Trait driven or neutral: understanding the change in functional trait diversity during early plant succession using Price partitions

doi:10.21203/rs.3.rs-1999208/v1

Download PDF

Research Article

Trait driven or neutral: understanding the change in functional trait diversity during early plant succession using Price partitions

https://doi.org/10.21203/rs.3.rs-1999208/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 08 Aug, 2023

Read the published version in Plant Ecology →

You are reading this latest preprint version

Habitat filtering, species interactions, and neutral colonisation and extinction dynamics govern the sequence of community assembly and functional diversity during primary plant succession. For an understanding we need to disentangle the underlying abiotic and biotic drivers behind these processes. Here we use data on plant seed size, number and specific leaf area from 107 initially homogeneous study plots along a seven year sequence of primary succession (2005–2011) in a 6 ha German artificial catchment and show that an additive partition of the temporal change in functional diversity (FD) into trait, richness, and abundance effects can according to a recently developed new Price partitioning provide insights into the relative importance of these drivers. Average FD steadily increased during the first four years of succession and reached a plateau in 2009. The annual change in FD was plot specific manifest in a highly positively skewed distribution. Average change in FD were comparatively low up to 2008 and later high. Initial soil characteristics and plant demand traits did not significantly influence the change in functional diversity. We observed trade-offs in the influence of species and abundance effects that tended to decrease the change in FD. We conclude that the high plot specific spatial variability of the annual changes in FD transform an initially homogeneous distribution of plants into a mosaic of very different local plant communities. Our partitioning results also indicate that the successional sequences in FD are in accordance with a hidden Markov series.

Functional traits

Markov chains

plant community assembly

Price equation

Plant functional traits are morphological, physiological, or life history characteristics that directly or indirectly influence plant fitness via their effects on survival, growth, and reproduction (Violle et al. 2012). At the community level the variability in the expression of a specific trait among species, below referred to as functional diversity (FD), informs about niche division and might serve as an indicator of variability during the course of community assembly (e.g. Lavorel and Garnier 2002; Díaz et al. 2007; Garnier et al. 2016).

Three basic processes govern plant community assembly: habitat filtering (Kraft et al. 2015), species interactions, particularly competition (Aschehoug et al. 2016), and neutral colonisation and extinction dynamics (Hubbell 2001). The first two process should predictably distinguish functional diversity from random assembly (Ulrich et al. 2016). In this respect, Kahmen and Poschlod (2004) and Ulrich et al. (2014) reported for several important traits, particularly plant height and seed mass, covariance with species richness and abundances during plant succession. However, the majority of traits analysed by Kahmen and Poschlod (2004) did not show predictable successional trends, indicating trait assembly to be independent of the specific processes of community assembly and to be in accordance with neutral processes. Similarly, Schleicher et al. (2011) failed to confirm that community assembly during plant succession can be predicted from trait distributions (being trait driven in the sense of Keddy 1992). It is possible that filtering and competition processes counterbalance, making it difficult to trace the drivers of FD (Cornwell et al. 2006; Kumordzi et al. 2015). For example, increased interspecific competition and strong habitat filter regimes during succession might lead to limited FD (MacArthur and Levins 1967) despite of increasing richness. Such a trade-off might be manifest in comparatively stable functional diversities during succession (Mason et al. 2008).

In a long-term study of primary plant succession, Zaplata et al. (2013) found that plant cover tended to increase in time, while being associated with high small-scale spatial variability in cover and in species composition. Such a spatial variability might mask the respective successional trends in trait expression and diversity. Additionally, environmental, particularly soil, variability is linked to the respective variability in community composition making it difficult to assess whether community composition or soil variability trigger changes in FD (Robertson et al. 1988).

To disentangle the drivers of FD during plant succession we would need to partition the major components that influence trait expression, particularly richness and abundance effects, and their environmental correlates. Recently, Ulrich et al. (2022) modified the approaches of Fox (2006), Fox and Kerr (2012), and Isbell et al. (2018) and used the Price equation for an additive partitioning of richness, total abundance, intraspecific change in functional diversity, and effects of relative species abundances. Such a partitioning offers an elegant way to disentangle different drivers of plant primary succession within a trait based approach.

