群落物种多度的分形模型和一般性分布规律的验证与探讨
Verification and Discussion on Fractal Model and the General Pattern on Species Abundance in Community
通讯作者: 苏强(1979-),男,辽宁沈阳人,副教授,主要从事生物多样性、生态化学计量学研究. E-mail:sqiang@ucas.ac.cn
收稿日期: 2021-03-25 修回日期: 2021-05-30 网络出版日期: 2021-07-22
基金资助: |
|
Corresponding authors: SU Qiang (1979-), male, Shenyang City, Liaoning Province, Associate professor. Research areas include biodiversity and ecological stoichiometry. E-mail:sqiang@ucas.ac.cn
Received: 2021-03-25 Revised: 2021-05-30 Online: 2021-07-22
作者简介 About authors
高俊峰(1997-),男,河南南阳人,硕士研究生,主要从事生物多样性研究.E-mail:gaojunfeng20@mails.ucas.ac.cn
解析群落物种间的个体数量关系,也称群落物种多度分布,被认为是理解群落物种多样性决定机制的关键。近年来已建立了许多物种多度分布模型,但却因模型众多且又难以区分而引发了许多争议。已有研究表明,对物种多度分布模型的筛选不仅要考察其与群落样本实测数据的拟合效果,还要检验其能否在更深层次上揭示某些宏观生态学现象。Su建立的分形模型拟合效果较好,并揭示了物种多度分布的一般性规律,即Nr/N1往往趋近于1∶1/2∶1/3……(Nr和N1分别为降序排序中第r位和第1位物种的个体数量)。但该模型尚未得到充分的重视,相关的验证研究也较为欠缺。鉴于此,通过一个全球性群落物种数据库资料,对该模型及物种多度分布的一般性规律进行更为详细的检验。结果显示,分形模型的实际拟合效果很好;物种多度分布的一般性规律可以得到该数据库的支持与验证。上述结果可为比较全面地理解物种多度分布、探究物种多样性的决定机制提供更加可靠的科学依据。
关键词:
The analysis of individual quantitative relationship among community species, also known as the Species Abundance Distribution (SAD), is considered to be the key to understanding what determines species diversity. In recent years, numerous SAD models have been proposed on various theoretical grounds, but it is difficult to draw general conclusions about which models provide the best fit to SADs. Previous studies have shown that the screening of SAD model should not only examine the goodness of fit of SAD model with the empirical data of community samples, but also evaluate model's ability to simultaneously explain some macro ecological patterns. The fractal model proposed by Su has good fit to the empirical data of community samples and reveals the general pattern of SAD; that is, Nr / N1 tends to be 1∶1/2∶1/3… (Nr/N1, Nr and N1 represent the number of individuals of the r-th and the first species in descending order). However, the model has not been given enough attention, and the relevant verification research is also lacking. This paper uses a global community species database to test the model and the general pattern of SAD. The results show that:
Keywords:
本文引用格式
高俊峰, 苏强.
GAO Junfeng, SU Qiang.
