Europe PMC

This website requires cookies, and the limited processing of your personal data in order to function. By using the site you are agreeing to this as outlined in our privacy notice and cookie policy.

Have superfetation and matrotrophy facilitated the evolution of larger offspring in poeciliid fishes?

Author information

Affiliations

  • 1. Posgrado en Ciencias Biológicas, Facultad de Ciencias, Universidad Nacional Autónoma de México, Ciudad Universitaria, Distrito Federal 04510, México.
      Authors
    • Olivera-Tlahuel C 1
    (1 author)
  • 2. Department of Biology, Indiana University, Bloomington, IN 47405, USA.
      Authors
    • Ossip-Klein AG 2
    (1 author)
  • 3. Colección Nacional de Peces, Instituto de Biología, Universidad Nacional Autónoma de México, Ciudad Universitaria, Distrito Federal 04510, México.
      Authors
    • Espinosa-Pérez HS 3
    (1 author)
  • 4. Departamento de Ecología y Recursos Naturales, Facultad de Ciencias, Universidad Nacional Autónoma de México, Ciudad Universitaria, Distrito Federal 04510, México.
      Authors
    • Zúñiga-Vega JJ 4
    (1 author)

ORCIDs linked to this article

Biological Journal of the Linnean Society. Linnean Society of London, 06 Sep 2015, 116(4):787-804
https://doi.org/10.1111/bij.12662 PMID: 26617418 PMCID: PMC4659389

Free full text in Europe PMC

Abstract 

Superfetation is the ability of females to simultaneously carry multiple broods of embryos, with each brood at a different developmental stage. Matrotrophy is the post-fertilization maternal provisioning of nutrients to developing embryos throughout gestation. Several studies have demonstrated that, in viviparous fishes, superfetation and matrotrophy have evolved in a correlated way, such that species capable of bearing several simultaneous broods also exhibit advanced degrees of post-fertilization provisioning. The adaptive value of the concurrent presence of both reproductive modes may be associated with the production of larger newborns, which in turn may result in enhanced offspring fitness. In this study, we tested two hypotheses: (1) species with superfetation and moderate or extensive matrotrophy give birth to larger offspring compared to species without superfetation or matrotrophy; (2) species with higher degrees of superfetation and matrotrophy (i.e. more simultaneous broods and increased amounts of post-fertilization provisioning) give birth to larger offspring compared to species with relatively low degrees of superfetation and matrotrophy (i.e. fewer simultaneous broods and lesser amounts of post-fertilization provisioning). Using different phylogenetic comparative methods and data on 44 species of viviparous fishes of the family Poeciliidae, we found a lack of association between offspring size and the combination of superfetation and matrotrophy. Therefore, the concurrent presence of superfetation and moderate or extensive matrotrophy has not facilitated the evolution of larger offspring. In fact, these traits have evolved differently. Superfetation and matrotrophy have accumulated gradual changes that largely can be explained by Brownian motion, whereas offspring size has evolved fluidly, experiencing changes that likely resulted from selective responses to the local conditions.

Free full text 

Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Biol J Linn Soc Lond. Author manuscript; available in PMC 2016 Dec 1.
Published in final edited form as:
Biol J Linn Soc Lond. 2015 Dec; 116(4): 787–804.
Published online 2015 Sep 6. https://doi.org/10.1111/bij.12662
PMCID: PMC4659389
NIHMSID: NIHMS714836
PMID: 26617418

Have superfetation and matrotrophy facilitated the evolution of larger offspring in poeciliid fishes?

Claudia Olivera-Tlahuel,1 Alison G. Ossip-Klein,2 Héctor S. Espinosa-Pérez,3 and J. Jaime Zúñiga-Vega4,*
Author information Copyright and License information Disclaimer

Abstract

Superfetation is the ability of females to simultaneously carry multiple broods of embryos, with each brood at a different developmental stage. Matrotrophy is the post-fertilization maternal provisioning of nutrients to developing embryos throughout gestation. Several studies have demonstrated that, in viviparous fishes, superfetation and matrotrophy have evolved in a correlated way, such that species capable of bearing several simultaneous broods also exhibit advanced degrees of post-fertilization provisioning. The adaptive value of the concurrent presence of both reproductive modes may be associated with the production of larger newborns, which in turn may result in enhanced offspring fitness. In this study, we tested two hypotheses: (1) species with superfetation and moderate or extensive matrotrophy give birth to larger offspring compared to species without superfetation or matrotrophy; (2) species with higher degrees of superfetation and matrotrophy (i.e. more simultaneous broods and increased amounts of post-fertilization provisioning) give birth to larger offspring compared to species with relatively low degrees of superfetation and matrotrophy (i.e. fewer simultaneous broods and lesser amounts of post-fertilization provisioning). Using different phylogenetic comparative methods and data on 44 species of viviparous fishes of the family Poeciliidae, we found a lack of association between offspring size and the combination of superfetation and matrotrophy. Therefore, the concurrent presence of superfetation and moderate or extensive matrotrophy has not facilitated the evolution of larger offspring. In fact, these traits have evolved differently. Superfetation and matrotrophy have accumulated gradual changes that largely can be explained by Brownian motion, whereas offspring size has evolved fluidly, experiencing changes that likely resulted from selective responses to the local conditions.

