Speculation but no data: a response to Hoagstrom's drift versus retention perspective

Abstract

Hoagstrom has challenged the interpretations of Wilde and Urbanczyk on the basis of their citation of an existing hypothesis, which suggests that ova and larvae of Great Plains pelagic-broadcast spawning cyprinids require current for success recruitment. Hoagstrom characterizes the hypothesis as ‘speculative’ and without empirical support, but his particular objections to the hypothesis are limited, at best. Hoagstrom advances his own speculative hypothesis, one that is without empirical support, to replace the existing hypothesis, committing the same lapse with which he finds fault

Keywords:

• pelagic-broadcast spawning
• Great Plains
• river fragmentation
• extinction
• fish conservation

Wilde and Urbanczyk (2013) presented an analysis of the relationship between river fragment length and persistence of two imperiled species of fish, sharpnose shiner Notropis oxyrhychus and smalleye shiner Notropis buccula, and concluded that persistence of these species appeared to require long reaches of an unfragmented river. Hoagstrom (2014) questions this conclusion and its implications, and uses the occasion of his comment to present an alternative hypothesis for which he lacks empirical support. Herein, we respond to Hoagstrom's comments.

Cross et al. (1985) described the decreasing distribution and abundance of the Arkansas River shiner Notropis girardi in Kansas and noted that, although the species had disappeared upstream from several impoundments, it was still present in downstream areas (i.e., longer downstream river fragments). Based on their observations, and those of Moore (1944), Cross et al. (1985) suggested that the Arkansas River shiner spawned semi-buoyant ova that required flowing water to remain afloat and successfully develop, which resulted in the downstream displacement of ova and young. They further hypothesized that this, in turn, necessitated a subsequent, upstream movement by adult fish to replenish populations in upstream areas. Fausch and Bestgen (1997) provide a detailed description of Cross et al.'s (1985) ‘drift-migration’ hypothesis. Based on the structure of his comment, we surmise that Hoagstrom (2014) takes issue with this hypothesis, which is now fairly widely accepted as a general life-history model for pelagic-spawning cyprinids (e.g., Winston et al. 1991; Fausch & Bestgen 1997; Luttrell et al. 1999; Bonner 2000).

Hoagstrom's comment takes three tacks: (1) criticize Moore's (1944) hypothesis, which presents a means and biological rationale for downstream displacement of ova and larvae; (2) argue that recruitment only occurs when ova and larvae are retained in the area in which they are spawned; and (3) discount the importance of long reaches of an unfragmented river, required by the ‘drift-migration’ hypothesis, for persistence of pelagic-spawning cyprinids.

Many cyprinids native to larger streams and rivers of the US Great Plains are known or are believed to be members of a reproductive guild that broadcast spawns semi-buoyant ova into the current (Moore 1944; Cross et al. 1985; Platania & Altenbach 1998). Moore (1944) first suggested this behavior and hypothesized that successful reproduction required current to keep ova and early larvae suspended in the water column in order to prevent abrasion, burial and suffocation, or other ill-adventures at the river bottom. Hoagstrom (2014) provides three criticisms of Moore's hypothesis. First, he and others, including Moore (1944), have hatched ova in static systems. We have too. However, static systems have neither the suspended sediment loads nor the potential for burial and suffocation that occurs in the field, so the relevance of observations from these systems is questionable. Abrasion or burial and suffocation, not suspension in the water column per se, was Moore's concern. Second, Hoagstrom argues that physical properties of sediments and ova preclude the possibility of ova being buried in flowing water; this is consistent with Moore's hypothesis, rather refuting it. Hoagstrom suggests that in the absence of current ‘pelagic eggs must be deposited in a similar manner and develop on the surface of the substrate, as in a finger bowl.’ Here, Hoagstrom offers speculation, but not data, to counter Moore's ‘speculative’ hypothesis for which results of Durham and Wilde (2006, 2009) do provide support. Third, Hoagstrom claims larval fish is not vulnerable to burial and suffocation because ‘Moore (1944) speculated that the swim-up behavior of protolarvae protects them from burial.’ Thus, Hoagstrom cites as evidence for his position the very hypothesis he criticizes. Hoagstrom's criticisms of Moore's hypothesis are not compelling.

Hoagstrom (2014) next builds, from the literature, a speculative hypothesis he refers to as the ‘local’ hypothesis. This posits that persistence of pelagic-spawning cyprinids is the result of developing ova and larvae being deposited in suitable nursery habitats located in the vicinity of spawning adults, and that ova and larvae displaced to downstream areas are lost to the system. Hoagstrom offers no empirical support for his hypothesis, a criticism he levels at Moore's (1944) hypothesis. Hoagstrom's hypothesis is inconsistent with the observations of Dudley and Platania (2007) and Durham and Wilde (2008), among others. His ‘local’ hypothesis is at odds with our observations on the Brazos and Canadian rivers, in which ova and larvae are displaced downstream and then move upstream as they mature. The very clearest evidence for this is presented by Bonner (2000).

Finally, Hoagstrom (2014) argues that long, unfragmented river reaches are not necessary for maintenance of pelagic-spawning cyprinids. In doing so, he eschews a large body of empirical evidence (Dudley & Platania 2007; Perkin & Gido 2011; Wilde & Urbanczyk 2013; Worthington et al. 2014) that relates persistence of 16 species of known or suspected pelagic-spawning cyprinids in at least 25 rivers to river fragment length. Instead, Hoagstrom selects a small number of samples (N = 4) from one data set (Dudley & Platania 2007) as support for his local hypothesis because, ‘…shorter fragments sometimes sustain remnant populations.’

Hoagstrom (2014) concludes, ‘With so few occupied river fragments remaining, it would be unfortunate to miss research or conservation opportunities in any river fragment.’ We agree, but we are not optimistic that much success will be observed in such reaches (e.g., Winston et al. 1991; Wilde & Ostrand 1999). Others may be more optimistic and that is fair enough. In the end, we believe that any success will require a keen understanding and manipulation of the interrelationship between discharge and river fragment length described by Dudley and Platania (2007) and Wilde and Urbanczyk (2013).

Finally, just as we would not dismiss possible research and conservation opportunities in smaller river reaches, we believe it would be, at best, imprudent to follow Hoagstrom's (2014) lead and discount the conservation importance of longer river fragments based solely on conjecture and speculation.

Acknowledgements

We thank J. Perkin for helpful discussions and comments on the manuscript.