sont difficilement applicables sur de vastes territoires et jugées inadaptées à un suivi régulier et harmonisé des habitats. L’objectif de cette thèse est d’explorer des approches innovantes afin de faciliter l’inventaire et la cartographie des habitats sur de grands sites naturels, en prenant comme cas d’étude le site Natura 2000 ‘Estuaire de la Loire’.
Un système expert a été développé pour l’identification de relevés phytosociologiques afin
d’établir la typologie des habitats du site. Cette démarche a permis de rattacher de manière formelle 1843 relevés de végétation à 89 habitats EUNIS et 17 habitats d’intérêt communautaire.
Des images satellites Sentinel-2 et des données aéroportées hyperspectrales et LiDAR ont été exploitées pour spatialiser les habitats du site par télédétection. Ces différentes données, aux caractéristiques complémentaires (résolutions spatiales, résolutions spectrales, répétitivité, 3D), ont permis de cartographier avec une très grande précision la majorité des habitats des 24 000 ha de l’estuaire de la Loire.
L’application de ces nouvelles approches démontre l’intérêt d’associer les systèmes experts et la télédétection pour typifier et cartographier des habitats de façon rentable et reproductible favorisant une gestion concertée du site Natura 2000.
... The quality of species identification in our survey is limited by expert knowledge. On the contrary, NFI data rely on inventories made by trained field operators with good botanical and phytosociological knowledge (Maciejewski et al., 2020), so that % >50, <50 is expected to be more robust to species misidentification than the survey of forest managers. The importance of Brachypodium pinnatum as a competitive species towards tree regeneration is thus probably underestimated by French forest managers. ...
... Maciejewski, L., Pinto, P.E., Wurpillot, S., Drapier, J., Cadet, S., Muller, S., Agou, P., Renaux, B.,Gégout, J.-C., 2020. Vegetation unit assignments: phytosociology experts and classification programs show similar performance but low convergence. ...
Le renouvellement forestier est une étape clef dans la vie des peuplements forestiers car il permet le maintien à long terme de la forêt, et des services écosystémiques qui lui sont associés. Parmi les facteurs susceptibles d’impacter négativement le renouvellement, la colonisation par des espèces végétales concurrentes peut ralentir la régénération ligneuse, voire la bloquer pendant plusieurs décennies.L’objectif de cette thèse est d’estimer l’impact, à une échelle régionale à nationale, de la végétation concurrente sur la régénération ligneuse. La thèse est principalement basée sur l’utilisation des données de l’inventaire forestier national français (IFN).Une première étape, utilisant les résultats d’une enquête auprès de gestionnaires forestiers, a permis de dresser une liste des principales espèces concurrentes pour la régénération ligneuse, et de déterminer les surfaces sur lesquelles chaque espèce est présente avec une forte abondance. La ronce (Rubus fruticosus), la fougère aigle (Pteridium aquilinum) et la molinie bleue (Molinia caerulea) sont les principales espèces concurrentes en France, et sont chacune présentes à forte abondance dans plus de 300 000 ha des forêts à faible couvert de canopée en France.Afin de mieux comprendre l’écologie des espèces concurrentes, le rôle de l’ouverture de la canopée sur la probabilité de présence et de forte abondance des espèces concurrentes a été étudié. Pour une majorité d’espèces, le couvert de la canopée a un effet faible sur la probabilité de présence des espèces. À l’inverse, l’abondance des espèces est fortement corrélée au couvert de la canopée, un couvert important étant pour la majorité des espèces concurrentes associé à une plus faible abondance.L’effet de la végétation concurrente sur le recouvrement de la régénération ligneuse a ensuite été modélisé à l’échelle de la France pour les trois principales espèces. Une forte abondance de P. aquilinum et M. caerulea est associée à une diminution relative du recouvrement de la régénération ligneuse d’environ 30 % et 40 % respectivement, par rapport à des situations où ces espèces sont présentes à faible abondance. À l’inverse, R. fruticosus a un effet plus ambivalent sur la régénération ligneuse. Le recouvrement de régénération ligneuse augmente en moyenne légèrement pour des abondances de R. fruticosus intermédiaires, et diminue sensiblement pour une forte abondance de cette espèce.Enfin, la probabilité de présence et de forte abondance des trois principales espèces concurrentes a été modélisée, afin d’identifier les conditions de fortes abondances et de prédire la localisation des surfaces potentiellement impactées par l’apparition de ces trois espèces lors de l’ouverture de la canopée. La difficulté de modéliser précisément la présence mais surtout le niveau d’abondance de ces espèces a conduit à des modèles de faibles capacités prédictives, ne permettant pas d’obtenir des estimations robustes des conditions à fort risque d’envahissement par les trois espèces étudiées.En conclusion, cette thèse a montré que les problèmes de régénération ligneuse liés à des espèces de végétation concurrente sont présents sur des surfaces notables dans les forêts françaises. P. aquilinum et M. caerulea ont globalement un effet négatif marqué sur la régénération ligneuse quelle que soit leur abondance, tandis que l’effet de R. fruticosus n’est globalement négatif que pour de fortes abondances. Enfin, cette thèse a également démontré l’importance de modéliser séparément la présence et l’abondance, et identifié les principaux facteurs à mieux prendre en compte pour mieux modéliser l’abondance des espèces.
... However, despite the information provided by the literature, identifying vegetation plots is prone to subjectivity, which can lead to a certain approximation in data identification (Oliver et al. 2013). It may lead different operators to obtain divergent interpretations (Kočí et al. 2003;Maciejewski et al. 2020). However, assigning a vegetation plot to a syntaxon is crucial and must be done accurately since it influences the future implementation of management measures, vegetation dynamics monitoring, or for assessing the conservation status of habitat types (Hearn et al. 2011;Cherrill 2016;Ullerud et al. 2018;Eriksen et al. 2019). ...
... Our method is based on the use of data from national and European reference systems, allowing us to transpose this approach to other sites and any other vegetation types. The ES is a standardized and adaptable tool that can be updated according to knowledge evolves (Maciejewski et al. 2020;Chytrý et al. 2020). It can also be used to develop a formal protocol for vegetation and habitat identifications at a national scale facilitating data sharing and use . ...
Vegetation is a good indicator that can help better manage and conserve nature. It is also essential for characterizing habitats that represent an essential component of European nature conservation policy, especially within the Natura 2000 network. However, identifying plant communities is a complex operation partly because of the lack of available tools to identify them accurately. This obstacle is particularly noticeable when we talk about diverse plant communities like meadows. This study aims to develop an expert system to apply formalized classifications of Atlantic estuarine wet meadow community types in the Natura 2000 site ‘Estuaire de la Loire’. The tool we created automatically assigns vegetation plots to the units of the French vegetation typology. It allows us to ensure the classification of the European habitat types (EUNIS and the Annex I of the EU Habitats Directive) with 91% accuracy. This expert system was applied to a dataset of 1898 vegetation plots from the study area. It allowed us to link 718 vegetation plots to 4 habitats of wet meadows including the habitat of community interest 1410 ‘Mediterranean salt meadows (Juncetalia maritimi)’. Using this approach, we have also defined the characteristic species of these habitats at a local scale. This tool enables the fast, objective and replicable identifications of wet meadows which are necessary to map or monitor the habitat type (sensu Habitat Directive). The method applied in this study can be easily adapted in other sites and for other habitat types.
... Ainsi, la végétation permet de définir un habitat puisqu'elle est dépendante des conditions stationnelles. Cette définition permet également de reconnaître le rôle de la phytosociologie dans la caractérisation des habitats terrestres(Maciejewski et al. 2016 ;Maciejewski et al. 2020). La phytosociologie est la science des humaines, etc.). ...
... The placement of the British variation within a broader European framework was investigated using floristic data from published studies by workers in continental Europe, where vegetation unit assignment is still ongoing (e.g. Maciejewski et al., 2020). ...