Here, we use data from long-term monitoring of early primary plant succession carried out between 2005 and 2011 in a German artificial catchment created to study the initial hydrological processes and the development of soil and vegetation (Frouz et al. 2020). In earlier work we reported already on the high small-scale variability in the pattern of species occurrences and phylogenetic community composition (Ulrich et al. 2017), the steady increase in the diversity of important functional traits (Ulrich et al. 2014) and the comparatively low although increasing degree in FD and multifunctionality (Winter et al. 2019), the latter being the manifold of ecological ecosystem services realized in a community (Hector and Bagchi 2007; Gamfeldt and Roger 2017). Based on these studies and the above discussion we ask three basic questions. 1) What is the major driver of the change in FD during early plant succession, increase in richness, change in abundance, or the dynamics in community assembly? 2) Are there trade-offs between these drivers that stabilize functional diversity? 3) Does the importance of each driver predictably change in time during plant succession?

From 2005 to 2011, we studied the early vegetation succession of a six ha area in the artificial catchment Chicken Creek (German: Hühnerwasser) within the partly decarburised lignite mine Welzow Süd in NE Germany (details in Gerwin et al. 2009) (supplementary material Appendices A and B, Fig. B1). For the present study we used annual data on quantitative plant surveys from 107 single plots of 25 m² (excluding a number of still empty plots in 2005 and 2006) (Fig. A1, details in Zaplata et al. 2013). For each plot, at the beginning of the study initial topsoil parameters (pH, organic carbon (C), and the percentage of sand in the soil) were determined. The complete set of raw data and the Pearson correlation coefficients between the soil variables are contained in the supplementary material Table A1 of Appendix A and Table B1 of Appendix B.

For each species, we estimated the cover degree (abundance) according to a modified Londo scale (Londo 1976; 0.1: ≤0.1%; 0.5: >0.1–0.5%; 1: >0.5–1%; 2: >1–2%, in 1% steps up to 10; 15: >10–15%, in 5% steps up to 30; 40: >30–40%, in 10% steps up to 100). Bryophyta and Marchantiophyta were not identified to lower taxonomic levels. In total, we found 168 morphospecies during the seven study years. The complete data of species identities and abundances of all study years used in this study are contained in Table A1 and in Ulrich et al. (2014, 2016).

We used the Leda (Kleyer et al., 2008) and BioFlor (Klotz et al., 2002) databases and compiled one plant morphological trait (specific leaf area, SLA), one reproductive trait (seed number), and one dispersal trait (seed weight). Additionally, we compiled data of three environmental demand traits assessed by Ellenberg indicator values (Ellenberg et al., 1992) (light, soil moisture, soil fertility). The traits served as proxies to the changing environmental conditions during succession. Raw data and pairwise correlations are contained in Tables A3 and B1.

For the present analysis we constructed three data matrices, a 168 × 4 species × traits matrix T containing for all 168 species the raw values of the four traits, a 168 × 642 species × plot/year matrix M containing the abundance data for all plot × consecutive study year combinations, and a 6 × 642 soil variable × plot/year matrix V, containing the soil parameters. These three matrices are provided in Appendix A.

The inner product T^TM provides the total trait expressions at each of the 642 plot/year combinations. We calculated the FD for each plot from the functional attribute metric (FAD) of Walker et al. (1999) being the normalized sum of all functional pairwise distances between the species of a community. FAD values strongly depend on species richness and abundances (Walker et al. (1999). Therefore, we used two approaches for subsequent data analyses. First, we compared observed FAD values with those obtained from a statistical null standard. This was obtained from a reshuffling of species names across trait values breaking of the plant cover - trait combinations. The associated null distributions were obtained from 200 randomizations. For comparisons we used standardised effect sizes ${SES}_{j}= \frac{{y}_{j}-\mu }{\sigma }$ where y_j denotes the observed parameter and µ and σ are the respective mean and standard deviation of the null model distribution.

Second, Ulrich et al. (2022) used the Price equation of evolutionary biology (Fox 2006) and partitioned the difference in the expression of FAD (ΔFAD) between two communities A and B into five parts:

$$\varDelta FAD=E\left({c}_{A}\right)\varDelta S+{N}_{B}\text{E}(\varDelta p\varDelta c)+{N}_{B}\text{E}(\varDelta c)+{N}_{B}\text{E}(\varDelta p)+f\left(\varDelta N\right)E\left({t}_{A}\right)$$

containing terms quantifying

1) the difference in species richness (ΔS)

2) the average difference in species trait expression (Δc)

3) the difference in relative abundance (Δp)

4) the combined effect of differences in relative abundance and trait expression (ΔpΔc)

5) the effect of differences in abundance (ΔN).

