Previous cultural evolutionary analyses have exploited the use of neutral models, or random copying models, analogous to neutral genetic drift, as null models to distinguish the unbiased copying and biased transmission of neutral cultural traits, but the dynamics and underlying mechanism of non-neutral cultural traits have not yet been explored. Adopting an ecological perspective, we define ecological and economic concepts as non-neutral cultural traits constitute a complex ecosystem of meaning representing sustainability ideas. We analyse the frequency distribution, turnover and innovation rates of nearly 4000 concepts appearing in People’s Daily from 1946 to 2015. The results reveal that, ecological concepts show stronger tendency towards random copying than economic concepts. The popular ecological and economic concepts may experience conformist bias and anti-conformist bias respectively. Semantic and pragmatic differences of popular concepts affect the variability which indicate stronger cultural drift in the ontological domain. The frequency change in popular concepts also suggests positive selection due to political strategy and governmental policy.

The data is retrieved from BCC corpus(BCC语料库 (blcu.edu.cn)), using a Boolean query of "生态+N", "N+生态", "经济+N" and "N+经济" (means "Ecological+Noun", "Noun+Ecology"，"Economic+Noun" and "Noun+Economy" in English). In Chinese, the character has no change in writing forms even they serve for different purpose, and in our case, as  head noun or modifier. A total list of different combination of concepts around "ecology" and "economy" is derived from the corpus. The corpus provide diachronic frequency of each concept which can be used in the further analysis of frequency distribution and turnover rate.

The BCC corpus public webpage doesn't include all concepts or phrases number of every year. In cultural evolution, a consideration of the population size is necessary. For that reason, this research also acquired a full list of yearly phrases number directly from the corpus building team to support the analysis. Any reuse of the data (file name "yearly_phrases_number.csv") should cite this, "Xun E, Rao G, Xiao X, Zang J. Construction of BCC corpus in the context of big data. Corpus Linguistics. 2016;3(1):93-109."