Nature conservation requires classification of vegetation types for site assessment and assignment. Species-rich floodplain meadows are a declining habitat in Britain and Europe yet their classification in Britain has been based on just a few samples and fails to describe community response to environmental change adequately. European classification, in opposite, has been based on samples from the wide geographical range with no environmental data/analysis supporting the choices. We propose a revised classification of the lowland meadow Alopecurus pratensis-Sanguisorba officinalis community of the British National Vegetation Classification (NVC) linked to variation in local water-table depth. Data have been collated from 58 British floodplain meadows. Based on botanical and hydrological data, four subcommunities within the Alopecurus-Sanguisorba community have been defined. Assessment of conservation sites at the subcommunity level allows temporal and spatial evaluation of the trends and suggests hydrological management towards desirable vegetation. This approach, developed on data from the British meadows, has much wider geographical applications if compared with European plant communities. Seventy-two British and European plant associations were compared via Canonical Correspondence Analysis (CCA). Species ordinations were used to study the coherence of floodplain syntaxonomic alliances across Europe from Ireland to Bulgaria. CCA confirmed the spread of the British subcommunities of the Alopecurus-Sanguisorba community along a strong hydrological gradient and highlighted their lower fertility compared to their Dutch counterparts. The hydrological gradient separating the British subcommunities should help inform site management for the conservation of the species-rich communities, especially where hydrological control is possible.
... At European level, rule-based expert classification system of floristic and ecological surveys has already been conducted to classify NFI and International Co-operative Programme on Assessment and Monitoring of Air Pollution Effects on Forests (ICP) plots (grid 16 × 16 km throughout Europe) into the 14 European Forest Types (Giannetti et al. 2018). At a more precise level of vegetation type such as phytosociological associations and alliances level, automatic classification based on a floristic typicality index (Gégout and Coudun 2012) has also been produced for French forest ecosystems (Maciejewski et al. 2020). The result of automatic classification was compared to ex-situ classification by different ecologists on plots located throughout France. ...
The monitoring of habitats at plant association level, has been developed by the French-National Forest Inventory (NFI) progressively since 2011, whereas ecological and floristic data exist since the mid-1980s. The NFI habitat monitoring is the French tool of surveillance of forest habitats decreed by Natura 2000 Directive (article 11). Determination of plant association in NFI plots concerns all the habitats, whether they are of community interest or not. The objective of this study is to compare different methods of automatic classification of floristic and ecological surveys into forest habitat groups. Indeed, enriching the old surveys, which contain only ecological, floristic and trees data, with information on habitats would increase the accuracy of the calculated statistical results on habitats. The uncertainty of the attribution of a habitat outside the field (ex-situ) by experts was quantified by comparison with the determination in the field (in situ). This result was used as a benchmark to compare to the error rates obtained by two methods of automatic classification: an algorithm inspired by the habitat identification key used in the field and Random forest, a learning classification method. The classification performance was evaluated for three levels of habitat groupings. The results showed that the lower the level of clustering, the higher the error rate. Depending on the classification method used and the level of aggregation, the error rates varied between 5 and 15%. In all cases, the error rates were below the estimated uncertainty of the expert attribution of ex-situ habitat.
... On the contrary, NFI data rely on inventories made by trained field operators with Page 14 of 17 Dumas et al. Annals of Forest Science (2022) 79:41 good botanical and phytosociological knowledge (Maciejewski et al. 2020), so that %A PC > 50, CC < 50 is expected to be more robust to species misidentification than the survey of forest managers. The importance of Brachypodium pinnatum as a competitive species towards tree regeneration is thus probably underestimated by French forest managers. ...
Key message
Fifteen species are most susceptible to require vegetation control during tree regeneration in the range of our study. Among these 15 species, Rubus fruticosus , Pteridium aquilinum , and Molinia caerulea cover each more than 300,000 ha of open-canopy forests.
Context
Vegetation control, i.e., the reduction of competitive species cover, is often required to promote tree seedling establishment during the forest regeneration stage. The necessity to control understory vegetation largely depends on the species to be controlled. In order to plan forest renewal operations, it is critical to identify which species require vegetation control during the regeneration stage and to quantify the forest area affected by these species.
Aims
We aimed at identifying the main species requiring vegetation control and at estimating the forest area they cover at the national level.
Methods
Using National Forest Inventory data, we created four indicators based on two levels of plant cover, cross-referenced with two levels of canopy opening, and compared them to the outcome of a survey of forest manager practices.
Results
The best indicator was the one that represented the proportion of forests with open canopy where the species was present with a large cover in the understory. In non-Mediterranean France, according to the indicator, a total of 15 species were found to frequently require vegetation control during the tree regeneration stage. Pteridium aquilinum , Molinia caerulea , and Rubus fruticosus were the main species, and each covered more than 300,000 ha of forest with open canopies, representing about 13% of the total forest area with open canopies outside of the Mediterranean area.
Conclusions
Forests covered by species requiring vegetation control according to forest managers represent a large share of the forest area undergoing regeneration. This study provides the first list of species that require vegetation control based on a methodological protocol that makes it possible to calculate the area associated with each species.
... Le fait que l'évaluation de l'état de conservation des habitats soit effectuée par différents experts implique la production d'une méthode standardisée si l'on veut diminuer le risque d'interpréter la notion d'état de conservation de différentes manières (Bottin et al., 2005). des habitats terrestres Maciejewski et al., 2020). La phytosociologie est la science des groupements végétaux, c'est-à-dire des syntaxons (Meddour, 2011). ...
Publication - ÉVALUATION DE L’ÉTAT DE CONSERVATION DES BAS-MARAIS CALCAIRES – VERSION FINALE.
Les grilles d’évaluation proposées pour évaluer l’état de conservation des bas-marais calcaires d’intérêt communautaire présents sur le territoire métropolitain ont été finalisées.
http://www.patrinat.fr/fr/actualites/evaluation-de-letat-de-conservation-des-bas-marais-calcaires-version-finale-6974
... Le fait que l'évaluation de l'état de conservation des habitats soit effectuée par différents experts implique la production d'une méthode standardisée si l'on veut diminuer le risque d'interpréter la notion d'état de conservation de différentes manières (Bottin et al., 2005). des habitats terrestres Maciejewski et al., 2020). La phytosociologie est la science des groupements végétaux, c'est-à-dire des syntaxons (Meddour, 2011). ...
Évaluation de l’état de conservation des bas-marais calcaires d’intérêt communautaire
... Following this rationale, Tichý (2005) developed four similarity indices using the phi coefficient as a fidelity measure. Among them, the frequency-positive fidelity index (FPFI) has been often used for the assignment of relevés that were misclassified or classified to more than one groups in supervised classifications (Boublík et al. 2007;Douda 2008;Boublík 2010;Janišová et al. 2010;Svitková and Šibík 2013;Landucci et al. 2013;Rodríguez-Rojo et al. 2014;Chytry and Tichy 2018;Maciejewski et al. 2020). Another approach employing the diagnostic species concept in a similarity index was proposed by Dai et al. (2006), who suggested using the total indicator value index (TIVI) to test the validity of a TWINSPAN classification and to refine the initial classification by reassigning relevés. ...
• Key message
Total fidelity value index can be used for the assignment of new relevés to existing vegetation units and it can be used to refine classifications derived from unsupervised clustering.
• Context
Diagnostic species is an important concept in vegetation classification. Apart from its usefulness to characterize species niche preferences, the diagnostic species concept is used in vegetation classification: (1) for the assignment of new relevés to the vegetation units of an existing classification; (2) to refine vegetation classifications by reassigning relevés that sustain the definition of vegetation units.• Aims
The main aims were to evaluate the relative predictive performance of different statistical fidelity measures for the reassignment of relevés to existing vegetation units, and in which cases reassignments improve the quality of the original classification.• Methods
We took the classifications produced by three commonly used unsupervised classification methods, and all relevés were reassigned to the closest vegetation unit according to the total fidelity value index (TFVI), where fidelity value had been calculated using one of eight distinct statistical measures, and according to the frequency-positive fidelity index (FPFI). Classifications obtained after relevé reassignments were compared to the initial ones using the Adjusted Rand Index. The quality of all classification solutions, including the initial ones, was evaluated using thirteen different evaluator statistics.• Results
The predictive performance of IndVal was the best among all eight fidelity indices in the TFVI framework, and also outperformed FPFI. The TFVI framework based on group-equalized fidelity indices produced better results than other assignment rules in terms of the chosen evaluator statistics. Re-assignments based on IndVal, r, or FPFI produced classifications with the best quality, when combining the results of all evaluators.• Conclusion
We conclude that TFVI based on IndVal and r has the best quality for assigning of new relevés to existing vegetation units, and it also could be used to refine classifications derived from unsupervised clustering. Consequently, our results reiterate that TFVI, which is new in vegetation sciences, can be a good alternative for FPFI, as the most commonly used in the assignment of vegetation plots (relevés), to predefined vegetation types in large datasets.