In Eq. 1 E denotes the statistical expectation (mean) in community A of 1) average trait expression c_A, the combined and pure changes in c (Dc) and relative abundance p (Dp), and the average species trait value t_A. N_B denotes the total abundance (here plant cover) in community B (details in Ulrich et al. 2022). We note that these partitions, although being additive, are not linearly independent. Particularly the present richness partitions DS for seed number and seed mass were moderately to strongly correlated with the relative (Dp) and absolute (DN) abundance partitions (Table B2).

Here, we used FAD as the trait of interest and consequently ∆FAD refers to the total change in FAD in a given plot between two consecutive years. We calculated ∆FAD and its partitions for all 643 plot × consecutive year combinations. Eq. 1 allows for a comparison of the drivers that directly influence the changes in FAD and for separate analyses of the underlying covariates.

To answer to our first and second starting question, we used general linear mixed modelling with robust error estimation as implemented in SPSS 28 to relate the average annual changes in FAD and its Price partitions across the 107 plots to species richness (S_A), total abundances (N_A), total trait expression (T_A), and environmental data (initial pH, concentrations of soil organic carbon, and proportion of soil sand), as well as the Ellenberg scores of plant light, soil humidity, and soil nitrogen demands. Study year served as random qualitative predictor. All variables were Z-transformed prior analysis. The environmental predictor variables used in the models were at most moderately corrected and multicollinearity should not influence parameter estimates (appendix B, Tab. B1). To reduce variability in bivariate comparisons we used average values and respective variance - mean ratios (VMR = σ²/µ; where σ and µ are the respective standard deviations and mean values) calculated from all 107 single plots. VMR = 1 indicates a Poisson random distribution. We note that the plots were temporarily not independent and significance levels might be inflated. However, this non-independence, expressed by the changes in trait space within each plot, is exactly what we wanted to study with this approach. Consequently, we focus on general trends and effect sizes.

The importance of single partitions differed between traits, although the change in trait expression Dc had in all three cases the highest impact on the change in FAD as quantified by ∆FAD (Table 1). With respect to seed numbers and SLA, ∆p and ∆N had always a comparatively low impact on ∆FAD, while the change in seed mass was strongly influenced by the change in richness (Table 1).

Table 1

Moments of Price partitions (Eq. 1) of 642 plot × year combinations for trait diversity: arithmetic mean (µ), standard deviation (σ), skewness (γ), and kurtosis (δ). µ_X/µ_T refers to the proportions of ΔS, ΔpΔc, Δc, Δp, and ΔN with respect to ΔFAD.
Moment	∆FAD	∆S	∆p∆c	∆c	∆p	∆N
	No. Seeds
μ	12379.56	3207.99	3472.35	7082.56	-789.42	-593.92
μ_X/μ_T		0.26	0.28	0.57	-0.06	-0.05
σ/μ	3.73	4.57	3.90	5.07	-10.45	-22.64
γ	4.34	6.96	2.64	5.63	-4.41	-2.41
δ	27.81	65.23	10.46	54.94	29.69	26.14
	Seed mass
μ	1914.26	761.82	311.90	940.41	450.08	-549.94
μ_X/μ_T		0.40	0.16	0.49	0.24	-0.29
σ/μ	6.26	4.19	14.58	7.61	8.08	-12.63
γ	4.82	6.94	6.31	8.71	0.96	-3.67
δ	39.94	57.41	83.78	117.42	52.67	34.18
	SLA
μ	56732.86	10774.27	13166.89	30363.78	-305.58	2733.51
μ_X/μ_T		0.19	0.23	0.54	-0.01	0.05
σ/μ	2.23	4.79	3.72	3.19	-51.62	8.95
γ	4.06	7.90	3.18	4.71	-0.89	7.60
δ	19.37	74.75	13.86	26.75	13.99	115.59

Average plot species richness constantly increased during the seven years of succession while average plot plant cover reached a plateau after four years (Figs. 1a, b). VMR of plot species richness scattered in all study years around unity indicating a Poisson random distribution of richness across the study plots (Fig. 1c). In turn, VMR of plant cover was from 2006 onwards well above unity and reached a plateau in 2009 (Fig. 1c).

Average FAD steadily increased during the first four years of succession and reached a plateau in 2009 (Fig. 1d). We did not find respective significant annual trends in the annual changes of FAD (∆FAD) and its partitions (Fig. B2). However, two other patterns were obvious. First, variability in ∆FAD and consequently in the partitions was small during the first three years (Fig. B2). There was a clear step in variability from 2007 to 2008 and from 2008 onwards variability in the annual change of FAD among the study plots was comparatively high (Fig. B2).

