The scale of the inventory is striking. Between 2023 and 2025, the researchers conducted quarterly field surveys across every administrative district of Shenzhen, from the forests of Wutong Mountain and Tanglang Mountain to the coastal zones of the Dapeng Peninsula, roadside green belts, urban parks, residential communities, and river shorelines. For each species encountered, they recorded geographic coordinates, habitat type, population size, and associated species, supplementing the fieldwork with herbarium specimen records, historical literature, and online databases. The final tally of 335 species, spanning 60 families and 206 genera, represents a dramatic increase from the 102 alien plants recorded in the city in 2004, 179 in 2016, 214 in 2019, and 258 in 2023.

Six plant families dominate the alien flora: Asteraceae, Fabaceae, Poaceae, Solanaceae, Amaranthaceae, and Euphorbiaceae together account for 178 species, or 53.1 percent of the total. The Asteraceae, or daisy family, alone contributes 54 taxa, having exploded from just 20 species two decades ago. The researchers attribute this expansion to distinctive biological traits: Asteraceae plants produce abundant, lightweight seeds equipped with pappi or spines that aid wind and animal dispersal, and some species release allelopathic chemicals that suppress competing vegetation. Legumes in the Fabaceae benefit from nitrogen fixation, which enriches soils and gives them a competitive edge, while grasses in the Poaceae combine prolific seed production with remarkable stress tolerance.

Of the 335 alien species, 144, or 42.9 percent, are classified as invasive, meaning they produce large numbers of reproductive offspring, can spread considerable distances, and cause measurable ecological or economic harm. Most of the alien flora consists of herbs, which make up 72.8 percent of species, followed by shrubs at 15.0 percent, climbers at 7.2 percent, and trees at 5.0 percent. High-risk species cluster in disturbed habitats such as roadsides, abandoned construction sites, and urban green spaces, with concentrations in the city’s southern and western districts. The dominant introduction pathway is intentional horticultural trade, responsible for 62.0 percent of alien species, followed by accidental seed contamination at 32.0 percent and natural dispersal from neighboring regions at 6.0 percent.

Shenzhen’s vulnerability is no accident. The city sits in a subtropical monsoon climate zone with an average annual temperature of 23.3 degrees Celsius and nearly two meters of rainfall per year, conditions ideal for plant growth and, unfortunately, for invasion. Decades of breakneck urbanization have transformed the landscape, yet protected areas like the Dapeng Peninsula Nature Reserve preserve pockets of high biodiversity alongside intense development. This unusual juxtaposition of high-intensity urban growth and rich native flora, combined with Shenzhen’s role as a port city and its appetite for ornamental landscaping plants, has made it one of the regions in China most severely affected by alien plants.

To manage this growing threat, the team constructed a two-tier evaluation architecture. The first tier is a pre-assessment procedure, structured as a decision tree, that rapidly screens out species with extremely high invasion risk and introduction proposals from unqualified entities, such as nurseries lacking proper quarantine credentials. Species that pass this initial filter move to the second tier: a comprehensive, hierarchical risk assessment system built around five primary criteria that mirror the stages of invasion. These are introduction risk, colonization risk, spread risk, potential harm and impact, and control risk. The system comes in two versions, one for species not yet introduced to Shenzhen, with 11 secondary indicators and 41 tertiary operational indicators, and one for species already present, which adds a growth and escape status indicator, bringing the total to 12 secondary and 44 tertiary indicators.

The weighting of every indicator was determined through the Analytic Hierarchy Process, a structured decision-making method in which a panel of 12 experts, including ecologists, botanists, agricultural scientists, and nature conservation scholars, independently constructed pairwise comparison matrices using the 1-to-9 Saaty scale. The geometric mean of their judgments produced collective matrices, and consistency ratios below 0.1 confirmed that the expert judgments were internally coherent. The final species score is simply the sum of all weighted indicator scores, allowing any plant to be assigned a quantitative risk value. Thresholds separating high, medium, and low risk were derived from natural breaks in the score distributions rather than imported from other regions, one of the study’s key methodological innovations.

Validation was rigorous. The researchers selected 50 alien plant species documented in Guangdong Province, deliberately excluding species used to calibrate the system to avoid circular reasoning. Forty-three of these are already established in Shenzhen, while seven occur in surrounding areas and represent potential future invaders. The test set spanned the full spectrum of impact, from highly destructive species such as Ageratum conyzoides, Lantana camara, and Chromolaena odorata, through moderate weeds like Veronica persica and Xanthium strumarium, to benign ornamentals including Bougainvillea glabra and Oxalis debilis. When the system’s risk classifications were compared against reference categories from documented invasion records, the agreement was perfect: 100 percent accuracy and a Cohen’s kappa coefficient of 1.00. A sensitivity analysis in which primary criterion weights were varied by plus or minus 20 percent kept accuracy above 92 percent and kappa above 0.90, demonstrating that the classifications are robust to reasonable uncertainty in the weighting.

The practical output is a three-level management framework. Species scoring in the highest band are prohibited from introduction or, if already present, flagged for strengthened monitoring and active control; the top scorers included Amaranthus spinosus, Spermacoce alata, Mimosa pudica, Alternanthera philoxeroides, and Praxelis clematidea, all classified as malicious invaders in national databases. Medium-risk species, about 22 percent of those assessed, may be introduced only with caution and preventive monitoring, while low-risk species, roughly a third of the test set, can be introduced under registration and routine surveillance. The authors emphasize that risk assessment is a dynamic process: as climate changes and new monitoring data accumulate, species must be periodically re-evaluated, and plants whose key indicators shift substantially require reassessment.

Beyond Shenzhen, the framework’s construction logic is designed to be transferable. The researchers argue that any city in tropical or subtropical regions can adapt the system by recalibrating parameters to local conditions, addressing a well-known weakness of national-scale assessment tools that often fail to capture regional ecological realities. Future work, they suggest, should integrate remote sensing and artificial intelligence to build dynamic invasion monitoring networks, expand long-term fixed-site observations, and engage the public through online platforms in reporting and managing alien plants. For now, Shenzhen has something it has never had before: a scientifically validated, locally calibrated yardstick for deciding which plants belong in its future and which must be kept out.

Subject of Research: Development and validation of a risk assessment system for alien plants in Shenzhen, China

Article Title: The investigation and construction of a risk assessment system for alien plants in Shenzhen

Article References: Liao, Y., Meng, L., Yang, T., Liu, Y., & Cheng, Y. (2026). The investigation and construction of a risk assessment system for alien plants in Shenzhen. Frontiers in Conservation Science, 7, Article 1901250. https://doi.org/10.3389/fcosc.2026.1901250

Image Credits: AI Generated

DOI: 10.3389/fcosc.2026.1901250

Keywords: alien plants, invasive species, Shenzhen, risk assessment, biodiversity, urban ecology, Analytic Hierarchy Process, plant invasion, Guangdong, conservation, biosecurity, Asteraceae