En 1992 en Europe, grâce à la Directive Habitats-Faune-Flore, les habitats naturels sont devenus des objets à conserver au même titre que les espèces, élargissant ainsi le domaine d’actions des politiques publiques à un autre niveau d’organisation de la biodiversité. Mais la reconnaissance tardive de leur valeur de conservation, ainsi que des lacunes dans leurs définitions sont en partie responsables de l’absence de séries temporelles de données sur les habitats à l’échelle nationale. Cela limite notre capacité à surveiller et évaluer leur état de conservation, et à adapter les actions de conservation aux niveaux national et local. Les objectifs de cette thèse sont d’abord d’explorer des approches rapides et formalisées de reconnaissance des habitats forestiers afin de pouvoir ensuite étudier leur dynamique récente au regard de deux grands changements survenus au cours des dernières décennies : le réchauffement climatique et la création du réseau Natura 2000.Nous avons d’abord étudié les incertitudes liées à la reconnaissance des habitats forestiers lors du rattachement d’un relevé floristique à un type d’habitat en comparant cinq experts et trois programmes automatiques de classement. Nous avons mis en évidence la forte variabilité de classement entre experts, et l’efficacité des programmes automatiques qui est comparable à celle des experts. Nous avons également montré que pour la reconnaissance des habitats forestiers, un nombre limité d’espèces est suffisant, et qu’il est possible d’utiliser des relevés réalisés en hiver. Ainsi, nous avons pu créer des séries temporelles de données standardisées sur les habitats forestiers à partir de différentes sources d’inventaires floristiques, rattachés ou non à un type d’habitat.Dans un second temps, la création de 5701 couples de relevés floristiques historiques (avant 1987) et récents (après 1997) a permis de mettre en évidence, en montagne, un changement de 11% des couples vers des habitats forestiers caractéristiques de conditions climatiques plus chaudes. L’augmentation de la dominance de ces habitats nous permet de conclure à une thermophilisation des habitats forestiers en montagne. Cependant, aucun changement significatif n’a été observé en plaine, ce qui conduit à un décalage important entre les exigences thermiques des communautés végétales et les températures actuelles : une dette climatique se développe. Face à des impacts différenciés, nous concluons que les politiques publiques pourraient être mises en place et priorisées de façon différente en montagne et en plaine pour être plus efficaces.Enfin, en étudiant 155 sites Natura 2000 français répartis sur tout le territoire métropolitain tempéré et montagnard, nous avons montré que, depuis la mise en place du réseau, l’augmentation de la quantité des très gros bois sur les zones où ils sont présents est significativement plus forte à l’intérieur du réseau Natura 2000 qu’à l’extérieur. Ainsi, nous avons mis en évidence que les actions de conservation mises en place dans les forêts au sein du réseau Natura 2000, qui sont gérées et exploitées, ont déjà eu des effets positifs sur les très gros bois, considérés comme une caractéristique de vieilles forêts, et utilisés aussi comme indicateur de biodiversité et du bon état de conservation des habitats forestiers.Ce travail de thèse était nécessaire pour compléter les nombreuses études déjà disponibles à l’échelle des espèces et des communautés végétales, car pour être efficace il est indispensable de travailler à la conservation de tous les niveaux d’organisation de la biodiversité simultanément. Connaitre les domaines de validité des moyens de reconnaissance des habitats forestiers, mais aussi comprendre leur dynamique récente et les facteurs qui l’influencent permettent de fournir des éléments pour mettre en place un suivi des habitats forestiers et adapter les politiques publiques et les actions de gestion afin d’en améliorer l’efficacité.
Due to a production error this article was inadvertently released early. It has been temporarily removed until the embargo is lifted.
Questions: Substantial variation between observers has been found when comparing parallel land-cover maps, but how can we know which map is better? What magnitude of error and inter-observer variation is expected when assigning land-cover types and is this affected by the hierarchical level of the type system, observer characteristics, and ecosystem properties?
Study area: Hvaler, south-east Norway.
Methods: Eleven observers assigned mapping units to 120 stratified random points. At each observation point, the observers first assigned a mapping unit to the point independently. The group then decided on a ‘true’ reference mapping unit for that point. The reference was used to estimate total error. ‘Ecological distance’ to the reference was calculated to grade the errors.
Results: Individual observers frequently assigned different mapping units to the same point. Deviating assignments were often ecologically close to the reference. Total error, as percentage of assignments that deviated from the reference, was 35.0% and 16.4% for low and high hierarchical levels of the land-covertype system, respectively. The corresponding figures for inter-observer variation were 42.8% and 19.4%, respectively. Observer bias was found. Particularly high error rates were found for land-cover types characterised by human disturbance.
Conclusions: Access to a ‘true’ mapping unit for each observation point enabled estimation of error in addition to the inter-observer variation typically estimated by the standard pairwise comparisons method for maps and observers. Three major sources of error in the assignment of land-cover types were observed: dependence on system complexity represented by the hierarchical level of the land-cover-type system, dependence on the experience and personal characteristics of the observers, and dependence on properties of the mapped ecosystem. The results support the necessity of focusing on quality in land-cover mapping, among commissioners, practitioners and other end users.
A historical overview of the phytosociological method in Europe is presented. Some methodological and procedural differences in the application of the Braun-Blanquet approach, from the selection of the sampling plots to the assignment of relevés to existing or newly described units, are briefly compared. The main advantages and limitations of the phytosociological vegetation classification are reviewed and discussed, also in light of their applications for vegetation mapping and monitoring.
Questions
Land‐cover maps are used for nature management, but can they be trusted? This study addresses three questions: (1) what is the magnitude of between field worker inconsistencies in land‐cover maps and what may cause such inconsistencies; (2) in which ways and to what extent do spatial scale and mapping system influence inconsistencies between maps; and (3) are some biomes mapped more consistently than others, and if so, why?
Location
Gravfjellet, Øystre Slidre municipality, southern Norway.
Methods
Two different mapping systems, designed for mapping at different spatial scales, were used for parallel mapping by three different field workers, giving a total of six maps for the study area. Spatial consistency of the resulting maps was compared at two hierarchical levels for both systems.
Results
The average pair‐wise spatial consistency at the highest hierarchical level was 83% for both systems, while the average pair‐wise spatial consistency at the lowest hierarchical level was 60.3% for the coarse system and 43.8% for the detailed system. Inconsistencies between maps were partly caused by the use of different land‐cover units and partly by spatial displacement.
Conclusions
Field workers made different maps despite using the same mapping systems, materials and methods. The differences were larger at lower hierarchical levels in the mapping systems and increased strongly with system complexity. Consistency among field workers should be estimated as a standard quality indicator in all field‐based mapping programmes.
While classification of vegetation can be conducted in many ways, international homogenization of procedures and typologies is desirable for human societies that are highly connected in terms of sharing biodiversity information. This Special Issue of Phytocoenologia includes 12 papers that document several of the plot-based classification approaches currently used throughout the world. The issue includes approaches from five continents, but noticeable gaps are South America, middle-eastern countries, northern Africa and southeastern Asia. We include in this editorial a brief synthesis of the papers included in the Special Issue, with respect to (1) the amount of vegetation-plot data and characteristics of the classification systems developed in different areas and (2) the concepts and procedures of classification approaches. One of the most important common attributes among the classification approaches is the need to define vegetation units at a low level of abstraction. ‘Association’ (and perhaps ‘alliance’ too) may be a classification level for which international homogenization of procedures would be most easy to achieve, perhaps establishing different consistent classification sections depending on ecological conditions. Several papers in this issue demonstrate that multiple approaches may coexist for higher levels, as long as they abstract vegetation from the same low level units by focusing on a specific set of concepts and defined from the perspective of applications.
Aims
Phytosociological classification of fen vegetation ( Scheuchzerio palustris‐Caricetea fuscae class) differs among European countries. Here we propose a unified vegetation classification of European fens at the alliance level, provide unequivocal assignment rules for individual vegetation plots, identify diagnostic species of fen alliances, and map their distribution.