We did not find plot specific patterns in the change of FAD (Table B3). Pearson correlations of plot ∆FAD values between study years returned only low correlations explaining less than 10% of variance (Table B3). There was one notable exception. For all three traits the change in FAD between 2007 and 2008 appeared to be synchronized among plots as indicated by the higher plot correlations of r ≥ 0.4.

The respective analysis of higher moments of the distributions of the Price partitions cross all study years revealed a high spatial variance in the average change in FAD and its partitions between the study plots (Table 1). The coefficient of variation was for all three traits and all of the partitions well above 1 or below − 1 (Table 1). The high skewness values (γ > > 0) indicate that this variability in ∆FAD stem from a small number of plots with comparatively high change (Table 1), whereas plot identity changed from year to year (Table B3). The detailed analysis of the spatial distribution of ∆FAD at its partitions revealed temporal trends and trait specific pattern (Fig. 2). The respective distributions deviated for most partitions significantly from zero and were mostly highly positively skewed consistent with a small number of plots with very high annual changes in functional diversity (Fig. 3). With respect to seed numbers (Fig. 3a) and size (Fig. 3c) this pattern was most pronounced in the FAD change from 2007 to 2008, where skewness values peaked. In line with these findings, kurtosis values were in all cases well above δ = 3, indicating a more flat distribution than expected from a normal distributions of FAD change across plots (Table 1).

The average interannual changes in FAD (∆FAD) of seed numbers and SLA, but not of seed size, significantly varied between studies years (Fig. 3). The Price partitioning revealed that the weak and insignificant influences of the model predictors on ∆FAD resulted from respective contrasting effects on these partitions (Fig. 3). Particularly, species richness had insignificant effects on ∆FAD for all three traits (Fig. 3). The reason was a trade-off between the richness and co-variance effects and the changes in single species trait expression and relative abundances (Fig. 3). Similar trade-offs occurred for T_A (seed number, seed size, and SLA) and N_A (seed number). Initial soil characteristics and spatial plot distance did not significantly influence the temporal changes in FAD (Fig. 3).

We have confirmed that an additive partitioning of an aggregate community trait, here FAD, into major components is able to reveal the effect of abiotic and biotic drivers behind this change. In ecology, Price partioning has first been applied to study the richness effect for ecosystem functioning (Loreau and Hector 2001), while Fox (2006), Isbell et al. (2018), and Ulrich et al. (2022) extended the approach to frameworks for the analysis of changes in community properties. Critics have argued that that Price partitions are not statistically independent (Pillai and Gouhier 2019) and lack a simple, unequivocal interpretation, or even become meaningless (Van Veelen 2020). However, the present work as well as Ulrich et al. (2022) and Loreau and Hector (2019) have shown that non-independence (cf. Table B2 for the present data) does not exclude application and that partitions have meaningful interpretation. In this respect they resemble variance partitions and non-orthogonal principle coordinates that can be interpreted despite of being correlated.

Our first starting question asked about the major drivers of functional diversity during early plant succession. This was for all three traits the change in the trait space of individual species followed by the richness effect (Table 1). Changes in relative and absolute abundances were of comparatively minor importance (Table1). If community assembly were temporarily neutral according to the ecological drift model (Hubbell 2001), the species functional equivalence (identical trait spaces) should have caused major effects of richness and abundance but not of trait space. Therefore, we argue that the change in functional diversity during early succession is mainly trait and richness driven while the increase in abundance (here cover) and changes of dominance order are of minor importance.

In this respect, much work has been devoted to the dynamics of trait expression during plant succession. It is generally accepted that local habitat filters in combination with the initial habitat conditions select for certain traits in a temporally ordered sequence leading to predictable changes in community composition at each successional stage (Weiher and Keddy 1999; Zaplata et al. 2013; Chang and Turner 2019, but see Götzenberger et al. 2012). In this respect, Douma et al. (2012) have shown that these changes can be related to changing light availability and initial abiotic soil conditions that shift the expression of major plant functional traits. Here we extend this knowledge to the change in functional diversity, and therefore to the total trait space spanned by a community at a certain successional state.

