Jianghao ZHAO, Yingying LIU, Xiaoguo BAI, Anping LI, Yanjiao LI, Shiping CHENG, Guang QI

School of Chemistry and Environmental Engineering, Pingdingshan University/ Henan Key Laboratory of Germplasm Innovation and Utilization of Eco-economic Woody Plant, Pingdingshan 467000, China

Abstract [Objectives] The paper was to reveal the phylogenetic structure of low altitude forest communities in Baotianman Mountain. [Methods] Ten 20 m×20 m plots were set at an altitude of 249-796 m in Baotianman Nature Reserve, in which 62 plant species were found. A phylogenetic tree was established on Phylomatic website, and the community related phylogenetic indexes at low altitude were calculated and analyzed, including net relatedness index (NRI), nearest taxon index (NTI), phylogenetic diversity Faith’s PD index, phylogenetic β diversity PhyloSor index and Dnn index. [Results] The NTI and NRI values of low altitude forest communities were generally greater than 0, showing phylogenetically clustered communities, and habitat filtration was the dominant factor in community construction. Combined with the species abundance matrix of plots, it was found that PD index had a significantly positive correlation with species richness. Phylogenetic β diversity PhyloSor index was larger in the lower altitude area, which was exactly opposite to Dnn index. In other words, the more similar the community, the closer the distance between species, and the communities were phylogenetically clustered. [Conclusions] The study can provide a scientific basis for forest community management and restoration.

Key words Phylogenetic tree, Community phylogeny, β diversity index

1 Introduction

Phylogeny refers specifically to the formation and development of a community, and it is relative to ontogeny. A common ancestor is often found when tracing the roots of any two species, so how did those ancestors evolve into current species? What are the factors at work in evolution? Whether the genetic relationship of species in a community is closely related or distant? The above questions can be answered to some extent through the study of phylogeny. Phylogenetic tree is often used to display phylogenetic research results, which represent the genetic relationships among species through tree-like bifurcation diagram, and then calculate and analyze to predict the evolutionary relationships between species[1].

Community phylogeny is the evolutionary result of long-term effects of various environmental factors[2]. Previous studies on community structure, composition and diversity mainly analyzed the ecological processes (such as symbiosis, interference and competition) and environmental conditions in the community, and rarely considered the influence of evolutionary history on the formation of the community, resulting in very one-sided conclusion. In order to overcome one of the shortcomings, current studies have applied the research methods of phylogenetic biology to community ecology, in which the phylogenetic status of species is used to infer the phylogenetic status of communities, to analyze the interspecific relationship of communities, to elaborate the influence of evolutionary history on community phylogeny, and to discuss the leading factors[1].

Phylogenetic community ecology is a science that studies the coexistence of species within a certain range in a community, the maintenance of biodiversity mechanism and the mechanism of community composition at the community level by using phylogenetic analysis[3]. Phylogenetic analysis can examine the role of competitive exclusion and habitat filtration in community construction. Modern hypothesis of species coexistence suggests that species evolution is not only a genetic process, but also an ecological process. Different evolutionary modes correspond to different ecological processes, and the functional structure and phylogenetic structure of community construction may also be different. Studies on community phylogenetic structure are helpful to understand the response degree of community diversity to disturbance and the community construction mechanism in different succession stages, and can provide more scientific basis for community management and restoration[4].

2 Overview of the survey area

Baotianman National Nature Reserve is located in Neixiang County, Nanyang City, Henan Province, 33°20′12″-33°35′ 43″ N, 111°46′55″ -112°03′32″ E. Baotianman is located in the transition zone from warm temperate zone to north subtropical zone, as well as the transition zone from the second terrain to the third terrain. The dominant community isQuercusvariabilisforest at an altitude below 1 200 m,Q.serratavar.brevipetiolataforest at an altitude of 1 100-1 300 m,Q.alienavar.acuteserrataforest at an altitude of 1 300-1 600 m, mixed coniferous broad leaved forest ofPinusarmandiiandQ.alienavar.acuteserrataat an altitude of 1 600-1 750 m, and elfin forest ofQ.alienavar.acuteserrataandBetulachinensisat an altitude above 1 700 m; there are alsoB.albosinensis,B.platyphylla,Sorbusalnifolia,Platycaryastrobilacea,PopulusdavidianaandAcerforests distributed in local district[5]. The unique conditions of the reserve provide the idea and practical basis for studying the phylogenetic characteristics of low altitude forest community in Baotianman Mountain.