1 引 言
自20世纪30年代以来,虽然建立了许多SAD模型,但是哪种模型能更好地解释自然群落的实测数据,目前尚无定论[9]。起初,Motomura[10]根据生态位原理建立了SAD的几何级数模型(Geometric-series model),并对湖泊底栖动物群落进行了分析;Fisher等[11]和Preston[12]则先后建立了对数级数模型(Log-series model)和对数正态模型(Log-normal model),并在昆虫和鸟类群落中得到了验证;此后,断棍模型(Broken-stick model)、中性模型(Neutral model)和Zipf-Mandelbrot分形模型(Fractal model)等理论模型也相继被提出[13,14]。这些模型一方面为SAD研究提供了更加广阔的理论视野[15];另一方面,如何对它们进行区分与甄别,在学界也引起了很多争议和讨论[16~19]。
Su[20]重新建立的SAD分形理论模型似乎可以较好地满足这些要求:
2 研究方法
表1 以群落物种间生物个体数量降序排列,物种排序位数(r)、新增物种倍数(Kr)和第r个物种多度(Nr)之间的关系
Table 1
排序位数r | 新增物种 倍数Kr | Nr与Nr-1的 计算关系 | Nr与N1的 计算关系 |
---|---|---|---|
1 | - | N1 | N1 |
2 | 2/1 | N1⋅(2)-1/d | N1⋅(2)-1/d |
3 | 3/2 | N2⋅(3/2)-1/d | N1⋅(3)-1/d |
r | r/(r-1) | Nr-1⋅[r/(r-1)]-1/d | N1⋅(r)-1/d |
如表1所列,群落内各物种的个体数量可表示为:
……
令p=1/d,则有:
式中:S为群落内物种总数,r为物种的排序位数。
p模型对群落样本实测数据的拟合优度由R2判定,其计算公式为:
式中:SSR为回归平方和,SST为总平方和,
3 数据资料来源及说明
图1
图1
布朗大学浮游有孔虫数据库(BFD)1 265个研究站位全球分布图
Fig. 1
Global distribution of 1 265 research stations in Brown University Foraminiferal Database (BFD)
表2 BFD中p值的计算结果与Su[21]对另外8个数据库p值的计算结果的详细信息
Table 2
数据库名称 | 最大值 | 最小值 | 中位数 | 平均数 | 样本数量 |
---|---|---|---|---|---|
BFD | 6.031 | 0.810 | 1.5336 | 1.649±0.032 | 1 265 |
Diatom | 5.825 | 0.335 | 1.272 | 1.343±0.008 | 3 224 |
Fish | 4.563 | 0.756 | 1.592 | 1.702±0.019 | 761 |
BBS | 2.375 | 0.548 | 0.938 | 0.984±0.004 | 2 769 |
CBC | 3.738 | 0.733 | 1.492 | 1.556±0.008 | 1 999 |
FIA | 2.229 | 0.235 | 0.907 | 0.931±0.003 | 10 355 |
Gentry | 1.851 | 0.352 | 0.827 | 0.872±0.019 | 222 |
MCDB | 3.265 | 0.495 | 1.547 | 1.587±0.052 | 103 |
NABC | 3.112 | 0.540 | 1.240 | 1.278±0.017 | 400 |
图2
图2
BFD中1 265组群落样本分形参数p的频数分布
Fig. 2
The frequency distribution of the fractal parameter p of 1 265 groups of community samples in BFD
本研究选择该数据库的主要原因如下:
4 研究结果
p模型对BFD数据库群落样本拟合优度的计算结果显示,R2的最小值为0.632,最大值接近1.00,平均值和中值分别为0.883±0.004与0.894。其中,R2>0.8的样本占比为86.7%。图3为p模型对其中4个群落样本实测数据的拟合效果。
图3
图3
采用Su[20]提出的分形模型对从BFD中随机选取的4个自然群落的拟合
r按物种多度降序排列后的排序位数,N1和Nr分别表示降序排序中第1位和第r位物种的个体数量;R2为拟合优度,取值范围在0和1之间,R2越接近1,拟合结果越好
Fig. 3
The Su's fractal model[20] fits four natural communities randomly selected from BFD
The r was arranged in descending order of species abundance. N1 and Nr are the abundances of the 1st and r-th species in descending order of species abundance, respectively. R2 is the goodness of fit. The closer R2 is to 1, the better the fitting result is
5 讨 论
目前,不同SAD模型的对比研究已得到了广泛关注[9,25~27],但由于研究方法与数据资料的局限性,大多数结论未能在研究者中形成共识[19~21]。分形理论与其他生态学理论(如生态位理论和中性理论)的比较[13~15,18]及p模型与其他分形模型(如Zipf-Mandelbrot分形模型)在条件假设、适用范围、拟合结果及参数的生态学意义等方面的差异[20, 21]已在过去的研究中得到详细论证,在此不再赘述。基于现有的评判标准[18~20],如何考量p模型的研究价值和科学意义,本研究将从以下3个方面进行分析和探讨:
首先,研究结果表明p模型对BFD数据库群落样本拟合效果很好。一方面,所有样本的R2值均大于0.6。这说明,p模型的拟合曲线与BFD群落样本的实测数据基本一致;另一方面,R2>0.8的样本占比为86.7%。这说明p模型的拟合结果与绝大多数群落样本非常吻合。因此,p模型可以较好地拟合BFD的实测数据。