We found significant differences in FAD change (∆FAD) among study years, although these were only weakly linked to initial soil conditions and only for SLA significantly related to plant habitat demands (Fig. 3). Unfortunately, we lack small scale light conditions at the plot level. The partitioning confirmed that this lack was not due to trade-offs between richness (∆S), composition (∆c∆p, ∆p), and abundance (∆N) partitions (Fig. 3). Our results instead point solely to the levels of richness, abundance and functional diversity that determine the degree of annual change in the course of succession (Fig. 3). Under this hypothesis the change in functional diversity of the three traits studied here would resemble a Markov chain process applied to successional processes and community assembly (Horn 1975; Usher 1981). In classical Markov chain models each subsequent state is solely determined by the previous state in combination with a specific transition rule. In this respect, these models resemble the present Price analysis where the change in trait value (here FAD) of two consecutive temporal states were compared and the partitions contained (except for the fixed multiplicator N_B) the initial state values only. Classical Markov chain models with fixed transition matrices have been found to be too rigid in most cases (e.g. Usher 1981) while extended models with adjustable transition matrices, particularly those with unobserved (hidden) states (Rabiner and Juang 1986), appeared to be suited to describe patterns of temporal and spatial species replacements (e.g. Logofet et al. 1997; McClintock et al. 2020). Forthcoming work has to show whether and how Price partitions can be reformulated within a hidden Markov framework. The associated transition matrices and their covariates would then inform about the quantitative parameters behind the succession process.

Classical Markov models are consistent with neutral community assembly (Steiner and Tuljapurkar 2012) whereas extended and hidden models include various degrees of niche and filter mechanisms (McClintock et al. 2020). The fact that S_A, N_A, and T_A explained generally less than 50% of the variance in ∆FAD indicates that other factors, presumably niche and filter effects, are important. This result does not point to simple neutral mechanisms of community assembly and successional change. Our results thus corroborate Ulrich et al. (2014), who already reported that filter processes gained importance for community assembly and FAD, particularly during the first four years of succession. This filter effect acted mainly via a reordering of species dominances (the ∆p and ∆p∆c partitions, Table 1)

Answering to our second starting question, the present Price partitions gave insight into the trade-offs between the drivers of the change in FAD (Fig. 3). For seed numbers and SLA we found contrasting effects of the richness (∆S) and relative abundance (∆p) components, and to a lesser degree between ∆p and the covariance between trait and relative abundance (∆c∆p) (Fig. 3). This result was unexpected as the temporal increase in richness (Fig. 1) should mainly affect rare species, while the total FAD should be largely unaffected. Apparently, this was not the case and newly colonising species immediately changed the dominance orders with major influence on FAD (Fig. 3). This was most apparent in the FD change from the third to the fourth year of succession (2007/2008, cf. Zaplata et al. 2013). Incoming species tended to increase FAD while the reordering of the abundance distributions mediated this trend leading to a stabilizing effect on the annual changes in FAD (Fig. 3). Indeed classical and current work (Bazzaz 1975; Yin et al. 2018) demonstrated that community evenness increases during plant succession. The higher proportions of species with intermediate abundances should stabilize trait spaces as the probability of rearrangements within this group of intermediate species increases in comparison to rearrangements within the groups of the most abundant and the rare species. Soliveres et al. (2015) called this the proper middle class effect, because these species might take over dominance in cases where dominants face population breakdown. This effect should particularly hold for FAD being calculated from the product of trait expression and abundance. Consequently, our findings support the hypothesis that the temporal variability in relative abundances mediates the increase in total FAD due to increasing species richness in the course of plant succession.

Finally, we asked whether the importance of each partition predictably changes in time during plant succession. This was not the case (Fig. B2). We did not see consistent annual trends for each partition. This lack of trend indicates that annual changes within the study system are spatially and temporarily point patterns triggered solely by the current local environmental state and community composition. Of course, in prior work we found a steady increase in plant species richness and cover (cf. Fig. 1), as well as a trend towards increasing spatial species segregation (increasing b-diversity among plots; Zaplata et al. 2013), but also a high small scale variability in richness and phylogenetic and functional diversity (Ulrich et al. 2014). The fact that environmental, particularly soil conditions were similar after catchment construction indicates that initial small local stochastic processes of plant colonisation or germination from the soil seed bank are able to trigger this high small scale variability in plant community composition. Prior, Wiegleb and Felinks (2001) already found that plant succession in comparable post-mining landscapes did not follow predictable chronosequences. Here we have shown that this finding also hold for the temporal change in functional diversity.