As for the study of community phylogeny of Baotianman forest, Renetal.[5]mentioned in the change of community phylogeny structure of deciduous broadleaved forest at different diameters in the transition zone between warm temperate zone and north subtropical zone that the ecological process of community phylogenetic structure of Baotianman forest could be analyzed by dividing the study area into different scales and diameter classes and comparing the changes of community phylogenetic structure under different classification conditions. Through the comparison of different zero models, it was found that the values ofNTI(nearest taxon index) andNRI(net relatedness index) decreased with the increase of the study scale and diameter class, indicating that the phylogenetic structure of the community in this study area diverged with time and space changes[6], and the phylogenetic structure of large-diameter individuals was more affected by phylogenetic density restriction than small-diameter individuals.

3 Survey methods

3.1 Setting and survey of sample plotsIn this study, 20 m×20 m plots were designed at an altitude of 249-796 m in Baotianman research area (Table 1). The longitude and latitude, altitude, slope’s gradient and aspect and other information of the plots were measured and recorded by tape measure, GPS and other instruments. All tree species withDBH≥1 cm in the plots were identified and named. The abundance, height, canopy width and canopy density of each species were investigated, and their habitats and disturbance status were recorded. The above information were stored in the computer and converted into the corresponding format for storage, providing basic information for future data analysis.

Table 1 Main information of sample plots at low altitude of Baotianman research area

3.2 Parameter selection

3.2.1Phylogenetic characteristic index. In this paper, two related phylogenetic relationship indexes were studied to represent the phylogenetic characteristics:NTIandNRI. By comparing the phylogenetic distance between species in plot communities with that in standardized zero model (assuming random distribution of species), and analyzing the differences between them, the presence or absence of the community phylogenetic structure can be detected[9]: If the phylogenetic distance simulated by zero model was significantly larger than that of the actual community, indicating that the community had a phylogenetic structure and was phylogenetically over-dispersed; on the contrary, if the results of zero model were significantly smaller, the community was phylogenetically clustered; if there was no obvious difference between them, it indicated that the community showed phylogenetic randomness or had no phylogenetic structure[8]. Darwin believed that species in the same genus are more similar in structure and living habits than those outside the genus, and intraspecific competition is much more intense than interspecific competition. This view accords with the phylogenetic conservative hypothesis in community studies: the closer the species are, the more similar their ecological characteristics are[3]. In a community, if habitat filtration plays a dominant role, phylogenetic cluster will be displayed by selecting species with similar adaptability and closely related relationships in the same habitat. Conversely, if competitive exclusion is the dominant role of community construction, it will prevent species with similar ecological niches from co-existing in the same habitat and the community is phylogenetically over-dispersed, and then the genetic relationship between species will be relatively distant. Therefore, on the premise of phylogenetic conservation, the main factors in community construction can be inferred from the phylogenetic performance of the community[8].

3.2.2Phylogenetic diversity index. (i) Phylogenetic diversityPDindex. Compared with other diversity indexes such as species diversity and functional diversity, phylogenetic diversity has a great advantage in explaining productivity variation. Phylogeny was first introduced into ecology primarily to remove the influence of common ancestors on the traits of related species. But as more research has progressed, ecologists have found that the process of community construction can be well represented by the phylogenetic relationships of species. Phylogenetic diversity is one of the best indicators to explain productivity variation when some key traits are missing. With the progress of science and technology, the relationship of evolutionary history of construction process has been more and more widely applied in ecology. For example, the phylogenetic diversityPDindex proposed by Faith[10]refers to the sum of all branch lengths in the phylogenetic tree that contains and only contains species in a community. The number of common ancestors of all species in the plots (number of phylogenetic tree nodes) determines the size of phylogenetic diversity index, such asPDand other similar indexes. However,PDindex mainly considers the existence of the community and ignores the influence of relative abundance on calculation results. Therefore, relative abundance is considered as a common index of phylogenetic diversity reflecting community evolutionary history and species abundance. Based on the study of phylogenetic diversity, phylogenetic trees can be constructed through software packages such as ape and picnate in Phylomatic or R[11].

