Staying ahead of water weed invasions


Aquatic systems are under threat due to the introduction of invasive exotic species such as water weeds. Modelling work by NIWA has provided new information on which water bodies may be at greatest risk.

The problem

The worst water weeds in New Zealand are able to tolerate a wide range of environmental conditions, and are able to invade diverse aquatic habitats including lakes, ponds and rivers.

As a result, it has been difficult to identify which water bodies are at the greatest risk of invasion, and to focus appropriate management initiatives.

Moreover, the cumulative records of weed invasion in lakes show the rate of spread is not decreasing, indicating that numerous other lakes are at risk.

The solution

Models of pest spread are mostly based on climate mapping and habitat suitability; this has limited use for predicting where a water weed is likely to invade in New Zealand in the short-term. NIWA took the alternative approach of modelling weed spread based on estimates of likely recreational use of lakes and the possibility of activities such as boating that are often responsible for the spread of weeds.

NIWA collated water weed records for 371 lakes, and built a GIS dataset for 3840 lakes that included:

  • measures of roading access (distance to the nearest highway, number of nearby roads)
  • the proximity and size of the local human population
  • lake area, as larger lakes are most attractive for recreational use
  • other environmental features.

The modelling had two aims: to identify the most important factors in the spread of weeds, and to model the probability of invasion for individual lakes.

The result

The model results clearly highlighted that roading accessibility and the proximity and size of human populations are linked to the probability of invasion of lakes by the worst water weeds in New Zealand.

Further, modelling was used to make forecasts of future weed spread. It was recognised that invaded lakes at the edge of the invasion front (where weed records are sparse) represented the greatest threat to adjacent un-invaded lakes. Therefore, the importance of these invaded lakes was increased (up weighted) in the model. Predictions indicated a large number of lakes remain at risk of invasion in the absence of management intervention. More importantly, the estimates of risk derived from the models are a potential tool to prioritize lakes for pest surveillance and protection efforts. For example, estimates of risk are being used together with other factors to develop strategic priorities for Northland Region's lakes.

On-going research aims to integrate these estimates of pest invasion risk within a GIS framework with information on the actual distribution of weeds, as well as the importance of a lake both regionally and nationally.

Links to reports/papers

Compton T. J.; de Winton, M.; Leathwick J. R.; Wadhwa, S. 2012. Predicting spread of invasive macrophytes in New Zealand lakes using indirect measures of human accessibility. Freshwater Biology, 57: 938–948, DOI: 10.1111/j.1365-2427.2012.02754.x. 

NIWA Contacts

placeholder image
Freshwater Ecologist
Page last updated: 
26 February 2014
Left: Surveillance for weeds should be targeted at lakes with high invasion risk. Credit: John Clayton. Right: Management response to a high weed invasion threat may be public education and information. Credit: Rohan Wells
Probability of weed invasion based on measures of human usage for lakes, with circles around those lakes that are confirmed to have the weed Ceratophyllum demersum.