Keywords: lecithotrophy, offspring size, phylogenetic comparative methods, Poeciliidae, simultaneous broods, viviparous fishes

INTRODUCTION

Superfetation is the ability of females to simultaneously carry multiple broods of embryos at different developmental stages (Turner, 1937, 1940; Scrimshaw, 1944). This mode of reproduction involves the occurrence of fertilization and development of a new brood before a preceding brood is born and can be clearly identified by the presence of groups of embryos within a single female that are in discretely different and well-separated stages of development (Pires et al., 2011a; Roellig et al., 2011). Superfetation is a relatively rare phenomenon that occurs in some viviparous species of three phylogenetically distant families of fishes: Clinidae (Gunn & Thresher, 1991), Zenarchopteridae (Reznick, Meredith & Collette, 2007a), and Poeciliidae (Turner, 1937, 1940; Scrimshaw, 1944; Schultz, 1961; Thibault & Schultz, 1978; Pollux et al., 2009, 2014). Some species of the bivalve family Sphaeriidae exhibit brood masses that contain developmentally discrete subsets of embryos, with each subset encapsulated in a different brood sac. This phenomenon has been termed “sequential brooding” but is equivalent to superfetation (Cooley & Foighil, 2000).

Within the fish family Poeciliidae, superfetation occurs in some genera such as Poeciliopsis, Poecilia, Heterandria, and Neoheterandria, whereas it is absent in other genera (e.g. Brachyrhaphis, Gambusia, Xhiphophorus, and Alfaro; Pires, Arendt & Reznick, 2010; Zúñiga-Vega, Macías-García & Johnson, 2010). The phylogenetic distribution of superfetation within the family Poeciliidae suggests that it has evolved independently multiple times (Reznick & Miles, 1989; Pollux et al., 2009; Pires et al., 2010; Meredith et al., 2011). In addition, the number of simultaneous broods present within females (i.e. degree of superfetation) varies widely among superfetating species. For example, Heterandria formosa exhibit a high degree of superfetation (females can carry up to eight different broods), whereas Poecilia parae exhibit a relatively low degree of superfetation (the maximum number of simultaneous broods is two; Travis et al., 1987; Pires et al., 2010). The multiple independent origins, along with the observed interspecific variation, suggest that superfetation might convey certain adaptive advantages that are not yet completely understood (Zúñiga-Vega et al., 2010).

In addition to superfetation, the amount of nutrients that are transferred from mothers to embryos during development varies widely among viviparous species of the family Poeciliidae (Reznick, Mateos & Springer, 2002; Pollux et al., 2009, 2014). Some species provide all the necessary nutrients for development before fertilization in the form of yolk without additional post-fertilization nutrient transfer. Hence, embryos lose mass during development due to metabolic costs (embryo mass decreases approximately 35% from fertilization to birth). This mode of maternal provisioning is called lecithotrophy (Wourms, 1981; Marsh-Matthews, 2011). In contrast, other species provide nutrients to developing embryos after fertilization through specialized morphological structures (i.e. placentas; Turner, 1940; Lombardi & Wourms, 1985; Mossman, 1991). This mode of maternal provisioning is called matrotrophy (Wourms, 1981; Marsh-Matthews, 2011). The relative amount of pre- and post-fertilization maternal provisioning varies widely among species (Pollux et al., 2009, 2014). Incipient matrotrophy refers to species in which females provide only small amounts of nutrients to developing embryos after fertilization in addition to the yolk that was previously allocated to eggs. In such cases embryo mass decreases only slightly (10–15%) or remains constant throughout development. Moderate matrotrophy refers to species in which embryo mass increases moderately throughout development, reflecting a relatively larger provisioning from the mother after fertilization, although still relying to some extent on pre-fertilization yolk allocation. Extensive matrotrophy refers to species in which yolk is absent from mature ova, the mother throughout gestation actively provides all the nutrients that are necessary for development, and the mass at birth is at least five times larger than the egg mass at fertilization (Reznick et al., 2002; Marsh-Matthews, 2011; Pires et al., 2011a). In some species, the increase in embryo size throughout development reaches 60,000% (Poecilia branneri and P. bifurca; Pires et al., 2010). There exists a continuum in the modes of maternal provisioning among poeciliid species, from strict lecithotrophy to extensive matrotrophy (Pollux et al., 2009, 2014).

