Post-settlement processes and early life history dynamics of zebra mussels (Dreissena polymorpha) in the Hudson River estuary

Wood, Helen

Permalink

https://hdl.handle.net/2142/44337 Copy

Description

Title

Post-settlement processes and early life history dynamics of zebra mussels (Dreissena polymorpha) in the Hudson River estuary

Author(s)

Wood, Helen

Issue Date

2013-05-24T22:08:15Z

Director of Research (if dissertation) or Advisor (if thesis)

Schneider, Daniel W.

Doctoral Committee Chair(s)

Schneider, Daniel W.

Committee Member(s)

• Cáceres, Carla E.
• Brawn, Jeffrey D.
• Suarez, Andrew V.

Department of Study

School of Integrative Biology

Discipline

Biology

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Date of Ingest

2013-05-24T22:08:15Z

Keyword(s)

• zebra mussel
• early post-settlement mortality
• density dependence
• settlement
• recruitment
• early life history stages
• population dynamics.

Abstract

Although progress has been made, we are still a long way from being able to predict and understand the dynamics of demographically “open” aquatic populations. Even though we often have a good understanding of events and processes that occur to visible adult stages, particularly those of economically important species, in most cases we know very little or nothing about processes occurring to early life history stages. One of the main barriers to our understanding is a lack of empirical data on young life stages including larval supply, the number of larvae settling and the number of settlers that survive the early benthic period (until they become visible). The study of early life stages of most aquatic species poses inherent challenges. In a coastal marine environment for example, it is difficult to identify larvae to species level, track them across long distances in complex oceanic circulation patterns and identify and count small and/or cryptic settler stages. It is becoming clear that, for some populations, events that occur to young life stages can have the most impact on population dynamics. Thus it is essential that we develop our understanding of the early life history period in order to fully comprehend the dynamics of open populations. The zebra mussel, Dreissena polymorpha, in an estuarine habitat such as the Hudson River, provides an ideal system for examining the early life history period and the effect of early life history processes on population dynamics. My dissertation research focuses on the early post settlement period of which little is known in zebra mussel populations, and indeed, in most other benthic invertebrate populations. I investigated early post settlement mortality and estimated the proportion of settling larvae that survived the early benthic phase (two weeks after settlement). Estimates of mortality for the first few days varied between 83 - 92% and 38 - 88% depending on location and model used (day or age). However, by the sixth day of the study, survivorship was high (~ 1) at both sites. Consequently, although total mortality over the 14 day period was ~ 55% and 40% (depending on site), virtually all mortality occurred over the first few days only; mortality after the first few days post settlement was absent or very low. Thus the first few days after settlement are a distinct and critical phase for zebra mussels. One caveat to note, however, is that the spacing of the settlement plates could have impeded large predators and afforded larvae that settled onto experimental plates some protection from crushing. I conducted several experiments to investigate density dependence, a process crucial to population regulation. In one experiment, I investigated whether recruitment was dependent on the density of just settled larvae and whether the settler – recruit relationship changed over time. I manipulated the number of just settled larvae on artificial substrate plates, generating four different settler treatment densities (100%, 50%, 25% and 12.5% natural settlement). I tracked the density of survivors (i.e. recruits) for a year, sampling at various intervals throughout the period. In the early period (up to ten weeks after settlement), a poisson GLMM (treatment x time x site) indicated only a significant treatment effect (p < 0.05) with nearly all treatments significantly different from each other (the Bonferroni correction threshold of 0.05/ 6 = 0.0083 was used to correct for multiple testing). In the later period (up to a year after settlement) a poisson GLMM (treatment x time x site) indicated significant treatment, time and sites effects (p values < 0.05). Only the highest settler treatment density (100%) and the lowest settler treatment density (12.5%) were different (the Bonferroni correction threshold of 0.05/ 6 = 0.0083 was used to correct for multiple testing). Thus the effect of differential settlement on recruitment differed depending on cohort age. In the early benthic phase zebra mussel settlement determined recruitment, with increases in settler density leading to increases in recruit density (density independent response). I found no evidence of density dependence until about a year after zebra mussels had first settled when recruit densities at all four settler treatment densities started to converge. Interestingly there was a 10 fold increase in filtration rates in the later period (zebra mussels almost doubled in length during a three month period from May to July). It seems likely that intraspecific competition for limited food resources is the mechanism generating density dependence later in the benthic phase in zebra mussel populations in the Hudson River. In a second experiment, I investigated whether recruitment was affected by the density of adult conspecifics. I manipulated the density of adults on artificial substrate plates, generating four different adult treatment densities (75 - 90% coverage, 45 - 60% coverage, 15 - 30% coverage and 0% coverage) and allowed natural settlement onto the plates. After five months, I removed all plates from the water and estimated recruit density at each adult density. A GLM (negative binomial distribution and log link function) indicated significantly higher recruitment on plates with 0% adult coverage (p < 0.05). Adult conspecifics had a negative effect on recruitment with significantly higher recruitment when adults were absent compared to when adults were present (at any density). The main driver for lower recruitment in the presence of adults is likely crushing by large predators. In addition I investigated broad patterns of early life history dynamics at the site and river scale in order to provide further insight into processes affecting population dynamics. I estimated larval supply and settlement over the reproductive season (May to October) for five years (2002 to 2006) in order to uncover any temporal changes in important events (for example spawning / settlement peaks). In addition I estimated recruitment and survival of post recruits and combined data on early life history dynamics with adult dynamics. I found distinct differences between early life stage dynamics during the first three years of the study (2002 to 2004) and those occurring during the final two years of the study (2005 & 2006). In later years, larval densities were significantly lower (p < 0.05) and spawning events were reduced to one per year compared with two to four in earlier years. Settlement was significantly higher (p < 0.05) and peaked later in the season in later years. Recruits were significantly smaller in later years (p < 0.05) and size frequency distributions highly skewed to small sizes compared to normal distributions in earlier years. Such differences in early life stage dynamics correspond with a steep decline in annual survivorship rates of zebra mussels in the Hudson River (annual survivorship was zero in later years). The phenomenon of heavy adult mortality in late summer seems to be the main driver for changes in early life history dynamics seen in this study. In addition, I found inverse linear relationships between larval density and settler density and between settler density and aggregate adult filtration rate. Such phenomena may be important regulation mechanisms for zebra mussels and are likely due to intercohort cannibalism. My research has identified two likely regulation mechanisms in zebra mussel populations in the Hudson River, both acting on young (less than one year) life stages. Thus any attempt to understand the dynamics of zebra mussel populations and predict their abundance and spread must take into account the early life history period.

Graduation Semester

2013-05

Permalink

http://hdl.handle.net/2142/44337

Copyright and License Information

Copyright 2013 Helen Wood

Owning Collections