Chapter One
THE MODEL AS MONSTER
One of the most enduring popular images of science is that of the cinematic mad scientist, scouring a graveyard for body parts from which to build a monster. Dr. Frankenstein is faced with two formidable tasks. First, he and his hunchbacked assistant must gather all of the pieces needed to build a living monster (usually while evading the local constabulary). Even more daunting, he then must find some way to animate the collection of unliving body parts-he must make the monster walk.
Dr. Frankenstein might seem to be a strange role model for scientists, but his method is perhaps a compelling model of reductionist science. In reductionist science, scientists study the various parts of a larger scientific problem or system in the hope that once they understand how all of the pieces work they will be able to put them together and achieve a mechanistic understanding of the whole. Although reductionist approaches are used widely and successfully in many branches of science, including ecology, some critics have claimed that ecological systems are so complex that the reductionist goal of constructing a satisfactory understanding of complex ecological systems from its parts is so difficult so as to be practically unachievable, as well as philosophically unsound (e.g., Peters 1991, Rigler and Peters 1995, Bormann 2005).
In some ways, the task of reductionist science is even more difficult than Dr. Frankenstein''s. Most often, a scientific field is not coordinated by a single person-it is as if we have dozens of hunchbacks scouring the countryside for body parts, but no scientist (mad or otherwise) to coordinate the monster-building. Further, unlike Dr. Frankenstein, we don''t usually have an explicit plan for animating the parts into a working theory. Although most of ecological science is reductionist in character, scientists rarely explicitly assess whether a reductionist approach to a particular ecological problem is feasible: Can we collect all the parts needed to build a mechanistic understanding of the problem? Can we animate them into a working whole? If these are feasible problems, what parts to we need, and how do we best integrate them into a working theory? Here, I explicitly consider what would be required to build a mechanistic understanding of one specific ecological problem: that of predicting the distribution and abundance of freshwater unionoid mussels (Fig. 1); I then assess the feasibility of this enterprise.
This exercise could be done with any group of organisms. I chose unionoid mussels as an example because they are important in conservation and freshwater ecology-thus, there is a specific interest in understanding the factors that limit the distribution and abundance of this particular group of organisms; and because there is no reason to think that unionoids are unrepresentative of the many other groups of organisms that ecologists try to understand or manage; and also because I enjoy working with the group and am familiar with the literature on these animals.
Our current understanding of unionoid ecology is inadequate in three important ways: (1) studies of single factors have not yet led to an adequate understanding of the importance of any single factor in nature, even for those factors that have received much study (e.g., habitat, fish hosts); (2) some factors (e.g., food) have not even been seriously considered as controls on natural populations; and (3) the simultaneous influence of multiple controlling factors, although probably important in nature, has not been seriously considered by unionoid ecologists. That is, both Igor and Dr. Frankenstein have work to do. These problems are not unique to unionoid ecology-I do not believe that there is anything especially pathological about unionoid ecology-but are common to many areas in ecology.
Before proceeding, I want to be clear about what, in my view, would constitute a satisfactory theory of mussel distribution and abundance. Ecology is often said to be concerned with predicting the distribution and abundance of organisms, and I take that goal literally. That is, I think we should strive to make quantitative predictions about the probability of occurrence or abundance of individual species based on environmental and biological variables. An adequate theory should be able to do this well; that is, the theory predictions should be close to actual values of these variables in the field.
Ecologists sometimes seem content to reach alternative goals. For instance, I have the impression that some ecologists would be satisfied if we could just list the factors that determine distribution and abundance of organisms. That is, instead of producing an equation of the form
Abundance = 0.15 + (0.5 x food - 1.66 x [predation).sup.-|temperature-20|]
they would be happy to say that
Abundance = f (food, predation, temperature)
Alternatively, much of contemporary ecology is focused on understanding individual processes (e.g., predation, disturbance, competition). This focus has been very fruitful in making generalizations across study systems, and I do not question its value. Nevertheless, both science and society often have compelling needs to know the actual distribution or abundance of individual species. Species, and the ecosystems in which they live, are subject to the integrated, simultaneous influences of multiple processes (cf. Greene 2005). Isolated studies of individual processes or a vague outline of the equation that describes the control of distribution and abundance will not meet the intellectual or practical goals that ecologists and society have set for our discipline. Therefore, I insist that, whatever the value of reaching any other goals, a literal, quantitative prediction of the distribution and abundance of individual species is properly a central goal of ecology.
The objectives of this book are to assess the feasibility of producing such a theory to predict the distribution and abundance of unionoid mussels and, to the extent possible, rough in parts of that theory. I see three possible conclusions of this assessment: (1) we can produce a mechanistic theory (following the reductionist model) that adequately predicts the distribution and abundance of unionoid mussels; (2) we can''t produce an adequate mechanistic theory, but we can devise some sort of acceptable alternative that provides useful predictions of unionoid distribution and abundance; or (3) we can''t produce adequate predictions by any means and must abandon the problem as scientifically intractable.