Some studies have demonstrated an evolutionary link between superfetation and matrotrophy (Reznick & Miles, 1989; Pollux et al., 2009, 2014; Trexler & DeAngelis, 2010; Meredith et al., 2011). Most species with superfetation also exhibit some degree of post-fertilization maternal provisioning, although a few exceptions exist (e.g. Priapichthys festae, Neoheterandria tridentiger, N. elgans, and Poeciliopsis monacha exhibit superfetation and lecithotrophy; Reznick & Miles, 1989; Pollux et al., 2009). This association has led to the hypothesis that the evolution of one of these two traits has facilitated the evolution of the other (Pollux et al., 2009, 2014; Trexler & DeAngelis, 2010; Meredith et al., 2011). The apparent coevolution of these reproductive strategies may have important consequences in terms of reproductive output. Hence, the adaptive significance of matrotrophy and superfetation could be associated with their patterns of covariation with life-history traits, such as number and size of offspring. For instance, species with superfetation tend to produce smaller broods and do so more often than non-superfetating species (Reznick & Miles, 1989; Pollux et al., 2009). This mode of reproduction allows females to give birth continuously, with an interval between broods being as short as a few days. In contrast, species without superfetation, which for the most part exhibit lecithotrophy or only incipient matrotrophy, tend to produce larger broods and give birth less frequently, with many days between broods (Fraser & Renton, 1940; Thibault & Schultz, 1978; Cheong et al., 1984; Leips et al., 2009). The production of larger broods in non-superfetating species could result in smaller offspring because viviparous females have limited space within their reproductive tracts and, hence, they should experience the widely recognized trade-off between number and size of offspring (Smith & Fretwell, 1974; Brockelman, 1975; Stearns, 1976).

In contrast, superfetating species that exhibit certain amounts of post-fertilization nutrient transfer might be able to increase the size of individual offspring. When matrotrophic females transfer an amount of nutrients during development that results in at least a slight increase in embryo mass from fertilization to birth (i.e. moderate matrotrophy), early stage embryos are smaller than late stage embryos. In combination with superfetation, where smaller broods imply fewer full-term embryos at any given time, this amount of post-fertilization nutrient transfer should promote less space requirements within the reproductive tract because these females bear a mix of early (small) and late (large) embryos compared to non-superfetating females that, at some point, would bear a high number of large full-term embryos. Therefore, the additional body volume that may be available to females exhibiting superfetation and at least moderate matrotrophy could be used to produce larger newborns. The combined effects of superfetation and matrotrophy on the potential reduction of space requirements within the reproductive tract could be even stronger in those species that exhibit extensive matrotrophy because the size of the ova at fertilization is remarkably smaller compared to the size of late stage embryos. Thus, superfetation and extensive matrotrophy promote the presence of very small early stage embryos, few large late stage embryos and, potentially, low ovarian volume. This in turn could more easily allow for increases in the size of individual neonates.

Producing large newborns may increase their probability of surviving to maturity, at least under particular ecological conditions (Reznick, 1982; Reznick, Bryga & Endler, 1990; Bashey, 2008; Gordon et al., 2009; Riesch et al., 2010a). Therefore, if the coevolution of superfetation and matrotrophy has facilitated the evolution of larger offspring in poeciliid fishes, then the adaptive significance of the combined presence of these two reproductive modes could be associated with enhanced offspring fitness.

In this study we used modern phylogenetic comparative methods to analyze the evolutionary relationship between offspring size and the combined presence of superfetation and matrotrophy in 44 species of viviparous fishes of the family Poeciliidae. We focused on testing two main hypotheses. (1) Species with superfetation and at least moderate matrotrophy give birth to larger offspring compared to species without superfetation, species with superfetation and lecithotrophy, and species with superfetation and incipient matrotrophy (after accounting for differences among species in the size of adult females). (2) Species with higher degrees of superfetation and matrotrophy (i.e. more simultaneous broods and increased post-fertilization maternal provisioning) give birth to larger offspring compared to species with relatively low degrees of superfetation and matrotrophy (i.e. fewer simultaneous broods and less post-fertilization provisioning). In other words, offspring size should increase as the number of simultaneous broods and the amount of post-fertilization nutrient transfer increase. This hypothesis is based on the observation that superfetating species with more simultaneous broods and relatively high levels of matrotrophy tend to produce fewer newborns per brood compared to superfetating species with fewer simultaneous broods and lower levels of matrotrophy. For example, Xenodexia ctenolepis (dry weight at birth is over 3 times larger than unfertilized ova) and Poecilia branneri (dry weight at birth is over 86 times larger than unfertilized ova) produce on average 4.13 and 4.77 simultaneous broods, with 4.40 and 4.13 newborns per brood, respectively (Reznick et al., 2007b; Pires et al., 2010). In contrast, Poeciliopsis monacha (embryo dry weight decreases during development) and P. turrubarensis (embryo dry weight barely increases during development) produce on average 2 and 2.12 simultaneous broods, with 11.80 and 7.64 newborns per brood, respectively (Thibault & Schultz, 1978; Zúñiga-Vega, Reznick & Johnson, 2007). For those species with advanced degrees of superfetation and matrotrophy, more smaller broods likely imply very few large full-term embryos, more small embryos of earlier stages, less total space requirements within the reproductive tract and, therefore, the possibility of producing considerably large offspring.

