Quantifying the population response of invasive water hyacinth, Eichhornia crassipes, to biological control and winter weather in Louisiana, USA

Abstract

Water hyacinth, Eichhornia crassipes, is an invasive, tropical, aquatic plant that has caused significant environmental and economic damage since its establishment in Louisiana, USA, in 1884. Both invasion control programs and freezing temperatures are known to negatively affect water hyacinth populations; however, the combined impact of these factors on water hyacinth population dynamics has not yet been quantified, thereby limiting the ability to isolate the effectiveness of biocontrol and other types of control under variable weather conditions. We built a seasonal logistic population model that included timevarying intrinsic growth and overwinter mortality parameters which were estimated by fitting the model to vegetation survey data. We estimated that annual overwinter mortality rates declined from a peak of 71 % in 1977 to the time series low of 11 % in the winter of 1992, followed by an average of 28 % per year from 1993 to 2013. After accounting for the magnitude and trend of overwinter dieback events, our model predicted that the intrinsic growth rate of the Louisiana water hyacinth population declined by 84 % between 1976 and 2013. Despite higher average winter temperatures in recent decades, the population has not rebounded. Our study reveals the dramatic effectiveness of Louisiana's biological control program to successfully suppress water hyacinth invasion.

Evidence-based economic analysis demonstrates that ecosystem service benefits of water hyacinth management greatly exceed research and control costs

“…Data gaps can be filled with after-the-fact mapping using remote sensing data ( Albright, Moorhouse & McNabb, 2004 ) and by using modeling to separate causes from effects. We demonstrated the latter approach by using the spring and fall water hyacinth coverage observations and historic weather records to distinguish effects of overwinter mortality from treatment effects, thereby refuting the hypothesis that cold winters that coincided with biological control release were solely responsible for plant declines ( Nesslage et al, 2016 ).…”

“…3 ) without treatment. The dramatic decline in coverage with control appeared to be supported by an 84% decline in growth rates over the study period (1976–2003) ( Nesslage et al, 2016 ). The spatial results reflected the north–south gradient in habitat suitability, since abundance was greater in the more suitable (warmer) southern areas ( Fig.…”

“…Although biological control can take time to show effectiveness and does not usually eradicate the target species ( Schooler, McEvoy & Coombs, 2004 ), it has potentially long-lasting, self-propagating, and self-sustaining benefits that offset the initial upfront investment costs. As evidence from this program, releases of agents in Louisiana were discontinued in the 1980’s, yet overall suppression has been maintained ( Nesslage et al, 2016 ). (Recent releases of M. scutellaris in Louisiana have not led to confirmed establishment.)…”

“…Water hyacinth distribution data collected in the spring and fall each year (usually April and November) from 1975 to 2013 ( Louisiana Department of Wildlife and Fisheries, 2014 ) enabled us to model plant growth rate response to biological control. We fit predicted to observed cover data using a logistic growth model that controlled for winter severity by including a variable of days with minimum temperature at or below freezing ( Nesslage et al, 2016 ). The model fit was substantially improved by incorporating a time-varying intrinsic growth rate for water hyacinth such that the growth rate declined through time.…”

Mottled Water Hyacinth Weevil Neochetina eichhorniae Warner (Insecta: Coleoptera: Curculionidae)

“…Since then, the insect has been introduced in more than three dozen countries worldwide (Winston et al 2014). Post-introduction studies indicate the insect suppresses the growth of water hyacinth, significantly reducing biomass and flower production (Grodowitz et al 1991;Center et al 1999;Tipping et al 2014;Nesslage et al 2016;Tipping et al 2017).…”