Location
Europe, western Siberia and SE Greenland.
Methods
29 049 vegetation‐plot records of fens were selected from databases using a list of specialist fen species. Formal definitions of alliances were created using the presence, absence and abundance of Cocktail‐based species groups and indicator species. DCA visualized the similarities among the alliances in an ordination space. The ISOPAM classification algorithm was applied to regional subsets with homogeneous plot size to check whether the classification based on formal definitions matches the results of unsupervised classifications.
Results
The following alliances were defined: Caricion viridulo‐trinervis (sub‐halophytic Atlantic dune‐slack fens), Caricion davallianae (temperate calcareous fens), Caricion atrofusco‐saxatilis (arcto‐alpine calcareous fens), Stygio‐Caricion limosae (boreal topogenic brown‐moss fens), Sphagno warnstorfii‐Tomentypnion nitentis ( Sphagnum‐ brown‐moss rich fens), Saxifrago‐Tomentypnion (continental to boreo‐continental nitrogen‐limited brown‐moss rich fens), Narthecion scardici (alpine fens with Balkan endemics), Caricion stantis (arctic brown‐moss rich fens), Anagallido tenellae‐Juncion bulbosi (Ibero‐Atlantic moderately rich fens), Drepanocladion exannulati (arcto‐boreal‐alpine non‐calcareous fens), Caricion fuscae (temperate moderately rich fens), Sphagno‐Caricion canescentis (poor fens) and Scheuchzerion palustris (dystrophic hollows). The main variation in the species composition of European fens reflected site chemistry (pH, mineral richness) and sorted the plots from calcareous and extremely rich fens, through rich and moderately rich fens, to poor fens and dystrophic hollows. ISOPAM classified regional subsets according to this gradient, supporting the ecological meaningfulness of this classification concept on both the regional and continental scale. Geographic/macroclimatic variation was reflected in the second most important gradient.
Conclusions
The pan‐European classification of fen vegetation was proposed and supported by the data for the first time. Formal definitions developed here allow consistent and unequivocal assignment of individual vegetation plots to fen alliances at the continental scale.
Aims: Syntaxonomic classification is widely used for vegetation survey in Europe. The long history of its use has produced many concepts and names of vegetation units that need to be revised and integrated into a single classification system. Here we (1) present a new, global hierarchical syntaxonomic systems of alliances, orders, classes of the Braun-Blanquet syntaxonomy for vascular plant, bryophyte, lichen and algal communities; (2) characterize all accepted syntaxonomic concepts in ecological and geographical terms, (3) link all available synonyms to these concepts, and (4) provide a list of diagnostic species for all classes of European vegetation.
Location: Europe, Greenland, Arctic archipelagos, Canary Islands, Madeira, Azores and Cyprus.
Methods: We reviewed about 10 000 bibliographic sources to compile the syntaxonomic systems of classes, orders and alliances, and species lists characterizing all classes. All known syntaxonomic concepts were critically evaluated by experts and their names revised according to the International Code of Phytosociological Nomenclature.
Results: The newly compiled EuroVegChecklist is a syntaxonomic conspectus consists of systems of classes, orders, and alliances for the communities dominated by vascular plants (EVC1), bryophytes and lichens (EVC2), and algae (EVC3). EVC1 comprises 110 classes, 300 orders and 1088 alliances (with 4067 synonyms for all ranks); EVC2 27 classes, 53 orders and 137 alliances (with 410 synonyms for all ranks); EVC3 13 classes, 24 orders and 53 alliances (with 188 synonyms for all ranks). 13 289 diagnostic taxa were assigned to classes of EVC1, 2099 to classes of EVC2 and 346 to classes of EVC3. Information on each accepted syntaxonomic concept was made accessible through the software tool EuroVegBrowser. An expert system for an automatic identification of class membership based on the proportion of character species was also developed.
Conclusions: The Conspectus is the first comprehensive and critical account of syntaxa synthesizing more than 100 years of classification effort of European phytosociologists. It aims at stabilizing the nomenclature of the syntaxa and of classification concepts for practical uses such as calibration of habitat classification used by the European Union, standardization of terminology for environmental assessment studies, management and conservation of nature areas, landscape planning and education.
Aims Vegetation sampling employing observers is prone to both inter-observer and intra-observer error. Three types of errors are
common: (i) overlooking error (i.e. not observing species actually present), (ii) misidentification error (i.e. not correctly
identifying species) and (iii) estimation error (i.e. not accurately estimating abundance). I conducted a literature review
of 59 articles that provided quantitative estimates or statistical inferences regarding observer error in vegetation studies.
Many current approaches to the formal definition of vegetation units have the disadvantage of leaving a large amount of relevés unclassified. In this paper I propose a new method for the unambiguous assignment of relevés, which is based on the summarised cover value of diagnostic species. In the first step, a relevé is assigned to the class with the highest cover score. For this purpose, the character and differential species of the class as well as the character species of all subordinated syntaxa are considered diagnostic. Once the class has been determined, the assignment proceeds successively to the lower ranks. This method, which may be called "summarised percentage cover approach", uses solely the diagnostic species of syntaxa for the assignment of relevés to vegetation units, making additional formal definitions unnecessary. As a test data set, I used 487 relevés of fringe vegetation and grasslands of nutrient-poor sites sampled in the Vienna Woods (Wienerwald), Austria. All relevés were provisorily classified at alliances level, mostly following the assignment of the original authors. TWINSPAN and DCA were applied to evaluate this preliminary classification. Diagnostic species for both the alliance and the class level were identified using the total cover value ratio as fidelity measure. The subjective classification was largely confirmed by the numerical methods. On basis of the summarised cover of Trifolio-Geranietea and Festuco-Brometea species, between 64% (Geranion sanguinei) and 99% (Seslerio-Festucion pallentis) of the relevés were reassigned to the same class as in the original classification. The fact that a considerable amount of relevés originally classified as Geranion sanguinei was reassigned to grasslands reflects the problems in delimiting fringe communities rather than a poor performance of the assignment method. The "summarised percentage cover approach" allows not only for the unequivocal assignment of relevés to an existing classification system, but may also help to improve classifications by clarifying the delimitation of higher syntaxa.
Questions: Are artificial neural networks useful for the automatic assignment of species composition records from vegetation plots to a priori established classes (vegetation units)? Is the assignment more accurate (1) if the classes are defined by numerical classification rather than by expert-based classification; (2) if the training data set is selected to include plots that are richer in diagnostic species of particular classes? Material: Species composition records (relevés) from 4186 plots of Czech grasslands. Methods: Plots were classified into 11 phytosociological alliances (expert classification) and into 11 clusters derived from numerical cluster analysis. Some plots were used for training the classifiers, which were the multi-layer perceptrons (MLP; a type of artificial neural network). Other plots were used for testing the performance of these classifiers. Plots used for training were selected (1) randomly; (2) according to higher representation of diagnostic species of particular classes. Results: Different MLP classifiers correctly classified 77-83% of plots to the classes of the expert classification and 70-78% to the classes of the numerical classification. The better result in the former case was mainly due to two classes in the expert classification, which were well recognized by the classifiers and at the same time contained a large proportion of the plots of the entire data set. Correct classification of the plots belonging to these large classes resulted in a good overall performance of the classifiers. After training with randomly chosen plots, the classifiers produced better results than after training with plots that contained more diagnostic species. This indicates that the biased selection of the training plots disables the classifiers to recognize the entire variation within the classes and results in errors when new plots are to be classified. Conclusions: MLP is suitable for assigning vegetation plots to already established classes. Unlike some other methods of supervised classification, it performs well even in communities that are poor in diagnostic species. However, the method does not provide clear assignment keys that could be used for class identification in field surveys. It is therefore more appropriate in applications that aim at a reliable class assignment rather than understanding the assignment rules.
Collection Patrimoines Naturels
Description of the subject. A field study has been conducted on 24 grasslands with five different botanical experts in order to assess inter-observer bias when making botanical surveys as well as the possible consequences in terms of descripting a semi-natural habitat.
Objectives. Fieldwork has been conducted to understand the most important factors of variability affecting botanical surveys conducted by several observers. These results were used to suggest practical solutions to enhance the quality of such surveys.