MATERIAL AND METHODS

Data collection

We conducted an extensive literature search to extract data on average standard length (mm) of reproductive females, embryo dry mass (mg) at the last stage of development (i.e. stage 11 according to Haynes [1995] or the equivalent stage in other classification methods of embryonic development), degree of superfetation (average number of simultaneous broods at different developmental stages), and matrotrophy index (MI) for species of the family Poeciliidae (Table 1). We considered dry mass at the last stage of development as an appropriate proxy for size (mass) at birth (Haynes, 1995). The matrotrophy index is a standard measure of the amount of post-fertilization maternal provisioning and is calculated as the dry mass of the offspring at birth divided by the dry mass of the egg at fertilization (Wourms, Grove & Lombardi, 1988; Trexler, 1997; Reznick et al., 2002; Marsh-Matthews, 2011). Thus, the MI represents the change in embryonic dry mass during development, with an MI value higher than 1 indicating an increase in embryo mass and lower than 1 indicating a decrease in embryo mass. An MI equal to 1 indicates that embryo mass remains constant from fertilization to parturition. When information was available for more than one population of a single species we used average values as input in our comparative analyses.

Table 1

Female size (standard length), offspring size (mass), degree of superfetation (average number of simultaneous broods per female), and matrotrophy index for 44 species of the fish family Poeciliidae. Offspring size corresponds to the dry mass at the last stage of development. Given that larger species produce significantly larger offspring, we also show the residuals from a linear regression between female length and offspring mass. The column PC1 shows the species-specific scores in the first principal component that represents a combined measure of superfetation and matrotrophy.

SpeciesFemale
size (mm)		Offspring
size (mg)		Offspring size
(residuals)		Superfetation
(number of
broods)		Matrotrophy
index		PC1References
Brachyrhaphis rhabdophora37.111.64−0.4581.000.77−0.623Johnson & Belk, 2001; Pollux et al., 2014
Brachyrhaphis episcopi30.082.430.8231.000.83−0.621Jennions & Telford, 2002
Brachyrhaphis holdridgei22.610.6−0.4851.000.66−0.628Pollux et al., 2014; This study
Heterandria formosa18.880.58−0.2443.0935.002.439Schrader & Travis, 2012; Leips & Travis, 1999; Pollux et al., 2014
Poeciliopsis lucida32.50.71−1.0663.002.000.968Thibault & Shultz, 1978; Pollux et al., 2014
Poeciliopsis pleurospilus45.011.04−1.6121.350.50−0.367This study
Poeciliopsis occidentalis34.201.66−0.2351.641.12−0.114Pollux et al., 2014; This study
Poeciliopsis prolifica24.630.59−0.6363.305.401.343Pires, McBride & Reznick, 2007; Pollux et al., 2014
Poeciliopsis infans28.750.82−0.6941.791.05−0.003Frías-Alvarez et al., 2014; Molina-Moctezuma, 2015
Poeciliopsis monacha32.51.26−0.5162.000.610.140Thibault & Shultz, 1978; Pollux et al., 2014
Poeciliopsis viriosa37.930.43−1.7262.230.930.329Pollux et al., 2014; This study
Poeciliopsis turneri48.53.390.4963.0041.402.642Thibault & Shultz, 1978; Pollux et al., 2014
Poeciliopsis gracilis30.340.92−0.7051.880.840.057Frías-Alvarez et al., 2014; Molina-Moctezuma, 2015
Poeciliopsis turrubarensis39.381.58−0.6772.121.050.251Zúñiga-Vega et al., 2007
Poeciliopsis latidens39.711.04−1.2401.120.86−0.526Pollux et al., 2014; This study
Poeciliopsis fasciata30.781.920.2641.340.91−0.355Pollux et al., 2014; This study
Poeciliopsis baenschi25.621.28−0.0151.640.98−0.122Molina-Moctezuma, 2011; This study
Heterophallus milleri23.761.220.0551.000.74−0.624Riesch et al., 2011
Gambusia panuco26.861.16−0.2221.000.79−0.622Ader, Zúñiga-Vega & Johnson, unpublished
Gambusia vittata24.531.16−0.0591.001.29−0.601Weldele, Zúñiga-Vega & Johnson, in press
Gambusia sexradiata24.191.810.6151.001.74−0.582Riesch et al., 2010a; Riesch, unpublished
Gambusia eurystoma23.635.744.5841.000.76−0.624Riesch et al., 2010a; Riesch, unpublished
Gambusia hubbsi28.393.221.7311.000.86−0.619Riesch et al., 2013
Gambusia yucatana23.450.8−0.3441.000.53−0.633This study
Gambusia holbrooki31.70.97−0.7501.000.64−0.629Meffe, 1990; Pollux et al., 2014
Gambusia affinis351.9−0.0511.000.62−0.629Stearns, 1983; Pollux et al., 2014
Gambusia speciosa28.580.94−0.5621.000.45−0.637This study
Gambusia aurata23.820.66−0.5101.000.82−0.621This study
Belonesox belizanus99.46.90.4501.000.70−0.626Turner & Snelson, 1984; Pollux et al., 2014
Pseudoxhiphophorus jonesii48.104.361.4941.000.65−0.628This study
Priapella chamulae30.32.310.6881.000.71−0.626Riesch et al., 2012
Priapella olmecae44.382.960.3561.000.76−0.623This study
Poecilia mexicana41.633.260.8461.000.63−0.629Riesch, Plath & Schlupp, 2010b; Pollux et al., 2014
Poecilia sulphuraria26.43.732.3801.000.69−0.627Riesch et al., 2010a
Poecilia butleri41.771.64−0.7841.002.30−0.558Zúñiga-Vega et al., 2011
Poecilia (Micropoecilia) bifurca14.450.48−0.0351.2055.051.836Pires et al., 2010; Pollux et al., 2014
Poecilia (Micropoecilia) branneri19.160.62−0.2244.7786.415.915Pires et al., 2010; Pollux et al., 2014
Poecilia (Micropoecilia) parae23.330.67−0.4651.036.75−0.346Pires et al., 2010; Pollux et al., 2014
Poecilia (Micropoecilia picta21.30.75−0.2431.000.78−0.623Pires et al., 2010; Pollux et al., 2014
Poecilia (Acanthophacelus) reticulata17.80.860.1111.000.66−0.628Pires et al., 2010; Pollux et al., 2014
Poecilia (Acanthophacelus) wingei27.11.860.4611.000.84−0.620Pires et al., 2010; Pollux et al., 2014
Phalloceros caudiomaculatus29.620.8−0.7751.002.14−0.565Arias & Reznick, 2000; Pollux et al., 2014
Phalloceros anisophallos29.90.79−0.8041.002.80−0.537Almeida-Silva & Mazzoni, 2014
Xenodexia ctenolepis473.570.7814.133.381.896Reznick et al., 2007b; Pollux et al., 2014