(ii) Phylogenetic β diversity index. Species β diversity refers to the variation of species with spatial gradient. In ecology, a variety of indexes can be used to quantify this change relationship, so as to study the importance of biological diversity and environmental factors on organisms[12]and to quantitatively reflect the spatial changes of species, but it is difficult to describe the information of genetic relationship, history, evolution and so on, which can provide new ideas for the study of community construction process. Hence, ecologists proposed the method of phylogenetic β diversity to describe the phylogenetic relationships between communities[13]. Compared with traditional species β diversity, phylogenetic β diversity is more advantageous because it can not only quantitatively describe the spatial variation of phylogenetic relationships, but also represent the genetic relationship between species. Therefore, phylogenetic β diversity has been introduced in community ecology, which reflects the degree of species change in phylogenetic relationships in different spaces and provides a way to assess changes in related traits and community structure in terms of environmental gradient and space. The study of community phylogenetic β diversity can provide a fundamental understanding of the interaction between the evolutionary and ecological processes of biodiversity patterns in different regions. In a small spatial range, when there is competitive exclusion (or natural enemy attack) among similar species in the habitat, the phylogeny of the community should be low, and the corresponding β diversity also shows a trend of change[5]. With the development of various related studies, a variety of phylogenetic β diversity indexes have been developed and applied, mainly includingPhyloSor(phylogenetic serensen’s index),Dpw(mean pairwise distance),Dnn(mean nearest taxa distance),RaoH,RaoD,etc.[14].PhyloSorindex andDnnindex mainly reflect the differences between the branching terminal communities of evolutionary trees, which can better reflect the utilization of resources in the community. It has been proved thatDnncan better reflect the phylogenetic situation of the community within the scope of 20 m× 20 m. Thus,PhyloSorindex andDnnindex were selected to represent the changes of β diversity in community during phylogeny[10].

3.3 Data processingThe survey data and information were sorted out, and the species abundance matrix was plotted by Excel. The species in the sample plots were matched with their Latin names one by one, and the family and genus information of each species was retrieved by using plantlist program package in R language[7]. Based on Zanne’s phylogenetic tree published in Nature in 2014, the information about the species and their families and genera were searched on Phylomatic website (http://phylodiversity.net/phylomatic), and the phylogenetic evolutionary trees of 62 species investigated in Baotianman plots were obtained[15].

Using picante, simba, and hydroTSM packages in R language[16], and combined with the constructed phylogenetic tree, thePDindex,PhyloSorindex,Dnnindex,NRI(formula 1) andNTI(formula 2) of community phylogenetic diversity in each section were calculated. The calculation principles ofNTIandNRIwere as follows.

The 62 species surveyed were taken as a species pool and randomly sampled from the pool; the same species of each plot of each section were selected and repeated 999 times, and the mean nearest phylogenetic taxon distance (MNTD) was obtained under the standard model; afterwards, the results were normalized to obtain the value ofNTI. The interspecific mean genetic distance (MPD) should be first obtained for the calculation ofNRI.

(1)

(2)

whereMPDandMNTDwith subscriptsamplerepresent their true values respectively;randsamplerepresents the average evolutionary distance obtained by random sampling of the phylogenetic tree;SDrepresents the standard deviation.

The trend of different indexes was plotted, and the quantified results were analyzed and compared.

4 Results and analysis

4.1 Construction of phylogenetic treeThrough the web version of Phylomatic[17], the phylogenetic tree (Fig.1) was built based on the phylogenetic tree published by Zanne in Nature in 2014[15], providing a basis for studying the influence of phylogenetic structure.

Fig.1 Phylogenetic trees constructed by 62 woody plants in Baotianman survey area

4.2 Analysis ofNRIandNTIAs shown in Fig.2, theNRIandNTIvalues in low altitude areas were generally greater than 0, which was significantly different from the hypothesis of zero model. There was phylogenetic structure in the community, and communities in low altitude areas were mainly phylogenetically clustered, especially in plot 6 and plot 7, which showed extremely significant results (P<0.01). In other words, habitat filtration played a major role in community construction and the results were significant.

Note: The P value in the figure represents significance; P<0.05 indicates significant difference, P<0.01 indicates extremely significant difference, and P>0.05 indicates insignificant difference.Fig.2 Genetic relationship index in low altitude areas

4.3 Phylogenetic diversityPDindexPDrefers to the sum of the length of phylogenetic branches of a taxon in a given sample plot. Combined with the species abundance matrix of various plots, it was found thatPDindex showed a significant positive correlation with species richness (Fig.3).

Fig.3 Changes of PD index with altitude in low altitude areas

4.4 Phylogenetic β diversity index

4.4.1PhyloSorindex of similarity.PhyloSorrefers to the proportion of the distance of the common species of the two plots to that of all species of the two plots in the phylogenetic tree. The larger the proportion, the greater the similarity between the plots.

(3)

whereBLabrepresents the distance sum of common species in the two plots;BLaandBLbrepresent the total distance sum of all species in plotaand plotb, respectively[18]. As shown in Fig.4, thePhyloSorindex of the communities in low altitude areas was larger, and the community similarity of adjacent plots was greater than that of other plots.