Method. Five observers performed a complete botanical survey of 24 grassland plots in the Famenne (Wallonia, Belgium) in June 2009. All surveys were statistically analyzed in order to detect and quantify the sources of variability between observers. The main parameters compared are the habitat diagnosis made on the field by the experts, the rate of detection of the characteristic species as well as their coverage in each plot.
Results. Regarding habitat identification, the biggest differences between observers are seen in plots where the composition is intermediate between a habitat in good and in bad status. Overall, there was a slight tendency to undervalue the quality of the habitat. The analysis revealed that the primary cause of variability between observers is the fact that the experts did not always strictly follow the criteria for habitat identification. As regards the comparison between observers, several sources of variability were identified. The main ones are the variability of the estimated coverage of some plants, the variability of the detection rate of characteristic species, as well as the variability of the prospecting effort that can be sub-optimal in each plot.
Conclusions. Some of the sources of variability that have been pointed out can be resolved easily, other have to be taken in consideration when comparing the results of surveys in the future. The solutions proposed to reduce the variability between observers are to encourage better self-control of the parameters to be taken into account at each step of the work, the organization of targeted training courses and more standardized prospecting efforts.
Keywords. Grassland, detection rate, cover rate, observer effect, bias, prospection, monitoring, habitat, identification.
This paper presents a new and simple method to find indicator species and species assemblages characterizing groups of sites. The novelty of our approach lies in the way we combine a species relative abundance with its relative frequency of occurrence in the various groups of sites. This index is maximum when all individuals of a species are found in a single group of sites and when the species occurs in all sites of that group; it is a symmetric indicator. The statistical significance of the species indicator values is evaluated using a randomization procedure. Contrary to TWINSPAN, our indicator index for a given species is independent of the other species relative abundances, and there is no need to use pseudospecies. The new method identifies indicator species for typologies of species releves obtained by any hierarchical or nonhierarchical classification procedure; its use is independent of the classification method. Because indicator species give ecological meaning to groups of sites, this method provides criteria to compare typologies, to identify where to stop dividing clusters into subsets, and to point out the main levels in a hierarchical classification of sites. Species can be grouped on the basis of their indicator values for each clustering level, the heterogeneous nature of species assemblages observed in any one site being well preserved. Such assemblages are usually a mixture of eurytopic (higher level) and stenotopic species (characteristic of lower level clusters). The species assemblage approach demonstrates the importance of the 'sampled patch size,' i.e., the diversity of sampled ecological combinations, when we compare the frequencies of core and Satellite species. A new way to present species-site tables, accounting for the hierarchical relationships among species, is proposed. A large data set of carabid beetle distributions in open habitats of Belgium is used as a case study to illustrate the new method.
The Prodrome of French vegetation is presented from its beginning in 1996 to the publication in 2004 of the first version of the national synsystem detailed up to the level of suballiance (PVF1). Work began in 2006 to produce a second edition, called PVF2, which aims to describe 78 of the 80 classes recorded in mainland France and Corsica, up to the level of association and subassociation. So far, 19 classes have been published, five classes are ready for publication and 54 classes are under preparation. The most important classes of PVF2, especially forest classes, should be completed in 2015.
Assignment of large numbers of vegetation plots to a priori vegetation classifications is increasingly being required to support natural resource management, monitoring and conservation at regional scales. Several automated systems have been developed that use quantitative synoptic tables and algorithm‐based plot‐to‐type assignment. However, where synoptic tables do not exist, and qualitative species lists characterise vegetation type classifications, existing systems may not apply. In these situations, vegetation experts may resort to manual assignment processes that can be slow, subjective and fraught with difficulties.
This study combines repeatable and objective quantitative analyses, with new software, to deliver a semi‐automated plot‐to‐type assignment process appropriate for a priori classifications based on qualitative species lists. The flexible semi‐automated assignment program ( SAAP ) calculates a quantitative goodness‐of‐fit score between plots and types, based on the species that characterise each a priori vegetation type, and the species that characterise groups of plots derived from quantitative analyses.
We applied the SAAP to a case‐study of 630 native vascular plant species from 930 plots, and an a priori classification of 99 vegetation types. We varied vegetation data set transforms [cover per cent (0–100%), cover score (0–6) and presence–absence (1, 0)] and analysis settings and tested the degree to which the SAAP provided plot‐to‐type assignment concordant with manual expert assignment.
Results provided clear evidence supporting the choice of particular data set transformations and analysis settings to maximise concordance. The SAAP allocated up to 50% of plots to the same expert‐assigned vegetation type, and more than 70% of plots to an expert‐assigned vegetation type ranked in the top five by the SAAP .
When coupled with repeatable and objective quantitative analyses, the SAAP provides vegetation experts with a new semi‐automated and quantitative decision support tool to assist with the assignment of vegetation plots within a priori vegetation classifications defined by characteristic species lists.
Delimitation of vegetation units in phytosociology is traditionally based on expert knowledge. Applications of expert-based classifications are often inconsistent because criteria for assigning relevés to vegetation units are seldom given explicitly. Still, there is, e.g. in nature conservation, an increasing need for a consistent application of vegetation classification using computer expert systems for unit identification.
We propose a procedure for formalized reproduction of an expert-based vegetation classification, which is applicable to large phytosociological data sets. This procedure combines Bruelheide's Cocktail method with a similarity-based assignment of relevés to constancy columns of a vegetation table. As a test of this method we attempt to reproduce the expert-based phytosociological classification of subalpine tall-forb vegetation of the Czech Republic which has been made by combination of expert judgement and stepwise numerical classification of 718 relevés by TWINSPAN. Applying the Cocktail method to a geographically stratified data set of 21794 relevés of all Czech vegetation types, we defined groups of species with the statistical tendency of joint occurrences in vegetation. Combinations of 12 of these species groups by logical operators AND, OR and AND NOT yielded formal definitions of 14 of 16 associations which had been accepted in the expert-based classification. Application of these formal definitions to the original data set of 718 relevés resulted in an assignment of 376 relevés to the associations. This assignment agreed well with the original expert-based classification. Relevés that remained un-assigned because they had not met the requirements of any of the formal definitions, were subsequently assigned to the associations by calculating similarity to relevé groups that had already been assigned to the associations. A new index, based on frequency and fidelity, was proposed for calculating similarity. The agreement with the expert-based classification achieved by the formal definitions was still improved after applying the similarity-based assignment. Results indicate that the expert-based classification can be successfully formalized and converted into a computer expert system.
Land cover data for landscape ecological studies are frequently obtained by field survey. In the United Kingdom, temporally separated field surveys have been used to identify the locations and magnitudes of recent changes in land cover. However, such map data contain errors which may seriously hinder the identification of land cover change and the extent and locations of rare landscape features. This paper investigates the extent of the differences between two sets of maps derived from field surveys within the Northumberland National Park in 1991 and 1992. The method used in each survey was the Phase 1 approach of the Nature Conservancy Council of Great Britain. Differences between maps were greatest for the land cover types with the smallest areas. Overall spatial correspondence between maps was found to be only 44.4%. A maximum of 14.4% of the total area surveyed was found to have undergone genuine land cover change. The remaining discrepancies, equivalent to 41.2% of the total survey area, were attributed primarily to differences of land cover interpretation between surveyors (classification error). Differences in boundary locations (positional error) were also noted, but were found to be a relatively minor source of error. The implications for the detection of land cover change and habitat mapping are discussed.
This paper presents a new and simple method to find indicator species and species assemblages characterizing groups of sites. The novelty of our approach lies in the way we combine a species relative abundance with its relative frequency of occurrence in the various groups of sites. This index is maximum when all individuals of a species are found in a single group of sites and when the species occurs in all sites of that group; it is a symmetric indicator. The statistical significance of the species indicator values is evaluated using a randomization procedure. Contrary to TWINSPAN, our indicator index for a given species is independent of the other species relative abundances, and there is no need to use pseudospecies. The new method identifies indicator species for typologies of species releves obtained by any hierarchical or nonhierarchical classification procedure; its use is independent of the classification method. Because indicator species give ecological meaning to groups of sites, this method provides criteria to compare typologies, to identify where to stop dividing clusters into subsets, and to point out the main levels in a hierarchical classification of sites. Species can be grouped on the basis of their indicator values for each clustering level, the heterogeneous nature of species assemblages observed in any one site being well preserved. Such assemblages are usually a mixture of eurytopic (higher level) and stenotopic species (characteristic of lower level clusters). The species assemblage approach demonstrates the importance of the "sampled patch size," i.e., the diversity of sampled ecological combinations, when we compare the frequencies of core and satellite species. A new way to present species-site tables, accounting for the hierarchical relationships among species, is proposed. A large data set of carabid beetle distributions in open habitats of Belgium is used as a case study to illustrate the new method.