For some species (i.e. Brachyrhaphis rhabdophora, B. episcopi, Gambusia vittata, G. panuco, Heterandria formosa, Poecilia butleri, Poeciliopsis baenschi, P. gracilis, P. infans, and P. turrubarensis), the authors that have published life-history descriptions provided us with their raw data, and we conducted a regression between stage of development and log-transformed embryo dry mass. From this regression, we estimated the dry mass at the last developmental stage as a proxy for size at birth. In addition, for three of these species (B. episcopi, G. panuco, and P. baenschi), we also used this regression to estimate the dry mass at fertilization (stage 4 according to Haynes [1995] or the equivalent stage in other classification methods) and then calculated an MI value as explained above (for most species MI values were available in Pollux et al., 2014).

In addition, we quantified the same variables (standard length of reproductive females, dry mass of embryos at the last stage of development, and degree of superfetation) from preserved females of the following species: Brachyrhaphis holdridgei, Poeciliopsis fasciata, P. latidens, P. occidentalis, P. pleurospilus, P. viriosa, Gambusia yucatana, G. speciosa, G. aurata, Pseudoxiphophorus jonesii, and Priapella olmecae (Table 1). These specimens were preserved in ethanol at the National Collection of Fishes (Instituto de Biología, Universidad Nacional Autónoma de México). We dissected 20 reproductive females per species. Before dissection, we measured the standard length (mm) of all females with a digital caliper. Upon dissection, we removed the ovary and separated embryos into distinct broods based on stage of development (according to Haynes, 1995). We quantified the degree of superfetation as the number of broods at different developmental stages within each female. We dried each brood of embryos at 55°C overnight, and then weighed embryos to the nearest 0.01 mg. To obtain an estimate of the dry mass at the last stage of development, we ran a regression between developmental stage and log-transformed embryo dry mass. In addition, for those species whose MI was not available in the literature (G. yucatana, G. speciosa, G. aurata, Poecilipsis pleurospilus, Pseudoxiphophorus jonesii, and Priapella olmecae) we also used this regression to estimate dry mass at fertilization and then calculated an MI as explained above. In total, we obtained data for 44 poeciliid species (Table 1).

Accounting for potential confounding factors

Given that larger species could produce larger offspring and/or have higher degrees of superfetation or matrotrophy simply as a consequence of a larger body size (Reznick & Miles, 1989), we conducted linear regressions (using species-specific average values) between female length and offspring mass, between female length and superfetation, and between female length and MI. Offspring mass increased significantly with species-specific female size (F = 31.7, df = 1, 42, P < 0.0001). Therefore, in all comparative analyses we used the residuals of this regression as size-corrected estimates of offspring size. In contrast, neither superfetation nor matrotrophy were affected by species-specific female size (superfetation: F = 0.003, df = 1, 42, P = 0.96; matrotrophy: F = 1.7, df = 1, 42, P = 0.19) and, therefore, we used the raw (unadjusted) values of number of simultaneous broods and MI.