Fig.4 PhyloSor index in low altitude areas

4.4.2Dnnindex.Dnnis the mean nearest taxa distance weighted by species abundance, indicating the average developmental distance between a species in one community and the closest related species in the other community.

(4)

whereSMrepresents the total number of species in the communityM;SNrepresents the total number of species in the communityN;farepresents the relative abundance of speciesain the communityM; minδaNrepresents the developmental distance between speciesain the communityMand the nearest species in the communityN; minδbMrepresents the developmental distance between speciesbin the communityNand the nearest species in the communityM;fbrepresents the relative abundance of speciesbin the communityN.

The overall phylogenetic distanceDnnis shown in Table 2. The farther the developmental distance between the two most closely related species in the community, the farther the phylogenetic relationship between species.

Table 2 Phylogenetic β diversity Dnn index in low altitude areas

5 Discussion

5.1 Phylogenetic index relationshipUnder the premise that phylogeny is conservative, the phylogenetic structure of the community is present along with changes in altitude, which is manifested as that the community phylogeny structure gradually changes from phylogenetically clustered to phylogenetically over-dispersed, or the cluster degree decreases with the increase of altitude gradient. Habitat filtration is the dominant factor in community construction in Baotianman low altitude area, in which NRI considers the genetic relationship by calculating the overall level of the community, reflecting the phylogenetic pattern occurring in the entire evolutionary tree. NTI is based on the phylogenetic relationships between closest species and mainly targets the species at the end of the phylogenetic tree[19]. We speculate that the main reasons for phylogenetic cluster in low altitude areas are as follows.

(i) The forest community at low altitude is greatly disturbed by human activities, and the forest vegetation under natural conditions is destroyed. Through human planting or felling, it gradually develops into artificial forest or secondary forest, and the remaining natural vegetations are not dominant in the phylogeny, which are in the early succession stage, presenting a very low naturalness. Human or natural disturbance (such as fire, felling,etc.) can filter the environmental factors and make the plant community move towards phylogenetic cluster.

(ii) The low altitude area is less affected by environmental factors such as temperature, moisture, sunlight and radiation than the high altitude area, and the heterogeneity of habitat and terrain is lower than that of the high altitude area. These conditions make communities with similar traits and similar degrees of evolution coexist in the environment.

5.2 Phylogenetic diversity indexPhylogenetic diversityPDindex does not change significantly at the small scale in the low altitude area, while phylogenetic β diversityPhyloSorindex increases with the increase of altitude. It is in agreement with the conclusion obtained by Luetal.[8]in the study of the phylogenetic structure of subtropical forest community in Ailao Mountain that changes along the altitude gradient: The diversity index decreases gradually with the increase of altitude. The phylogenetic β diversityDnnindex increases with the rising of altitude. However, compared with other plots, the third plot in the low altitude area shows a larger value ofDnn. In combination with the investigation of plots, it was found that the vegetation type of plot 3 community isP.orientalisartificial forest, which is strongly influenced by human activities, and is quite different from natural vegetation communities in other plots, that is, theDnnof the communities is larger.

The overall altitude of the sample plots investigated in this experiment is less than 1 200 m, with relatively mild environmental conditions suitable for biological growth, and there is no strong environmental stress, with small environmental pressure. Moreover, the low altitude area is disturbed by human activities, which also has a certain influence on the results. Due to the dominant role of habitat filtration in community construction, environmental factors play a restriction role in low altitude areas, and coexisting species must have similar traits in order to adapt to the environment, which leads to an aggregation of closely related species and screens species with similar functional traits and adaptability, resulting in closer phylogenetic relationship of the communiy[20].

6 Deficiency and prospect

This paper only considers the influence on the phylogenetic related indexes in the small-scale spatial range at low altitude, and the study content of the influence of community phylogenetic structure is relatively simple and has certain limitations. The research object of this experiment was small scale plots of 20 m×20 m. Due to the limitation of workload and work experience, the scale of selected sampling plots was not expanded. Some studies have pointed out that with the increase of the spatial scale of the research scope, phylogenetic cluster is more likely to occur in communities[20]. In addition, the community phylogenetic structure is influenced by many factors, such as functional traits, study scale and environmental factors. In future experiments, when the community phylogenetic structure is used to judge the ecological process and construction mechanism of the community, it is necessary to expand the sampling range, add more samples, and carry out comprehensive studies combining with the functional traits of plants. For example, if a cluster of phylogenetic structures is found in a community, the environmental factors that may lead to such a structure should be screened out first, whether the functional traits exhibited by these environmental factors are based on phylogenetic conservation is examined, and finally the role of different effects in community construction or ecological processes can be judged[21].