During the last decade many electronic databases of vegetation plots, mainly phytosociological relevés, were established in different European countries. These databases contain information which is extremely valuable for both testing various macroecological hypotheses and for nature conservation surveying or monitoring. The aim of this paper is to provide estimates of the number of vegetation plots there are in Europe, how many are stored in an electronic format and to assess their distribution across European countries and regions.We sent a questionnaire to the managers of national or regional databases of vegetation plots and other prominent vegetation ecologists. Meta-data obtained in this way indicate that there are > 4,300,000 vegetation-plot records in Europe, of which > 1,800,000 are already stored electronically. Of the electronic plots, 60% are stored in TURBOVEG databases. Most plot records probably exist in Germany, the Netherlands, France, Poland, Spain, Czech Republic, Italy, UK, Switzerland and Austria. The largest numbers of plots per unit area are in the Netherlands, Belgium, Denmark and countries of central Europe. The most computerized plots per country exist in the Netherlands (600,000), followed by France, the Czech Republic and the UK. Due to its strong phytosociological tradition, Europe has many more vegetation plots than any other part of the world. This wealth of unique ecological information is a challenge for future biodiversity studies. With the alarming loss in biodiversity and environmental problems like global warming and ongoing changes in land use, there is an urgent need for wide-scale scientific and applied vegetation research. Developments of information systems such as SynBioSys Europe and facilitation of data flow between the national and regional databases should make it easier to use these vegetation-plot data
Broadening the scope of conservation efforts to protect entire communities provides several advantages over the current species-specific focus, yet ecologists have been hampered by the fact that predictive modeling of multiple species is not directly amenable to traditional statistical approaches. Perhaps the greatest hurdle in community-wide modeling is that communities are composed of both co-occurring groups of species and species arranged independently along environmental gradients. Therefore, commonly used "short-cut" methods such as the modeling of so-called "assemblage types" are problematic. Our study demonstrates the utility of a multiresponse artificial neural network (MANN) to model entire community membership in an integrative yet species-specific manner. We compare MANN to two traditional approaches used to predict community composition: (1) a species-by-species approach using logistic regression analysis (LOG) and (2) a "classification-then-modeling" approach in which sites are classified into assemblage "types" (here we used two-way indicator species analysis and multiple discriminant analysis [MDA]). For freshwater fish assemblages of the North Island, New Zealand, we found that the MANN outperformed all other methods for predicting community composition based on multiscaled descriptors of the environment. The simple-matching coefficient comparing predicted and actual species composition was, on average, greatest for the MANN (91%), followed by MDA (85%), and LOG (83%). Mean Jaccard's similarity (emphasizing model performance for predicting species' presence) for the MANN (66%) exceeded both LOG (47%) and MDA (46%). The MANN also correctly predicted community composition (i.e., a significant proportion of the species membership based on a randomization procedure) for 82% of the study sites compared to 54% (MDA) and 49% (LOG), resulting in the MANN correctly predicting community composition in a total of 311 sites and an additional 117 sites (n = 379), on average, compared to LOG and MDA. The MANN also provided valuable explanatory power by simultaneously quantifying the nature of the relationships between the environment and both individual species and the entire community (composition and richness), which is not readily available from traditional approaches. We discuss how the MANN approach provides a powerful quantitative tool for conservation planning and highlight its potential for biomonitoring programs that currently depend on modeling discrete assemblage types to assess aquatic ecosystem health.
Aims
Expert systems are increasingly popular tools for supervised classification of large datasets of vegetation‐plot records, but their classification accuracy depends on the selection of proper species and species groups that can effectively discriminate vegetation types. Here, we present a new semi‐automatic machine‐learning method called GRIMP (GRoup IMProvement) to optimize groups of species used for discriminating among vegetation types in expert systems. We test its performance using a large set of vegetation‐plot records.
Methods
We defined discriminating species groups as the groups that are unique to each vegetation type and provide optimal discrimination of this type against other types. The group of discriminating species of each vegetation type considerably overlaps with the group of diagnostic species of this type, but these two groups are not identical because not all diagnostic species have sufficient discriminating power. We developed the GRIMP iterative algorithm, which optimizes the groups of discriminating species to provide the most accurate vegetation classification, using a training set of a priori classified plot records. We tested this method by comparing classification accuracy before and after the GRIMP optimization of species groups using vegetation‐plot records from the Czech Republic a priori classified to 39 phytosociological classes, and three initial sets of candidate discriminating species from different sources.
Results
The GRIMP algorithm improved the classification accuracy at the class level from 65% correctly classified plots in the test dataset before group optimization to 88% thereafter. The other plots were misclassified or unclassified, but misclassifications were reduced by adding further expert‐based criteria considering dominant growth forms.
Conclusions
GRIMP‐optimized groups of discriminating species are very useful for semi‐automatic construction of expert systems for vegetation classification. Such expert systems can be developed from an a priori unsupervised or expert‐based classification of at least some vegetation plots.
Over the past half century, nature protection in the European Union has been increasingly controlled by commitments to policy and legislative frames, notably the Habitats Directive, originating from the European Union and adopted by an enlarging constituency of member states. Habitat (or biotope) classifications developed in association with these policies, first the Palaearctic habitat classification and CORINE, then the EUNIS habitat classification, have provided typologies with definitions of habitat types intended to aid their recognition, mapping, protection and monitoring. Phytosociological expertise and classifications of formally defined plant communities or syntaxa have played a part in the development of these typologies and in interpretation of the Habitats Directive from the start, though this involvement has been complex and sometimes unclear. This paper catalogues this history and shows how the development of increasingly robust definitions of EUNIS habitat types, an overarching European framework of phytosociological syntaxa and very substantial point-source data (relevés) are converging to aid the interpretation and delivery of environmental policy. In particular, crosswalks between EUNIS habitat types and syntaxa, lists of constant, differential and dominant species, standardised habitat descriptions as well as distribution, predictive and indicative maps are now becoming available. The European Red List of Habitats, also based on the EUNIS typology, provides images and other complementary information on distribution, pressures and threats and a Red List assessment. A comprehensive factsheet with complementary fuller environmental parameterisation for each EUNIS habitat type remains a realistic goal.
Fruit de l'exceptionnelle connaissance des ecosystemes forestiers de Jean-Claude Rameau et d'une vaste synthese de la litterature phytosociologique europeenne, le document Les habitats forestiers de la France temperee ― Typologie et caracterisation phytoecologique propose pour la premiere fois un synsysteme decline jusqu'au niveau de l'association pour les forets de la France metropolitaine (hors zone mediterraneenne et Corse), qu'elles soient ou non concernees par la directive "Habitats". Encore en cours d'amelioration, il s'inscrit dans le prolongement des documents de reference en la matiere (Cahiers d'habitats, Prodrome des vegetations de France, document Gestion forestiere et diversite biologique...) ; il se propose de les completer, en lien etroit avec la declinaison du Prodrome des vegetations de France, travail dans lequel il s'inscrit pleinement afin d'aboutir a une typologie commune aux deux documents. Ce document donne acces, pour la premiere fois a cette echelle, a une caracterisation precise et quantitative des conditions floristiques, climatiques et edaphiques de plus de 180 associations forestieres realisee a l'aide de 9 000 releves floristiques et phytoecologiques issus essentiellement de la base de donnees EcoPlant.
The decision to launch a permanent programme of forest inventory in France dates back to 1958, when it was put into law.
Soil moisture and nutritional characteristics are frequently assessed using plant species and community bioindication, e.g., the Ellenberg system of species indicator values. This method, based on complete inventories of plant species present in plots, is time-consuming, which could prevent its general use for forest or other natural land management. Our aim was to determine the impact of a reduction in the time spent to carry out a floristic inventory on the quality of soil characteristic assessment using plant bioindication. We compared the measurements of soil pH-H2O (pH), organic carbon to total nitrogen ratio (C:N) and base saturation (BS) in the 0–5 cm soil layer of 470 plots with the same variables estimated from floristic inventories of increasing duration, using plant indicator values (IV) from the EcoPlant database. The performance of predictions was evaluated by the square of the linear correlation coefficient between measured and predicted values (R2) and the root mean square error (RMSE) of predictions.
