Workpackage 5: The influence of hydrology on sexual and vegetative regeneration strategies in floodplain species
Participating partners:
University of Cambridge (UK) (1)
University of Lethbridge (Canada) (4)
Centre National de la Recherche Scientifique –Toulouse (F) (5)
3.1 Objectives
· Completion of genetic analyses of 450+ P. nigra leaf samples from the Garonne study sites. Statistical analysis of this data and field survey data.
· Write up of the Garonne study with the aim of sending off a manuscript by month 33. The write up prompted the cross-checking of some of the genetics data and the need for additional dendrochronological measurements. Statistical analyses of data and technical drawings also needed some refining.
· Produce a conceptual model of the influence of flow regimes on regeneration strategies.
· Completion of genetics analyses of 400 leaf samples from the Drôme study site. Completion of GIS study to describe disturbance in channelised and non-channelised river reaches from a series of aerial photographs and production of a map of leaf collection points. Analysis of combined datasets. Write up of Drôme study with the aim of sending off a manuscript for publication by month 33.
3.2 Methodology and scientific achievements
3.2.1 Drôme River
Molecular analysis of plant material
This work is completed, but results are difficult to interpret. This has slowed progress on joint analysis and publication of work.
Mapping work + GIS study describing disturbance in contrasting river reaches
Maps of the relative positions of leaf sample collection points have been produced. GIS work has been completed using available aerial photographs from the past decade. The results reveal a trend for most erosion of established pioneer stands to occur on the upstream tips. When not swept away, changes to established pioneer stands were more significant in the channelised than non-channelised reaches and there was a greater turnover rate of vegetated islands in channelised than non-channelised reaches.
3.2.2. Garonne River
Molecular analyses of plant material
Microsatellite analyses of over 450 P. nigra samples have been successfully completed using 5 primers pairs (dinucleotide repeats). Sizes of amplified alleles were determined and visualised by automated florescent detection with an ABI 3100 autosequencer. A high number of alleles per locus (between 12 and 32 alleles) and wide allelic ranges among these suggested that the 5 microsatellite loci selected for this study were sufficient to cover the polymorphism present in the Garonne population. Replicate samples and re-runs of the same leaf samples revealed the high repeatability of this molecular genetic technique. Identical genotypes were scored where allele sizes at all five microsatellite loci were identical or the same with one base pair size difference at no more than a single loci.
Statistical analysis of field survey and genetic data sets + relating genetic results to indicators of hydrological disturbance
One way ANOVA’s were applied to test for significant differences in a number of stand descriptors (e.g. stand density, stem diameters, shoot height…) between ‘young’, ‘middle-aged’ and ‘old’ stands. The frequency of identical genotypes in the different stand age categories was also tested using One-way ANOVA’s. Relationships between indicators of stand disturbance (i.e. angle of slant of trees, frequency and magnitude of clumped growth) and the incidence and frequency of identical genotypes within stands in each age group were investigated using Pearson’s correlation coefficient.
Frequency of identical genotypes within young, middle-aged and old stands
A high proportion of black poplar stands contained replicate genotypes; i.e. 50 % of young stands, 100% of middle-aged stands and 50% of old stands. However, when considering the entire data set, the total number of unique, non-replicated genotypes remains high; i.e. out of 450 trees sampled, 351 represented distinct, non-replicated genotypes, 98 represented replicated genotypes and 1 sample yielded no results. The high numbers of unique genotypes despite high apparent frequency of clonal replication, is explained by low levels of replication of any given ‘clonal’ genotype within stands; i.e. clonal genotypes were replicated 1.3 times on average. High numbers of copies of replicated genotypes within stands were otherwise relatively rare with only seven instances of three identical genotypes detected in the same stand and one instance of four identical genotypes detected within the same stand.
The number of different genotypes differed between young, middle-aged and old stands. In young stands, the number of genotypes found among the 15 trees sampled in each stand was significantly greater (p<0.05) than in the middle–aged or old stands, suggesting that there is less clonal replication in young stands. There was no significant difference between middle-aged and old stands in this respect, where the numbers of genotypes per stand averaged 12.5 and 13.1, respectively. In terms of the number of replicated genotypes found among the 15 trees sampled per stand, significantly more (p<0.01) were found in the middle-aged than the young, or old stands. The greater number of replicated genotypes in middle-aged stands also tended to be represented by a significantly (p<0.05) greater number of stems than replicated genotypes in the young or old stands.
The location of replicate genotypes within stands (nearest neighbours in 85% of cases) suggested that flood-training was primarily responsible for the promotion of vegetative copies of distinct genotypes. However, translocated fragments of parental material also contributed to the number of replicated genotypes within and among stands, travelling several kms downstream in some instances.