This result is strongly corroborated by the high spatial variability in ∆FAD and its partitions (Tables 1 B3, Figs. 3, B2). Ulrich et al. (2014) already pointed to the high small-scale spatial variability of phylogenetic diversity in this system. The high positive skewness and kurtosis values in the ∆FAD distribution across plots particularly point to a few plots in each year with extraordinary high annual temporal variability in FAD. Importantly, these plots varied in time as indicated by the low correlations of ∆FAD between plots, particularly among consecutive years (Table B3). Therefore, we could not identify highly variable and more predictable plots leaving the successional process being even more driven by hidden factors. Consequently, the annual change in functional diversity appeared again to be a process without memory, typical for Markov processes with variable transition probabilities. However, our study design could not clarify whether an accounting for plot specific annual characteristics, manifest in the species composition, might reveal the hidden underlying patterns. Subsequent plot-specific analyses should uncover predictable successional patterns in this system.

Acknowledgments

The authors thank the working group Z1 (monitoring) members of the SFB/TR 38, who helped us to perform this study and the Vattenfall Europe Mining AG for providing the research site.

Funding

This study was part of the TransRegio Collaborative Research Centre 38 (SFB/TR 38: ecosystem assembly and succession), which was financially supported by the Deutsche Forschungsgemeinschaft (DFG, Bonn) and the Brandenburg Ministry of Science, Research and Culture (MWFK, Potsdam). WU acknowledges funding from the Polish National Science Centre (UMO-2017/27/B/NZ8/00316).

Competing interests

The authors have no relevant financial or non-financial interests to disclose.

Author contributions

WU provided the theoretical background. WU performed analyses and wrote the first draft. MKZ collected the floristic data and created the database. Both authors contributed to revisions.

