László Orlóci

Late Quaternary Araucaria forest, grassland (Campos), fire and climate dynamics, studied by high-resolution pollen, charcoal and multivariate analysis of the Cambará do Sul core in southern Brazil

Abstract Late Quaternary vegetation, fire and climate dynamics have been studied based on high-resolution dated pollen and charcoal samples and multivariate data analysis. The samples were taken from a 212-cm-long sediment core of a bog in the Cambara do Sul region on the highlands of northeastern Rio Grande do Sul State, Brazil. The records, including seven AMS radiocarbon dates, span 42 840 14C years, for the first time extending the reconstruction of past environmental changes on the southern Brazilian highlands back to the Last Glacial Maximum (LGM) and pre-LGM times. The last 1100 years provide a decadal resolution. Initially the site was a permanent shallow lake which became seasonally dry after 26 900 14C yr BP. Seasonal climate with a long annual dry period prevailed until the late Holocene. The climate was somewhat wetter from 42 840 to 41 470 14C yr BP and from 41 470 to 26 900 14C yr BP than during the LGM and the late-Glacial period. Natural fires were rare, but became very frequent after 7400 cal BP, suggesting human occupation of the southernmost Brazilian highlands since that time. The records suggest that a species-rich Campos (grassland) vegetation existed in the area under a relatively dry and cold climate during glacial times under possibly as low as −10°C. The record also suggests that small populations of Araucaria were probably only present in refugia of deep and protected valleys and/or on wetter coastal slopes. Campos vegetation existed through the early and mid-Holocene until 4320 cal yr BP, after which Araucaria forest expanded into the network of gallery forests along the streams. By 1100 cal yr BP the Araucaria forest replaced the Campos vegetation reflecting the onset of the wettest period with no marked annual dry season. The marked expansion of the Araucaria forest coincided with the reduction in fire. Between AD 1520 and 1770 Weinmannia became a common taxa in the Araucaria forest, suggesting a shift to warmer climatic conditions on the highlands. This interval was synchronous with a cool phase within the Little Ice Age known from North Atlantic land records. After about AD 1780 human activities changed the original forest composition, first by introducing cattle into the forest and than by selective logging of Araucaria trees. Multivariate analysis of the pollen data shows compositional changes that follow a trajectory alternating undirectional, random phases and phases with directional, sometimes fast transitions. The results also show that compositional changes in the vegetation are slower during cool periods (LGM compared to pre-LGM) and faster in warm periods (Holocene).

On randomization testing in vegetation science: multifactor comparisons of relevé groups

. Hypothesis testing in phytocoenological applications is likely to be hindered when based on conventional statistical methods. The problem created by unrealistic assumptions can, however, be overcome by randomization. This paper discusses the general idea of randomization testing, describes a method and interprets its application in group comparisons. Two sets of variables are involved, the vegetation set on the basis of which the groups are compared and the environmental factors which delimit the groups under different analytical designs. Although simple partitioning of sum of squares is at the core of the test, the method has versatility of testing uni- or multifactor designs, which is novel in phytocoenological applications. The algorithm has been implemented in programs SYNCSA and MULTIV by V.P. Data from the Campos of southern Brazil are used for illustration.

ANALYSIS OF CONCENTRATION AND DETECTION OF UNDERLYING FACTORS IN STRUCTURED TABLES

If the rows and columns of a data table are rearranged in such a way that the non-zero values are concentrated in distinct blocks, the table is said to be structured. In such tables, two specific properties are often subject to interpretations: the sharpness of the block structure and the compositional variation among the blocks. A simple chi square analysis is satisfactory to test the overall sharpness of the block structure. The interpretation of compositional variation is more difficult, since the dimensionality of such variation will almost certainly exceed the dimensionality of the table. A solution can nevertheless be obtained based on a two-step analysis. In the first step, the canonical variates of compositional variation are identified. These are characteristics of the vegetation. In the second step, the environmental variables are identified which are highly correlated with the canonical variates. The example in the paper uses data from fossil flood plains. The data table is structured into 18 blocks of 6 species groups and 3 relevé groups. The chi square test indicates a sharp block structure. The structuring generated two canonical variates. The dominant of these two signifies the environmental influence of changing elevation on the flood plain. The second seems to be a response of the vegetation to disturbance. The algorithm which performs analysis of concentration is AOC. A program is available to enquirers free of charge. The programming language is BASIC.

Data Centering: A Review and Evaluation with Reference to Component Analysis

The purpose of the work described in this paper is to review several data centering techniques and to evaluate their effect on the outcome of component analysis. Data centering is defined as a translation of the co-ordinate origin to the centroid of the system. The conditions under which the Rand Qtechniques of component analysis are applicable as alternative strategies are discussed and the techniques are illustrated in a numerical example. A mathematical appendix is included.

Information analysis in phytosociology: partition, classification and prediction.