Environmental assessments and land-use planning require reliable information on the botanical composition and distribution of habitats. There have been numerous academic studies of inter-observer variation in species-inventory and habitat mapping, but studies addressing the prevalence of inter-observer variation and consequences of poor quality data in professional practice are lacking. This paper addresses these questions via a questionnaire survey of environmental professionals, using the standard Phase 1 and National Vegetation Classification (NVC) survey methods in the United Kingdom. The survey revealed that misidentification of habitat types within survey reports was relatively common (approximating to 20% of all reports seen by respondents over the previous five years). Approximately 40% of respondents who had encountered erroneous reports stated that these had led to inaccurate initial site ecological assessments. Additional field surveys and discussions with surveyors were commonly used to resolve these issues, but for Phase 1 and NVC 26% and 34% of respondents, respectively, had encountered one or more cases where errors resulted in negative consequences for clients commissioning surveys (in terms of extra costs and project delays). Net loss of biodiversity arising from inaccurate reports was reported in at least one instance by 32% and 38% of respondents for Phase 1 and NVC surveys, respectively – results that may contribute to the attrition of natural capital within the UK. The study highlights the need to extend studies of inter-observer variation to consider impacts on environmental assessments and decision-making in professional practice. The potential benefits of introducing an accreditation scheme (favoured by the majority of respondents to the questionnaire) are discussed.
Aim
The unsupervised nature of traditional numerical methods used to classify vegetation hinders the development of comprehensive vegetation classification systems. Each new unsupervised classification yields partitions that are partly inconsistent with previous classifications and change group membership for some sites. In contrast, supervised methods account for previously established vegetation units, but cannot define new ones. Therefore, we introduce the concept of semi‐supervised classification to community ecology and vegetation science. Semi‐supervised classification formally reproduces the existing units in a supervised mode and simultaneously identifies new units among unassigned sites in an unsupervised mode. We discuss the concept of semi‐supervised clustering, introduce semi‐supervised variants of two clustering algorithms that produce groups with crisp boundaries, k ‐means and partitioning around medoids ( PAM ), provide a free software tool to perform these classifications and demonstrate the advantages using example data sets of vegetation plots.
Methods
Semi‐supervised methods use a priori information about group membership for some sites to define centroids ( k ‐means) or medoids ( PAM ) of site groups that represent previously established vegetation units. They identify these groups in a species hyperspace and assign new sites to them. At the same time, they search for a user‐defined number of new groups. We compared the unsupervised, supervised and semi‐supervised methods using an example of a forest vegetation data set that was previously classified using expert knowledge, and assessed how well these methods reproduced vegetation units defined by experts. Then we compared supervised and semi‐supervised methods in a task when a grassland vegetation classification established in one country was extended to two neighbouring countries.
Results and conclusions
Example analyses of vegetation plot data sets demonstrated that semi‐supervised variants of k ‐means and PAM are extremely valuable tools for extending existing vegetation classifications while preserving previously defined vegetation units. They can be used both for identifying so far unrecognized vegetation types in the regions where a vegetation classification already exists and for extending a vegetation classification from a particular region to neighbouring regions with partly identical but partly different vegetation types. Both k ‐means and PAM provide site groups with crisp boundaries, which makes them a simpler alternative to fuzzy clustering methods.
Question
How to classify forest vegetation relevés automatically in the traditional phytosociological system and thus distinguish typical relevés – easier to classify – from relevés with non‐diagnostic species and intermediate relevés?
Material
A data set of 11 324 forest vegetation relevés, including 4880 relevés classified a priori by experts in phytosociology down to the association level (100 associations in 30 alliances and ten classes, covering all forest vegetation and environmental conditions encountered in France).
Methods
A new typicality index was formalized to quantify the probability of automatically classifying a given relevé in the same vegetation unit as would do an expert in phytosociology. Computation of the typicality index is based on two parameters: a level of affinity linked to the number of diagnostic species of the vegetation unit present in the relevé, and a level of differentiation that is greater when the risk of confusion in classification of the relevé between different units is smaller.
Results
The automatic classification was identical to expert judgement for 60% of the 4080 calibration relevés. The typicality index isolated atypical relevés that are more difficult to classify. The model was successfully transferred to classify an independent data set of 6444 relevés from the French National Forest Inventory performed in 2008 and to distinguish 1114 (17%) typical relevés of phytosociological associations.
Conclusions
Interest in this new typicality index is manifold. It is: (1) operational for the current phytosociological system that forms the basis of the Natura 2000 system; (2) easy to implement from characteristics of species and communities; and (3) based on criteria of uncertainty used by phytosociological experts. This study establishes a clear bridge between recent works on collecting, storing and analysing vegetation relevés, on one hand, and the traditional phytosociological approach, on the other. It should trigger further studies on the spatial and temporal distribution of European habitats.
We have applied the recently developed technique of random amplification of polymorphic DNA (RAPD) to the analysis of the relationships among ten cultivars of papaya (Carica papaya L.). Eleven ten-base synthetic oligonucleotides were chosen that gave multiple PCR amplification products using papaya DNA as template. These 11 primers amplified a total of 102 distinct fragments. Cultivars were scored for presence or absence of RAPD fragments and grouped by cluster analysis using simple matching coefficients of similarity. A dendrogram of the ten cultivars was constructed. Of the ten cultivars seven were of the Hawaiian type, and all of these grouped to one branch of the tree. Divisions within the Hawaiian, branch were mostly consistent with the known genetic background of these cultivars. Three non-Hawaiian, cultivars were also analyzed. The minimum similarity detected was 0.7 suggesting that the domesticated papaya germ plasm is quite narrow. Our results show that RAPD technology is a rapid, precise and sensitive technique for genomic analysis.
Question: How may sampling time affect exhaustiveness of vegetation censuses in interaction with observer effect and quadrat species richness?
Location: French lowland forests.
Methods: Two data sets comprised of 75 timed, one‐hour censuses of vascular plants carried out by five observers on 24 400‐m ² forest quadrats were analysed using mixed‐effect models.
Results: The level of exhaustiveness increased in a semi‐logarithmic way with sampling time and decreased with quadrat species richness. After one hour, 20 to 30% of the species remained undetected by single observers. This proportion varied among observers and the discrepancy increased with increasing sampling time. Fixing the sampling time may make richness estimates vary less between observers but the time limit should be at least 30 min to reduce the bias in exhaustiveness between rich and poor quadrats.
Conclusions We advocate the use of sampling methods based on spatially or temporally‐replicated censuses and statistical analyses that correct for the lack of census exhaustiveness in vegetation studies.
Questions: Is it possible to develop an expert system to provide reliable automatic identifications of plant communities at the precision level of phytosociological associations? How can unreliable expert‐based knowledge be discarded before applying supervised classification methods?
Material: We used 3677 relevés from Catalonia (Spain), belonging to eight orders of terrestrial vegetation. These relevés were classified by experts into 222 low‐level units (associations or sub‐associations).
Methods: We reproduced low‐level, expert‐defined vegetation units as independent fuzzy clusters using the Possibilistic C‐means algorithm. Those relevés detected as transitional between vegetation types were excluded in order to maximize the number of units numerically reproduced. Cluster centroids were then considered static and used to perform supervised classifications of vegetation data. Finally, we evaluated the classifier's ability to correctly identify the unit of both typical (i.e. training) and transitional relevés.
Results: Only 166 out of 222 (75%) of the original units could be numerically reproduced. Almost all the unrecognized units were sub‐associations. Among the original relevés, 61% were deemed transitional or untypical. Typical relevés were correctly identified 95% of the time, while the efficiency of the classifier for transitional data was only 64%. However, if the second classifier's choice was also considered, the rate of correct classification for transitional relevés was 80%.
Conclusions: Our approach stresses the transitional nature of relevé data obtained from vegetation databases. Relevé selection is justified in order to adequately represent the vegetation concepts associated with expert‐defined units.