Another potential confounding factor is the way in which specimens were originally preserved. Some species were preserved in formalin, whereas others were preserved in ethanol (Supporting Information, Table S1). Ethanol can extract lipids from the tissues and, therefore, might bias the estimates of offspring mass and matrotrophy index (Shields & Carlson, 1996). We tested this potential confounding effect by means of general linear models using offspring dry mass and MI as response variables, the preservation method (formalin or ethanol) as a main factor, and female length as a covariate. We also included in these models the interaction between preservation method and female length. Neither the preservation method nor its interaction with female length significantly affected offspring dry mass (preservation method: F = 0.45, df = 1, 40, P = 0.51; interaction: F = 0.10, df = 1, 40, P = 0.76) or MI (preservation method: F = 0.14, df = 1, 40, P = 0.71; interaction: F = 0.31, df = 1, 40, P = 0.58). Again here, female length had a highly significantly effect on offspring mass (F = 12.07, df = 1, 40, P = 0.001), whereas it did not affect MI (F = 1.29, df = 1, 40, P = 0.26). Therefore, most of the variation among species in offspring mass was explained by the size of adult females, with no detectable effect of the preservation method on either offspring mass or MI.

To further examine the potential confounding effect of the preservation method, we implemented all our comparative analyses using only the 33 species that were preserved in ethanol (Table S1). All results were qualitatively similar to those obtained using the complete data set (44 species). Therefore, we report here the results obtained with the 44 species and those from the subset of 33 species in the Supporting Information.

A combined measure of superfetation and matrotrophy

We hypothesized that the combined effect of superfetation and at least moderate matrotrophy may allow for increases in size at birth in poeciliid fishes. Therefore, we conducted a principal components analysis on our species-specific values for degree of superfetation and MI. This analysis resulted in a single principal component with an eigenvalue > 1 (1.59) in which both superfetation and MI had high loadings (0.89 for both variables). Hence, we considered this principal component as a combined measure of superfetation and matrotrophy. Species with high positive scores corresponded to those with the highest degrees of superfetation and the largest amounts of post-fertilization maternal provisioning (e.g. Poecilia branneri: 4.77 simultaneous broods and MI = 86.41), whereas those with relatively high negative scores corresponded to lecithotrophic species without superfetation (e.g. Gambusia speciosa: lacks superfetation and MI = 0.45; Table 1). This combined measure of superfetation and matrotrophy was used as input in our comparative analyses.

Phylogeny and branch lengths

We used a well-resolved phylogeny of the family Poeciliidae that was reconstructed by Pollux et al. (2014) using a large molecular data set (20 nuclear and 8 mitochondrial markers). We pruned their topology to our genera and species of interest. Even though this phylogenetic reconstruction did not include all the 44 species that we considered in our study, we used it as the main representation of the relationships among genera. However, it did not include the genus Heterophallus. Therefore, the phylogenetic position of H. milleri with respect to other genera was based on Doadrio et al. (2009). Phylogenetic relationships within genera, among those species that were not included in the topology of Pollux et al. (2014), were obtained from the following studies: Mojica, Meyer & Barlow (1997) (Brachyrhaphis), Lydeard, Wooten & Meyer (1995), Langerhans et al. (2012) (Gambusia), Ptacek & Breden (1998) (Poecilia), Meredith et al. (2010) (Poecilia, subgenus Micropoecilia), Morales-Cazan & Albert (2012) (Poeciliopsis), and Meyer, Schories & Schartl (2011) (Priapella). The composite phylogeny that we used for comparative analyses is shown in Fig. 1.

An external file that holds a picture, illustration, etc. Object name is nihms714836f1a.jpg
An external file that holds a picture, illustration, etc. Object name is nihms714836f1b.jpg

Composite phylogeny showing the relationships among 44 species of the family Poeciliidae as well as estimated ancestral states of the combined presence of superfetation and at least moderate matrotrophy (i.e. matrotrophy index > 1) based on parsimony. Green branches indicate presence of superfetation and at least moderate matrotrophy, whereas white branches indicate absence of superfetation, superfetation with lecithotrophy, or superfetation with incipient matrotrophy. Both trees (A, reconstruction 1; B, reconstruction 2) were equally parsimonious.

Phylogenetic comparative methods require information about the expected amount of evolutionary change along each branch of the phylogenetic tree (i.e. branch lengths; Harvey & Pagel, 1991). Given that we based our comparative analyses on a composite phylogeny, we lacked branch lengths in units of DNA sequence divergence. Therefore, we used a conservative approach that consisted of generating 1,000 possible sets of branch lengths in units of time by means of computer simulation, assuming a standard branching process of speciation (Martins, 1996). We conducted each analysis on each of these 1,000 phylogenies and report average results. For analyses for which the calculation of average results across 1,000 phylogenies was not automated, we generated five sets of branch lengths and report the average results across those five phylogenies.