I will begin by considering individually the pieces that I think are probably necessary for a working theory of unionoid distribution and abundance, reviewing what we know and what we might ultimately need to know about each part. There are many ways in which to divide up and define these parts, but I will use a five part structure that seems natural to me. The five pieces that I think might be needed to predict unionoid distribution and abundance are dispersal, habitat, fish hosts, food, and predation. Except for the piece on fish hosts, which is needed to account for the unionoids'' peculiar parasitic life history, these pieces probably are needed to explain the distribution and abundance of any kind of organism. After I review each of these pieces, I will discuss various ways in which they might be put together. Finally, I will assess our prospects for actually collecting all of the necessary pieces and integrating them into a working theory, a central goal of unionoid ecology.
Chapter Two
THE CASE OF PEARLY MUSSELS
IMPORTANCE OF PEARLY MUSSELS
Pearly mussels of the superfamily Unionoidea (including the families Unionidae, Margaritiferidae, and Hyriidae) are common and widespread in rivers, streams, lakes, and ponds around the world, living on all continents except Antarctica. They can form locally dense populations of >100 animals/m2 (Fig. 42), and often vastly outweigh other animals in benthic communities, reaching biomasses (not including their shells) sometimes exceeding 100 g dry mass/[m.sup.2] (e.g., Hanson et al. 1988, Strayer et al. 1994). Although their roles in freshwater ecosystems have not been fully investigated (Vaughn and Hakenkamp 2001), they can be important suspension-feeders, influencing water chemistry and clarity, and the amount and kind of suspended particles in the water (e.g., Welker and Walz 1998, Vaughn and Hakenkamp 2001). Shell production by unionoids can be of the same order of magnitude as wood production by trees in a temperate forest (Gutierrez et al. 2003, Strayer and Malcom 2007b), providing important physical structure to other organisms (e.g., Chatelain and Chabot 1983, Beckett et al. 1996, Vaughn et al. 2002, Gutierrez et al. 2003). Waste products from mussels can enhance local populations of algae (Vaughn et al. 2007) and macroinvertebrates (Vaughn and Spooner 2006). Thus, the effects of pearly mussels on freshwater ecosystems can be important and pervasive.
Pearly mussels also are economically important to humans. Pearly mussels have been harvested as a source of pearls, mother-of-pearl, and human food since prehistoric times (e.g., Kunz 1898, Morrison 1942, Claassen 1994, Ziuganov et al. 1994, Anthony and Downing 2001, Walker et al. 2001). Freshwater pearl fisheries were one of the reasons that Julius Caesar invaded Britain (Ziuganov et al. 1994), and were one of the most important sources of new capital in 19th century American rural economies (Claassen 1994). Most of these fisheries have disappeared because of overharvest, habitat destruction, or pollution, or because the products they provided have been replaced by other materials (e.g., we now make "mother-of-pearl" buttons and ornaments out of plastic), but regionally important fisheries for shell and pearls still exist (Bowen et al. 1994, Claassen 1994, Neves 1999, Beasley 2001).
EVOLUTION AND CLASSIFICATION OF PEARLY MUSSELS
The major groups of unionoid mussels, their geographic distributions, and the approximate number of species that each contains are now well known (Table 1), but evolutionary relationships among both higher-level taxa and species still are incompletely understood. Traditional classifications based on characters of the shell and soft anatomy have largely been invalidated by molecular studies (e.g., Lydeard et al. 1996, Hoeh et al. 2001, Huff et al. 2004, Campbell et al. 2005). However, molecular data have not yet been collected on enough species to provide a clear picture of evolutionary relationships in the Unionoida.
The order Unionoida usually has been divided into two superfamilies: the Etherioidea, whose larva is a lasidium, and which live in tropical fresh waters around the world; and the Unionoidea, whose larva is a glochidium. The Etherioidea are relatively poorly known ecologically, and will not be dealt with further here. The superfamily Unionoidea contains three families: the Unionidae, by far the largest and most widespread family in the order, the Margaritiferidae, and the Hyriidae.
Members of the Unionidae occur on all of the continents except for Antarctica. Several recent studies (e.g., Lydeard et al. 1996, Hoeh et al. 2001, Campbell et al. 2005) have succeeded in defining several more or less well defined groups of genera ("tribes") within the Unionidae (Table 2), although the placement of all unionid genera in these tribes is not yet known. The tribe Lampsilini is usually thought to be the most derived evolutionarily (Campbell et al. 2005).
The Margaritiferidae usually are regarded as primitive relatives of the Unionidae. The family is small and restricted to the Northern Hemisphere, but margaritiferids often are extremely abundant where they occur (cf. Fig. 26). Many margaritiferids use salmonids as hosts. Because of their abundance and the current peril of many of the species, the margaritiferids are perhaps the best-studied of the unionoids (e.g., Ziuganov et al. 1994, Bauer and W?chtler 2001, Huff et al. 2004, and references cited therein).