1. In the UK, Phase 1 survey is a standard method of habitat mapping that has been used widely for environmental assessment and management planning. In this paper we make the first rigorous test of the precision with which environmental consultants apply the technique.
2. Six ecologists surveyed independently the same upland site in northern England. In pairwise comparisons between maps, spatial agreement was found to average 25·6% (with a range of 17·3–38·8%) of the area of the study site. The numbers of land cover types that were identified ranged from 13 to 21. Four or more surveyors agreed on the classification of 19% of the study site, while the area of land upon which all six agreed was only 7·9% of the study site. Spatial errors in the positioning of habitat boundaries occurred, but were a relatively minor source of the differences between maps. The majority of differences between maps were due to classification errors. Land cover types with similar species compositions were most frequently confused.
3. Spatially referenced field ‘target notes’ giving additional information on the vegetation mapped in each survey varied in number between 18 and 56. The contents of target notes were inadequate to allow a retrospective assessment of mapping decisions. The total numbers of species listed in target notes varied between surveys from 25 to 145. Sorenson's similarity for species lists derived from pairs of surveys ranged from 18·8% to 63·7%, and was not related to spatial agreement between surveys.
4. Time spent at the field site was not a correlate of any aspect of the results or cost of the survey. Three surveys conducted by members of a professional institute for ecologists were the most expensive, and also recorded larger numbers of target notes and species than the other surveys. However, their maps were no more similar than other pairs of maps.
5. Analysis of the survey results and comparisons with other methods of vegetation mapping suggest that mapping precision could be increased by (i) placing a greater emphasis on use of aerial photographs and other extant map data prior to (and during) field work; (ii) making greater provision for mapping of mosaics and increasing the level of floristic information in habitat definitions; (iii) recording a greater number of more detailed target notes in the field; and (iv) providing office-based support to assist in the interpretation of aerial photographs, and the cross-checking of field surveyors’ preliminary classifications against the contents of target notes and habitat definitions. The current application of the Phase 1 approach by environmental consultants places too great a reliance on decision-making by the (frequently) unsupported lone surveyor whilst in the field.
Question
Detecting species presence in vegetation and making visual assessment of abundances involve a certain amount of skill, and therefore subjectivity. We evaluated the magnitude of the error in data, and its consequences for evaluating temporal trends.
Location
Swedish forest vegetation.
Methods
Vegetation data were collected independently by two observers in 342 permanent 100‐m ² plots in mature boreal forests. Each plot was visited by one observer from a group of 36 and one of two quality assessment observers. The cover class of 29 taxa was recorded, and presence/absence for an additional 50.
Results
Overall, one third of each occurrence was missed by one of the two observers, but with large differences among species. There were more missed occurrences at low abundances. Species occurring at low abundance when present tended to be frequently overlooked. Variance component analyses indicated that cover data on 5 of 17 species had a significant observer bias. Observer‐explained variance was < 10% in 15 of 17 species.
Conclusion
The substantial number of missed occurrences suggests poor power in detecting changes based on presence/absence data. The magnitude of observer bias in cover estimates was relatively small, compared with random error, and therefore potentially analytically tractable. Data in this monitoring system could be improved by a more structured working model during field work.
Every proposed vegetation classification is sooner or later confronted with an accumulation of new data, which has to be assigned to existing vegetation units. Calculation of similarity indices between new relevs (vegetation plots) and constancy columns of established vegetation units is a suitable method for computerised assignment of relevs to these units. This paper compares several similarity indices using simulated data set where either randomly distributed or diagnostic species prevail in the species composition of the tested relev. Traditional indices, based only on species composition, produce different results than similarity indices that consider species fidelity. However, both types of indices failed in some situations and thus cannot be widely accepted as suitable methods of additional relev assignment. Therefore a combined Frequency-Positive Fidelity Index (FPFI) is proposed. This new index includes compositional similarity of an assigned relev with vegetation unit and retains the advantages and lacks the disadvantages of tested indices. The calculation of all these indices is available in the JUICE program (http://www.sci.muni.cz/botany/juice.htm).
McGraw-Hill publications in the botanical sciences, Edmund Sinnott, Consulting Reimpresión en 1972 Incluye bibliografía e índice Traducciòn del Alemàn
Question: How does a newly designed method of supervised clustering perform in the assignment of relevé (species composition) data to a previously established classification. How do the results compare to the assignment by experts and to the assignment using a completely different numerical method?
Material: Relevés analysed represent 4186 Czech grassland plots and 4990 plots from a wide variety of vegetation types (359 different associations or basal communities) in The Netherlands. For both data sets we had at our disposal an expert classification, and for the Czech data we also had available a numerical classification as well as a classification based on a neural network method (multi‐layer perceptron).
Methods: Two distance indices, one qualitative and one quantitative, are combined into a single index by weighted multiplication. The composite index is a distance index for the dissimilarity between relevés and vegetation types. For both data sets the classifications by the new method were compared with the existing classifications.
Results: For the Czech grasslands we correctly classified 81% of the plots to the classes of an expert classification at the alliance level and 71% to the classes of the numerical classification. Correct classification rates for the Dutch relevés were 64, 78 and 83 % for the lowest (subassociation or association), association, and alliance level, respectively.
Conclusion: Our method performs well in assigning community composition records to previously established classes. Its performance is comparable to the performance of other methods of supervised clustering. Compared with a multi‐layer perceptron (a type of artificial neural network), fewer parameters have to be estimated. Our method does not need the original relevé data for the types, but uses synoptic tables. Another practical advantage is the provision of directly interpretable information on the contributions of separate species to the result.
As the major part of a Habitat Survey of Wales, over 80% of the land surface was surveyed in the field between 1987 and 1997 using the Phase 1 method. A resurvey of 294 randomly selected points was carried out during the early stages to audit the quality of the data being collected, leading to the development of a set of recommendations for the surveyors to improve the consistency and accuracy of habitat mapping. Recent studies have indicated a high level of discrepancy between organisations in field habitat mapping using the Phase 1 method. The findings of the Phase 1 audit in Wales are presented here to show the level of repeatability that was achieved within an organisation. There was 76% correspondence in habitat mapping between 'surveyor' and 'assessor' at the level of individual Phase 1 habitat classes. The degree of repeatability varied according to habitat strata: it was highest for modified land cover types (88%), lowest for semi-improved types (56%) and intermediate for semi-natural types (75%). An overall estimate of the repeatability of Phase 1 survey in the study area of 83% was obtained by weighting the figures for the three strata by the proportion of land area occupied by each stratum. This figure increased to 85% when habitats were amalgamated into Broad Habitat groups. These results are considerably better than those reported by studies of consistency between organisations. Most of the discrepancies between surveyor and assessor were caused by differences in habitat identification. However, at almost two thirds of the points where such a difference occurred, the assessor noted that the vegetation was transitional or borderline with that mapped by the surveyor.
Countryside Survey 2000 Quality Assurance Exercise
- M.V. Prosser
- H.L. Wallace
How reliable is the monitoring of permanent vegetation plots? A test with multiple observers
- P Vittoz
- A Guisan
Vittoz, P. and Guisan, A. (2007) How reliable is the monitoring of permanent
vegetation plots? A test with multiple observers. Journal of Vegetation
Science, 18, 413-422. https://doi.org/10.1111/j.1654-1103.2007.
tb025 53.x.
Determination of diagnostic species with statistical fidelity measures
- M Chytrý
- L Tichý
- J Holt
- Z Botta-Dukát
Chytrý, M., Tichý, L., Holt, J. and Botta-Dukát, Z. (2002)
Determination of diagnostic species with statistical fidelity
measures. Journal of Vegetation Science, 13, 79-90. https://doi.
org/10.1111/j.1654-1103.2002.tb020 25.x
Evaluation of the "observer effect
- J.-M Couvreur
- V Fiévet
- Q Smits
- M Dufrêne
Couvreur, J.-M., Fiévet, V., Smits, Q. and Dufrêne, M. (2015) Evaluation
of the "observer effect" in botanical surveys of grasslands.
Interpretation manual of European Union habitats
European Commission DG Environment (2013) Interpretation manual of
European Union habitats. EUR 28.
Additional supporting information may be found online in the Supporting Information section
- S U Pp O Rti N G I N Fo R M Ati O N
S U PP O RTI N G I N FO R M ATI O N
Additional supporting information may be found online in the
Supporting Information section.