Abstract Information content is a physical property, related to the likelihood of an observation. Because it is independent from the observer, information content represents an ideal medium for the analysis of biological relationships. The application of information analysis to phytosociology is rapidly gaining momentum, particularly in the field of diversity studies and, to a lesser extent, in descriptive, hierarchical, classifications. While its potential as a classification technique is further considered, the use of information analysis for the partition of data vectors and measurement of the predictive value of classifications is also discussed in detail in the present paper. Concerning partition, the information content of a species may be visualized as a composite of two independent components, arising from thefactors of establishment and survival, and variable performance, respectively in the different sites. If the performance component is large, the replacement of the quantitative observations by presence/absence scores probably results in a significant loss of information. The loss, however, can be measured and also, it can be tested for significance. Regarding classifications, several sorting procedures are discussed whereby, on the basis of the use of information content, individual sites can be assigned to classes of known parameters in which they best fit or, alternatively, hierarchical classifications can be constructed from sites of a previously unclassified plant community or habitat type. Site is defined as the location of a vegetation stand or habitat. The degree of the relatedness of two classifications ( A , B ), produced by independent criteria from the same set of sites, may be used as a measure of the overall predictive value of A with respect to B , or vice versa. Numerical examples are given in the Appendix.

On information flow in ordination

The path of information flow is described and major sources of information loss are identified in ordinations. These sources are shown to be related to the resemblance function and to the transformations which produce the ordiantion co-ordinates. Preventive steps to minimize information loss are suggested.

Chaotic dynamics in a multispecies community

It is shown that community dynamics is neither haphazard nor completely directed. This is quite clear from our examination of a concrete example where recovery dynamics in vegetation progressed from an early phase of strong linear determinism to intense randomness with phase transition defined by density. Is it possible to reconstruct the two phase structure in simple mathematical terms? The results show that it is, and that the model is very simple: a discrete-time Markov chain with white noise. Interestingly, the long-term behaviour of the model is complex chaotic and explosive, suggesting that progression from dominant randomness to determinism is a distinctly probable event. And thus a conceptual foundation is laid, through interlinking initial condition, phase structure and explosive chaoticity, for a unifying theory, in which the classical hypotheses of community dynamics appear as special cases.

Vegetation survey of the Alaska Highway, Yukon Territory: types and gradients.

The vegetation and environment in the vicinity of the Alaska Highway in the southern Yukon were surveyed. Our methods included stratified random sampling and automated data analysis. The vegetation units and compositional gradients, which we recognized, and the nature of the environmental conditions which characterize them, are described. Our results clearly indicate a significant group structure at the level of different vegetation units, and a close relationship of compositional variation in the vegetation to changing environmental conditions. We offer the contents of this report for predictive use in applications where site sensitivity or other ecological conditions have to be assessed, or a framework is needed for a more complete ecological inventory. The authors also thank Misses C. Emo, D. State, Messrs. P. Fewster and P. Merluzzi for their participation in the field work; Miss D. State for assistance in the preparation of the manuscript; Drs. G.W. Douglas and P. Vaartnou for sharing floristic information; Ms. B.J. Woodfield, Messrs. L.W.Bouckhout, R.A. Owens, C. Fouks, M. Romaine, C.E. Brown, E. Nyland, and D. Schuler for contributions in various ways to facilitate completion of the report. The Federal Department of the Environment, ALCAN Task Force, and Foothills Pipe Lines (South Yukon) Ltd., Calgary, Alberta, funded the project, and also supported this publication. Foothills also made available photo mosaics, aerial photographs, and other necessary maps.

Data-Processing in Phytosociology: Retrospect and Anticipation

The Working Group for Data-Processing in Phytosociology of the International Society for Vegetation Science is now to finish its activities. Through four conferences of the Working Group at Trieste (1969), the Netherlands (1971), Prague (1973) and Montpellier (1975), during the 14th Symposium of the Society which was devoted to Basic problems and methods in phytosociology and included 12 contributions from members of the Group (Van der Maarel & Tuxen 1972; see Van der Maarel 1971), during small meetings at the Society’s Symposia in Rinteln, as well as through numerous personal contacts, the Working Group largely achieved its original aim, viz. the assimilation of computer-based data-processing in Braun-Blanquet phytosociology. Furthermore, theories and new methods were developed which are of general interest for quantitative and theoretical plant ecology.

Edge Detection in Vegetation: Jornada Revisited

The paper returns to vegetation edge detection on the Jornada Long Term Ecological Research site, based on contiguous transect sampling, species-free vegetation descriptions, and new continuity/discontinuity criteria. The derivation of these criteria is consistent with the notion that vegetation pattern along the transect confounds random variation, serial dependence, and possibly, real edge-generating effects. Since serial dependence is likely to promote the appearance of continuity, edge detection is focussed on the properties of a residual pattern in the data, remaining after the serial dependences were removed. Two structures help probing the residual pattern for evidence of edges: an ordination map and a vegetation transect profile. The ordination map is a chain of connected line segments in vegetation hyperspace on which the transect segments appear as apices and their paired compositional dissimilarities as inter-apex distances. The transect profile displays deviations of the observed vegetation composition within the transect segments from a state of theoretical continuity for the entire transect. The continuity condition is such that only random variation affects the residual pattern. Sharp apices (small subtending angles), isolated from others by long inter-apex distances in the ordination map mark the points of vegetation discontinuity in the transect. While the subtending angle to an apex reflects changes in population proportions, a criterion for the edge’s sharpness, the height of the deviations profile is a criterion for the edge’s intensity, the expression of a sheer quantitative effect. The deviations profile model does not seem to have direct parallel in previous edge detection techniques. The ordination model is similar to the Ludwig-Cornelius method with its focus on what emanates as an inter-apex distance in our model, but it does not use residuals and ostensibly blurs the edges.