Frequency of identical genotypes and indicators of stand disturbance
In relating the angle of slant of trees and the number of identical genotypes detected, the positive correlation coefficients derived suggest that the greater the angle of slant of trees, the greater the frequency of identical genotypes in stands, although this relationship was significant (p<0.01) only when combining the young and middle-aged stands or when considering the old stands independently (p<0.1). A weak positive correlation was also observed between trees in stands displaying clumped growth and the incidence of identical genotypes; the relationship, however, was only found to be significant (p<0.05) among middle-aged stands and where stands of all age categories were combined (p<0.1).
3.2.3 Conceptual model of the influence of flow regimes on regeneration strategies
Figure 1.
A. Low frequency, medium impact, early spring floods promote seedling recruitment and some vegetative recruitment. Seedling mortality is high during the first years of recruitment, while vegetative expansion gradually increases during this period with each successive flood. Competition occurs between seed-derived trees and vegetative sprouts in the middle-aged phase. Vegetative sprouts also compete among themselves resulting in some mortality of vegetative recruits and consequently an increase in the proportion of unique genotypes in the old stands and hence in the established populations.
B. High frequency, high magnitude and/or irregular flooding events reduce the starting number of seed-recruited individuals, but favour asexual recruitment. Asexual recruits further reduce the number of seed-recruited individuals in young and middle-aged stands through competition. The proportion of asexual recruits continues to increase in middle-aged stands as successive floods promote vegetative expansion. A decline in the number of replicated genotypes occurs eventually, however, in the middle-aged and old stands when vegetative recruits compete amongst themselves. The result is a slight increase in the proportion of unique genotypes in the old stands.
3.3 Socio-economic relevance and policy implication
The policy implications of these results relate to the planning stages of initiatives aiming at improving river flow management to sustain floodplain biodiversity and particularly where there is a will to sustain the genetic diversity of naturally regenerating populations of black poplar and other related early successional floodplain woodland species (e.g. Salix spp.). The conceptual model (figure 1) clearly illustrates the potential loss to genetic diversity resulting from frequent, irregular and/or high magnitude flooding events and this loss may be more significant for willow species which have excellent re-sprouting abilities, but comparatively poor seedling survivorship. Flood-training will be promoted in favour of recruitment from seed where (i) flooding occurs during periods of seedling establishment in the early summer, (ii) there is a high frequency of medium to high magnitude flooding events. Since the risks of frequent, sustained disturbance are greatest along channelised river reaches, ideally the active alluvial zone should be as wide as possible to increase the ‘heterogeneity of disturbance’; the effect would be that some areas of the floodplain would remain undisturbed for sufficiently long time periods to allow significant numbers of seedlings to become established.
Where monitoring of the genetic diversity of populations is considered these results indicate the importance of taking tree age and location into account. Sampling should thus, be undertaken for a single age category, samples should be collected widely across the population and sample points should be widely spaced (5m+) to avoid sampling in clonal clumps. Small scale sampling can give an idea of the relative proportions of sexual to asexual recruits, or this can be extrapolated from the findings in this study. In making final estimates of the genetic diversity of a population and/or a Minimum Viable Population estimate for a given site, the age distribution of the population and expected clonal frequency should be included in the estimate.
3.4 Discussion and Conclusions
While it has been demonstrated that P. nigra reproduces by both sexual and asexual regeneration strategies in a number of riparian contexts (channelised and non-channelised river reaches of a braided river system and along a semi-constrained reach of a meandering river), asexual regeneration through re-sprouting was not found to be a dominant regeneration strategy in any of these situations. Asexual regeneration tended instead to be highly localised although the number of replicated genotypes (clones) and the number of stems representing each of these replicated genotypes varied with respect to stand age and the particular hydrological disturbance history at a given site. Overall, for similar age groups, there were fewer identical genotypes present in the braided Drôme River compared to the meandering Garonne River. This might be explained by the more erratic shifts in flow direction along the Drôme River (vegetation is subjected to high flow for limited periods and if in a vulnerable position, displaced) compared with the Garonne River where changes in river flow direction are less dramatic and thus, vegetation must sustain the fluctuating forces of river flow over longer periods at a given point which inevitably leads to flood-training in areas subject to the highest flow rates.
3.5 Plan and objectives for the next period
· Sending off manuscript of Garonne work for publication and starting work on a second related publiucation.
· Reopening discussions (May 2003) on solutions to problem results and potential publications to be produced from the Drôme work.