Characteristics of mangrove swamps managed for mosquito control in eastern Florida, USA: a re-examination

Abstract

A recent study by Middleton et al. (2008; Mar Ecol Prog Ser 371:117-129) characterizing mangrove swamps impounded for mosquito control in Florida has a number of problems that weaken some of the resulting conclusions. We examine some of these issues, which we divide into (1) errors of fact, (2) methodological problems, (3) unwarranted conclusions. Errors of fact include incorrect historical assertions, incorrect site descriptions, and inaccurate statements on the general effects of impoundments. Methodological problems include inadequate sampling for the stated purpose, in particular using results of one-time sampling of very small marsh areas to characterize entire marshes and groups of marshes; defining inadequate controls; use of heterogeneous experimental treatment groups, including highly atypical and unrepresentative sites within experimental treatments; selective sampling; and others. Unwarranted conclusions include the characterization and comparison of marshes and management strategies based upon unrepresentative sampling, and the attribution of differences in habitat function as a result of management strategy without any experimental evidence to suggest such an effect. We make suggestions for mitigating some of these issues.

Perspectives on mosquito impoundments in eastern Florida, USA: Reply to Rey et al. (2009)

“…The sampling design used by Middleton et al (2008) involved streamlined sampling of replicate sites in RIM, breached-RIM, and control categories (total of 20 sites), but with limited sub-sampling within sites. Rey et al (2009) criticize this approach and suggest more intensive sampling at fewer sites. However, such an approach would have reduced the statistical inference space, because type of impoundment is the source of variation that requires replication and allows inference to other sites of the same type.…”

“…1). Rey et al (2009) suggest that the only appropriate choice of controls is high marsh habitat, which characterized managed sites prior to impoundment. We disagree.…”

“…Site information was based on historical photography, published literature, and extensive communications with site managers (including D. Carlson, co-author of the Comment, and his staff, who accompanied Middleton and co-workers to field sites). This information, which Rey et al (2009) portray as being incorrectly attributed, was ascribed generally to managers during editing with no intent to misquote.…”

“…Choices made during ecological studies are partly a reflection of the difficulties inherent in sampling at various landscape scales in complex systems. Rey et al (2009) list 'errors of fact' and 'methodological problems', which are minor or unfounded, and which do not invalidate the findings of Middleton et al (2008). Most issues raised by Rey et al (2009) were in fact carefully considered during study design.…”

“…Rey et al (2009) list 'errors of fact' and 'methodological problems', which are minor or unfounded, and which do not invalidate the findings of Middleton et al (2008). Most issues raised by Rey et al (2009) were in fact carefully considered during study design. Site information was based on historical photography, published literature, and extensive communications with site managers (including D. Carlson, co-author of the Comment, and his staff, who accompanied Middleton and co-workers to field sites).…”