Aschehoug ET, Brooker R, Atwater DZ, Maron JL, Callaway RM (2016) The mechanisms and consequences of interspecific competition among plants. Ann Rev Ecol Evol Syst 47: 263–281. https://doi.org/10.1146/annurev-ecolsys-121415-032123
Bazzaz FA (1975) Plant species diversity in old field successional ecosystems in southern Illinois. Ecology 56: 485–488
Chang CC, Turner BL (2019) Ecological succession in a changing world. J Ecol 107: 503–509. https://doi.org/10.1111/1365-2745.13132.
Cornwell WK, Schwilk LD, Ackerly DD (2006) A trait-based test for habitat filtering: convex hull volume. Ecology 87: 1465–71. https://doi.org/10.1890/0012-9658.
Díaz S, Lavorel S, de Bello F, Quétier F, Grigulis K, Robson TM (2007) Incorporating plant functional diversity effects in ecosystem service assessments. Proc Natl Acad Sci U S A 104: 20684–20689. https://doi.org/10.1073/pnas.0704716104
Douma JC, de Haan MWA, Aerts R, Witte J-PM, van Bodegom PM 2012 Succession-induced trait shifts across a wide range of NW European ecosystems are driven by light and modulated by initial abiotic conditions. J Ecol 100: 366–380. https://doi.org/10.1111/j.1365-2745.2011.01932.x.
Ellenberg H, Weber HE, Düll R, Wirth V, Werner W (1992) Zeigerwerte von Pflanzen in Mitteleuropa. Scripta Geobot 18: 1–248.
Fox, JW (2006) Using the Price equation to partition the effects of biodiversity loss on ecosystem function. Ecology 87: 2687–2696. https://doi.org/10.1890/0012-9658.
Fox JW, Kerr B (2012) Analyzing the effects of species gain and loss on ecosystem function using the extended price equation partition. Oikos 121: 290–298. https://doi.org/10.1111/j.1600-0706.2011.19656.x
Frouz J, Bartuška M, Hošek J, Kučera J, Leitgeb J, Novák Z, Šanda M, Vitvar T (2020) Large scale manipulation of the interactions between key ecosystem processes at multiple scales: why and how the falcon array of artificial catchments was built. Eur J Environm Sci 10: 51–60. https://doi.org/10.14712/23361964.2020.7
Gamfeldt L, Roger F (2017) Revisiting the biodiversity–ecosystem multifunctionality relationship. Nat Ecol Evol 1: 0168. https://doi.org/10.1038/s41559-017-0168
Garnier E, Navas M-L, Grigulis K (2016) Plant Functional Diversity. Organism traits, community structure, and ecosystem properties. Oxford University Press, Oxford.
Gerwin W, Schaaf W, Biemelt D et al. (2009) The artificial catchment ‘Chicken Creek’ (Lusatia, Germany) - A landscape laboratory for interdisciplinary studies of initial ecosystem development. Ecol Engineer 35: 1786–1796. https://doi.org/10.1016/j.ecoleng.2009.09.003
Götzenberger L, de Bello F, Bråthen KA, Davison J, Dubuis A, Guisan A, Lepš J, Lindborg R, Moora M, Pärtel M, Pellissier L, Pottier J, Vittoz P, Zobel K, Zobel M (2012) Ecological assembly rules in plant communities—approaches, patterns and prospects. Biol Rev 87: 111–127. https://doi.org/10.1111/j.1469-185X.2011.00187.x
Hector A, Bagchi R (2007) Biodiversity and ecosystem multifunctionality. Nature 448: 188–190. https://doi.org/10.1038/nature05947
Horn HS (1975) Markovian properties of forest succession. In Cody ML, Diamond JM (eds.) Ecology and Evolution of Communities, Belknap, pp. 196–211.
Hubbell SP (2001) The unified neutral theory of biodiversity and biogeography. Princeton Univ Press.
Isbell F, Cowles J, Dee LE, Loreau M, Reich PB, Gonzalez A, Hector A, Schmid B (2018) Quantifying effects of biodiversity on ecosystem functioning across times and places. Ecol Lett 21: 763–778. https://doi.org/10.1111/ele.12928
Kahmen S, Poschlod P (2004) Plant functional trait responses to grassland succession over 25 years. J Veg Sci 15: 21–32. https://doi.org/10.1111/j.1654-1103.2004.tb02233.x.
Keddy PA (1992) Assembly and response rules: two goals for predictive community ecology. J Veg Sci 3: 157–164. https://doi.org/10.2307/3235676
Kleyer M, Bekker RM, Knevel IC, et al. (2008) The LEDA traitbase: a database of life-history traits of Northwest European flora. J Ecol 96: 1266–74. https://doi.org/10.1111/j.1365-2745.2008.01430.x
Klotz S, Kühn I, Durka W (2002) BIOLFLOR - Eine Datenbank zu biologisch-ökologischen Merkmalen der Gefäßpflanzen in Deutschland. Schriftenreihe für Vegetationskunde 38. Bonn.
Kraft NJB, Adler PB, Godoy O, James EC, Fuller S, Levine JM (2015) Community assembly, coexistence and the environmental filtering metaphor. Funct Ecol 29: 592–599. https://doi.org/10.1111/1365-2435.12345
Kumordzi BB, de Bello F, Freschet GT, Le Bagousse-Pinguet Y, Lepš J, Wardle DA (2015) Linkage of plant trait space to successional age and species richness in boreal forest understorey vegetation. J Ecol 103: 1610–1620. https://doi.org/10.1111/1365-2745.12458
Lavorel S, Garnier E (2002) Predicting changes in community composition and ecosystem functioning from plant traits: revisiting the Holy Grail. Funct Ecol 16: 545–556 https://doi.org/10.1046/j.1365-2435.2002.00664.x.
Logofet DO, Golubyatnikov LL, Denisenko IA (1997) Inhomogeneous Markov models for succession of plant communities: new perspectives on an old paradigm. Biol Bull 24: 506–514.
Loreau M, Hector A (2001) Partitioning Selection and Complementarity in Biodiversity Experiments. Nature 412: 72–76. https://doi.org/10.1038/35083573