The evolutionary position of the hyriids continues to be unclear. These animals are common and widely distributed in South America, Australia, New Zealand, and the Pacific Islands. They were originally placed with the etherioids because of their Gondwanaland distribution and anatomical characteristics, but then moved to the Unionoidea when the peculiar lasidium larva of other etherioids was discovered (hyriids have glochidia). More recent analyses have again united the hyriids with the etherioids (Graf 2000, Graf and Cummings 2006), or suggested that they occupy a basal position in the Unionoida (Hoeh et al. 2001, Walker et al. 2006). Like the Unionidae, the hyriids have been divided into several subfamilies (Table 2). Although this subfamilial classification hasn''t been fully tested with molecular methods, it has received partial support (Graf and ? Foighil 2000).
The evolutionary relationships among the major groups of the Unionoida still are unclear (see Graf and Cummings 2006 for a critical discussion). In particular, the position of the Hyriidae is unclear (indeed, it is possible that they don''t even belong in the superfamily Unionoidea despite possession of a glochidium larva), and the relationships and memberships of the subfamilial groups of the Unionidae and Hyriidae remain to be fully defined and tested. Further molecular data and analyses should help us understand the evolutionary relationships, geographic origins and spread, and the development of biological traits of the unionoid mussels.
The genus- and species-level taxonomy of the unionoids also is in a state of flux. Many familiar genera now appear to be polyphyletic (Baker et al. 2004, Huff et al. 2004, Campbell et al. 2005) and will need to be redefined. At the species or subspecies level, cryptic speciation and geographic differentiation both appear to be common (e.g., Davis 1983, 1984, King et al., 1999, Baker et al. 2004, Jones et al. 2006, Serb 2006), so that traditional views about the limits and internal phylogeographic structure of some species will have to be rethought. This fine-scale differentiation has profound implications for the conservation and management of rare pearly mussels.
The advent of molecular methods and statistical analyses has led to very rapid progress in the areas of unionoid evolution and classification over the past 10-20 years. Although the field is very much in flux, I expect that this rapid progress will continue, and that many of the important questions about unionoid evolution will be satisfactorily resolved in the next few years.
BIOLOGY OF PEARLY MUSSELS
Sexes are separate in most unionoid species, although a few species are normally hermaphroditic (van der Schalie 1970, Kat 1983). Hermaphroditism occurs occasionally in many other species, and apparently can be induced by low population density (Bauer 1987a, Walker et al. 2001). Sperm is shed into the water, taken up by females, and fertilizes the eggs held in the females'' gills. The fertilized eggs develop into specialized larvae (glochidia) that are held in the females'' gills for weeks to months. The developed larvae are obligate, more or less species-specific parasites of fish (known host relationships were complied by Cummings and Watters 2005). Although it was once believed that glochidia were simply broadcast into the water to await a chance contact with the proper host, it has become increasingly clear that unionoids use a wide range of sophisticated (sometimes almost unbelievable) methods to get their larvae onto hosts: the females have elaborate moving lures (Fig. 1; Kraemer 1970, Haag and Warren 1997, Haag et al. 1999, Corey et al. 2006), or the glochidia are packaged to resemble fish food (Haag et al. 1995, Hartfield and Hartfield 1996, Watters 1999, 2002, Haag and Warren 2003). Some mussels even catch and hold their fish hosts while infesting them with glochidia (Barnhart 2006)! It is worth noting that at least one species bypasses the parasitic stage altogether (Barfield and Watters 1998, Lellis and King 1998, Corey 2003), simply releasing small juveniles onto the sediments, and this short-circuit may occur in other species (Lefevre and Curtis 1911, Howard 1914). Once the larvae attach to the host, they encyst and transform into juveniles. This parasitic period lasts for several days to several months (Coker et al. 1921, Young and Williams 1984b, Watters and O''Dee 1999), and is the main opportunity for the mussel to disperse. The juvenile mussel falls to the sediment after transformation is complete. Not much is known about the juvenile phase, but most juveniles apparently live an interstitial life, buried in the sediments (Yeager et al. 1994, Sparks and Strayer 1998, Smith et al. 2000). They may deposit-feed as an alternative or supplement to suspension-feeding (Yeager et al. 1994). The juvenile phase lasts for one to a few years (Coker et al. 1921, Jirka and Neves 1992, Haag and Staton 2003), after which sexual maturity is attained and the adults are more or less epifaunal (but see Smith et al. 2000, 2001, Schwalb and Pusch 2007), living at or near the sediment surface. Adults probably are mainly suspensionfeeders (this will be discussed in more detail below), and may live for one to several decades (e.g., Bauer 1992, Haag and Staton 2003, Howard and Cuffey 2006).
(Continues...)
Excerpted from FRESHWATER MUSSEL ECOLOGYby David L. Strayer Copyright © 2008 by The Regents of the University of California. Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