Phylogenetic comparative methods

We used the program COMPARE 4.6b (Martins, 2004) to implement different phylogenetic comparative methods. To compare offspring size (adjusted for female length) between species with different degrees of superfetation and matrotrophy we used the adaptation-inertia model of Hansen (1997). This comparative method accounts for phylogenetic relatedness when analyzing the relationship between a phenotypic trait (e.g. offspring size) and a particular selective force. Hansen (1997) described this as a model of evolutionary change based on strong stabilizing selection in which phenotypes evolve towards an adaptive optimum. This model accounts for competing selective forces that might result in different optimal phenotypic values depending on the particular evolutionary context. We implemented this method twice, with each implementation representing a different hypothesis. First, we tested the hypothesis that superfetation accompanied by at least moderate matrotrophy would be enough to facilitate the evolution of larger offspring. For this purpose, we classified our species into one of the following two groups, that we considered alternative reproductive modes. (1) Species with superfetation and at least moderate matrotrophy (i.e. all superfetating species with an MI > 1), and (2) species without superfetation, species with superfetation and lecithotrophy (MI < 0.8), and species with superfetation and incipient matrotrophy (0.8 ≤ MI ≤ 1). We treated these two groups as alternative evolutionary forces and estimated how much of the variation across species in offspring size can be explained by the relative time each species has evolved with either of these two reproductive modes (r2).

In the second implementation we tested the hypothesis that the combination of superfetation and extensive matrotrophy would provide even further opportunities for the evolution of larger offspring. Thus, we classified our species into one of the following two groups (alternative reproductive modes). (1) Species with superfetation and extensive matrotrophy (i.e. all superfetating species with an MI > 5), and (2) all other species. Again here, we estimated the proportion of interspecific variance in offspring size that can be explained by these two alternative evolutionary forces (r2).

This method also allowed us to estimate optimal values of offspring size for species that exhibit these different combinations of the presence/absence of superfetation and relative amount of matrotrophy. Based on our hypotheses, we expected a larger optimal size at birth for superfetating species that exhibit moderate or extensive matrotrophy compared to that for non-superfetating species, superfetating species with lecithotrophy, and superfetating species with incipient matrotrophy. Given that we adjusted our data to account for interspecific variation in female length, we actually expected a positive optimal value for superfetating species with moderate or extensive matrotrophy (a positive residual indicates a larger offspring size than that expected for a given female length) and a negative optimal value for non-superfetating species, superfetating species with lecithotrophy, and superfetating species with incipient matrotrophy (a negative residual indicates a smaller offspring size than that expected for a given female length).

Hansen’s (1997) method requires previous knowledge about the evolutionary history of the putative selective agent across the phylogeny. Thus, we reconstructed ancestral states for (1) the presence of superfetation with at least moderate matrotrophy and (2) the presence of superfetation with extensive matrotrophy using maximum likelihood and parsimony approaches implemented in the program MESQUITE 3.0 (Maddison & Maddison, 2009).

To estimate the magnitude of the relationship between offspring size and our combined measure of superfetation and matrotrophy we used different phylogenetic comparative methods. The first was Felsenstein’s (1985) independent contrasts (FIC). This method solves the statistical problem of non-independence in the data due to shared ancestry and assumes that traits evolve along a phylogeny at a constant rate. Therefore, this mode of evolution can be described by Brownian motion, in which phenotypes diverge under random genetic drift. Hence, the expected phenotypic difference between sister species grows proportional to the time since they shared a common ancestor. From this method, we obtained a phylogenetically corrected correlation coefficient (r). We compared this to a Pearson correlation coefficient calculated using the uncorrected data (i.e. a non-phylogenetic approach referred to as TIPS; Martins & Garland, 1991). The TIPS method assumes that traits adapt immediately to the local environment, leaving behind no trace of phylogenetic relationships.

We also used the phylogenetic generalized least squares method (PGLS) (Martins & Hansen, 1997), which uses an Ornstein-Uhlenbeck (OU) process as the underlying model of evolution. An OU process assumes that traits evolve under stabilizing selection. Similar to the FIC approach, PGLS estimates the relationship between two traits by means of a phylogenetically corrected correlation coefficient, r. In addition, PGLS estimates an additional parameter, α, which may vary between 0 and 15.5 (in COMPARE). When α is small, the rate of local adaptation is very slow and phenotypes appear to evolve by Brownian motion, which results in high phylogenetic signal. In contrast, when α is large the rate of adaptive change is fast and phenotypes evolve without a strong phylogenetic signal. Given that PGLS is a least squares regression model, we used our combined measure of superfetation and matrotrophy as predictor variable (X) and size-corrected offspring size as response variable (Y).