其次,根据分形参数p的统计分析结果,SAD的一般性规律可以得到BFD数据资料库的支持。BFD中群落样本p的中值和平均值分别为1.533、1.649±0.032,与Su[21]对另外8个数据库p值的计算结果进行对比(表2),9个数据库中p的平均值和中值均接近1。同时,根据p的频数分布图(图2),p值很少有远大于1或非常接近0的情况,其峰值集中在1.35~1.65。上述两点说明,BFD群落样本的p值在大于1的范围内趋近于1。根据
6 结 语
参考文献
Dominance and diversity in land plant communities: numerical relations of species express the importance of competition in community function and evolution
[J]. ,
Neutral theory and relative species abundance in ecology
[J]. ,
Unveiling the species-rank abundance distribution by generalizing the Good-Turing sample coverage theory
[J]. ,
Deep-sea coral forest
[J]. ,
深水珊瑚林
[J]. ,
Species abundance distribution pattern of typical grassland plant communities under different utilization patterns
[J]. ,
不同利用方式下典型草原植物群落物种多度分布格局
[J]. ,
Species abundance distribution models of Toona ciliata communities in Hubei Province, China
[J]. ,
Fractal analysis of community species abundance pattern
[J]. ,
群落物种多度格局的分形解析
[J]. ,
A meta-analysis of species-abundance distributions
[J]. ,
The relation between the number of species and the number of individuals in a random sample of an animal population
[J]. ,
Species diversity as a fractal property of biomass
[C]// NOVAK M, ed.
Research progress on species abundance relationship model of multi species aggregation
[J]. ,
多物种集合的种—多度关系模型研究进展
[J]. ,
Modeling macroscopic patterns in ecology
[J]. ,
A comment on Hubbell's zero-sum ecological drift model
[J]. ,
Species abundance distributions: moving beyond single prediction theories to integration within an ecological framework
[J]. ,
An extensive comparison of species-abundance distribution models
[J]. ,
Analyzing fractal property of species abundance distribution and diversity indexes
[J]. ,
Environmental controls on the geographic distribution of zooplankton diversity
[J]. ,
Research on the determinability of determinable coefficient in regression model
[J]. ,
回归模型可决系数的可决性研究
[J]. ,
Abundance inequality in freshwater communities has an ecological origin
[J]. ,
Characterizing species abundance distributions across taxa and ecosystems using a simple maximum entropy model
[J]. ,
Commonness and rarity in the marine biosphere
[J]. ,
Challenges and development trend of species distribution model
[J]. ,
物种分布模型面临的挑战与发展趋势
[J]. ,
Neutral theory and niche theory of community construction
[J]. ,
群落构建的中性理论和生态位理论
[J]. ,
On measures of entropy and information
[C]//NEYMAN J.
Diversity and evenness: a unifying notation and its consequences
[J]. ,
The impact of neutrality, niche differentiation and species input on diversity and abundance distributions
[J]. ,
Review: on the species abundance distribution in applied ecology and biodiversity management
[J]. ,
Testing species abundance models: a new bootstrap approach applied to Indo-Pacific coral reefs
[J]. ,
A unified model of species immigration, extinction and abundance on islands
[J]. ,
A new species abundance distribution model based on model combination
[J]. ,
/
〈 | 〉 |