Loreau M, Hector A (2019) Not even wrong: Comment by Loreau and Hector. Ecology 100: e02794. https://doi.org/10.1002/ecy.2794
MacArthur R, Levins R (1967) The limiting similarity, convergence, and divergence of coexisting species. Am Nat 101: 377–385. https://www.journals.uchicago.edu/doi/10.1086/282505
McClintock BT, Langrock R, Gimenez O, Cam E, Borchers DL, Glennie R, Patterson TA (2020) Uncovering ecological state dynamics with hidden Markov models. Ecol Lett 23: 1878–1903. https://doi.org/10.1111/ele.13610
Mason NWH, Lanoiselée C, Mouillot D, Wilson JB, Argillier C (2008) Does niche overlap control relative abundance in French lacustrine fish communities? A new method incorporating functional traits. J Anim Ecol 77: 661–669. https://doi.org/10.1111/j.1365-2656.2008.01379.x
Pillai P., Gouhier TC (2019) Not even wrong: the spurious measurement of biodiversity’s effects on ecosystem functioning. Ecology 100: e02645. https://doi.org/10.1002/ecy.2645
Rabiner LR, Juang BH (1986) An introduction to hidden Markov models. IEEE Assp Magazine, 16pp.
Robertson GP, Hutson MA, Evans FC, Tiedje JM (1988) Spatial Variability in a Successional Plant Community: Patterns of Nitrogen Availability. Ecology 69: 1517–1524.
Schleicher A, Peppler-Lisbach C, Kleyer M (2011) Functional traits during succession: Is plant community assembly trait-driven? Preslia 83: 347–370.
Soliveres S, Maestre FT, Ulrich W, Manning P, Boch S, Bowker M, et al. (2015) Intransitive competition is widespread in plant communities and maintains species richness. Ecol Lett 18: 790–798. https://doi.org/10.1111/ele.12456
Steiner UK, Tuljapurkar S (2012) Neutral theory for life histories and individual variability in fitness components. Proc Natl Acad Sci USA 109: 4684–4689. https://doi.org/10.1073/pnas.101809610
Ulrich W, Piwczynski M, Zaplata MK, Winter S, Schaaf W, Fischer A (2014) Soil conditions and phylogenetic relatedness influence total community trait space during early plant succession. J. Plant Ecol 7: 321–329. https://doi.org/10.1093/jpe/rtt048
Ulrich W, Zaplata MK, Winter S, Schaaf W, Fischer A, Soliveres S, Gotelli NJ (2016) Interspecific interactions and random dispersal rather than habitat filtering drive community assembly during early plant succession. Oikos 125: 698–707. https://doi.org/10.1111/oik.02658
Ulrich W, Zaplata MK, Winter S, Fischer A (2017) Spatial distribution of functional traits indicates small scale habitat filtering during early plant succession. Persp Plant Ecol Evol Syst 28: 58–66. https://doi.org/10.1016/j.ppees.2017.08.002
Ulrich W, Zaplata MK, Gotelli NJ (2022) Reconsidering the Price equation: a new partitioning based on species abundances and trait expression. Oikos 131: e08871. https://doi.org/10.1111/oik.08871
Usher MB (1981) Modelling ecological succession, with particular reference to Markovian models. Vegetatio 46: 11–18.
van Veelen M (2020) The problem with the Price equation. Phil Trans R Soc B 375: 20190355. https://doi.org/10.1098/rstb.2019.0355
Violle C, Enquist BJ, McGill BJ, Jang L, Albert CH, Hulshof C, Jung V, Messier J (2012) The return of the variance: intraspecific variability in community ecology. Trends Ecol Evol 27: 244–52. https://doi.org/10.1016/j.tree.2011.11.014
Walker B, Kinzig A, Langridge J (1999) Plant attribute diversity, resilience, and ecosystem function: the nature and significance of dominant and minor species. Ecosystems 2: 95–113. https://doi.org/10.1007/s100219900062
Weiher E, Keddy P (1999) Ecological Assembly Rules: Perspectives, Advances, Retreats. Cambridge University Press.
Wiegleb G, Felinks B (2001) Predictability of early stages of primary succession in post-mining landscapes of Lower Lusatia, Germany. Appl Veg Sci 4: 5–18. https://doi.org/10.1111/j.1654-109X.2001.tb00229.x
Winter S, Zaplata MK, Rzanny M, Schaaf W, Fischer A, Ulrich W (2019) Increasing ecological multifunctionality during early plant succession. Plant Ecol 220: 499–509. https://doi.org/10.1007/s11258-019-00930-3
Yin Z-Y, Zeng L, Luo S-M, Chen P, He X, Guo W, Li B (2018) Examining the patterns and dynamics of species abundance distributions in succession of forest communities by model selection. PlosOne: journal. pone.0196898 . https://doi.org/10.1371/journal.pone.0196898
Zaplata MK, Winter S, Fischer A, Kollmann J, Ulrich W (2013) Species-driven phases and increasing structure in early-successional plant communities. Am Nat 181: E17-E27. https://doi.org/10.1086/668571

No competing interests reported.

Download PDF

Journal Publication

published 08 Aug, 2023

Read the published version in Plant Ecology →

Editorial decision: Major revision
10 Jan, 2023
Reviews received at journal
28 Dec, 2022
Reviewers agreed at journal
12 Dec, 2022
Reviewers agreed at journal
06 Sep, 2022
Reviewers invited by journal
30 Aug, 2022
Editor assigned by journal
26 Aug, 2022
Submission checks completed at journal
26 Aug, 2022
First submitted to journal
25 Aug, 2022

You are reading this latest preprint version

Trait driven or neutral: understanding the change in functional trait diversity during early plant succession using Price partitions

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Methods

Results

Discussion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1