Finally, we used the phylogenetic mixed model (PMM) (Lynch, 1991; Housworth, Martins & Lynch, 2004), which also estimates the correlation between two traits after accounting for phylogenetic relatedness among species. The PMM model assumes that phenotypes are the result of a linear combination of gradually accumulated evolutionary changes that have occurred along the phylogenetic history of the species and rapid evolutionary changes that likely resulted from adaptive responses to sudden changes in the environment. Thus, this model estimates the relative proportion of gradual Brownian motion-like change (h2, defined as phylogenetic heritability) for both traits (offspring size and the combined measure of superfetation and matrotrophy). When h2 approaches one, most of the interspecific variation in the phenotypes is explained by the phylogenetic history. When h2 approaches zero, phenotypes show little trace of phylogenetic signal and evolutionary changes have occurred faster than expected by Brownian motion (Housworth et al., 2004). Using PMM, we also calculated h2 separately for degree of superfetation and MI.

RESULTS

Ancestral reconstructions of superfetation and matrotrophy

The parsimony analysis resulted in two equally parsimonious ancestral reconstructions of superfetation and at least moderate matrotrophy, referred to herein as reconstructions 1 and 2 (Fig. 1). In reconstruction 1, the combination of superfetation and MI values > 1 originated in several occasions; once within the subgenus Micropoecilia, three times within the genus Poeciliopsis, with one loss (in P. pleurospilus), and once in Hererandria formosa (Fig. 1A). Reconstruction 2 was quite similar, with the only difference being two origins (instead of three) within Poeciliopsis and two loses (in P. pleurospilus and P. gracilis) (Fig. 1B). In both reconstructions we found uncertainty on whether the common ancestor of the family exhibited superfetation and at least moderate matrotrophy or not. If it did, then Xenodexia ctenolepis retained the ancestral state. If not, an additional independent origin occurred in X. ctenolepis.

The maximum likelihood reconstruction also indicated an origin of superfetation with at least moderate matrotrophy within the subgenus Micropoecilia, two origins within Poeciliopsis and two loses (in P. pleurospilus and P. gracilis), one additional origin in H. formosa, and high uncertainty with respect to the presence or absence of this reproductive mode in the common ancestor of the entire family (proportional likelihoods: 0.53 for superfetation and at least moderate matrotrophy and 0.47 for any other strategy) (Fig. 2). Again here, X. ctenolepis represents either an additional independent origin or retention of an ancestral state. We note that the proportional likelihoods of the ancestral nodes in the group conformed by P. turneri, P. gracilis, and P. turrubarensis indicated high uncertainty with respect to the presence or absence of superfetation and MI values > 1 (Fig. 2). Given all this uncertainty, we decided to implement Hansen’s (1997) method using both parsimony reconstructions 1 and 2 (the maximum likelihood reconstruction summarized these two) and assuming either presence or absence of superfetation and at least moderate matrotrophy in the most ancestral node. All results were qualitatively similar. Thus, for simplicity we report here only the results using reconstruction 1 and assuming that the ancestor of the family exhibited both superfetation and an MI > 1.

An external file that holds a picture, illustration, etc. Object name is nihms714836f2.jpg

Composite phylogeny showing the relationships among 44 species of the family Poeciliidae as well as estimated ancestral states of the combined presence of superfetation and at least moderate matrotrophy (i.e. matrotrophy index > 1) based on maximum likelihood. Circles depict proportional likelihoods for the presence (green) and absence (white) of superfetation and at least moderate matrotrophy.

Regarding the ancestral reconstructions of superfetation and extensive matrotrophy, the parsimony analysis resulted in a single most parsimonious tree and identified four evolutionary origins (Fig. 3A). The first occurred within the subgenus Micropoecilia, particularly in the common ancestor of Poecilia (Micropoecilia) bifurca, P. (M.) branneri, and P. (M.) parae, the second in Poeciliopsis turneri, the third in P. prolifica, and the fourth in Heterandria formosa. The maximum likelihood reconstruction gave identical results with minimum uncertainty with respect to the presence of superfetation and extensive matrotrophy in ancestral nodes (Fig. 3B).

An external file that holds a picture, illustration, etc. Object name is nihms714836f3a.jpg
An external file that holds a picture, illustration, etc. Object name is nihms714836f3b.jpg

Composite phylogeny showing the relationships among 44 species of the family Poeciliidae as well as estimated ancestral states of the combined presence of superfetation and extensive matrotrophy (i.e. matrotrophy index > 5) based on (A) parsimony and (B) maximum likelihood. Red branches in A indicate presence of superfetation and extensive matrotrophy. Circles in B depict proportional likelihoods for the presence (red) and absence (white) of superfetation and extensive matrotrophy.

Potential evolutionary relationships between offspring size and the combination of superfetation and matrotrophy

Interspecific variation in offspring size could not be explained well by the evolution of superfetation and at least moderate matrotrophy (Hansen’s r2 = 0.05) (Table 2). Contrary to what we expected, the optimal offspring size (after adjusting for female length) for species with superfetation and MI values > 1 was −0.38, whereas that for non-superfetating species, superfetating species with lecithotrophy, and superfetating species with incipient matrotrophy was 0.16.

Table 2

Phylogenetic estimates of the magnitude of the evolutionary relationship between offspring size and the combination of superfetation and matrotrophy calculated using different phylogenetic comparative methods.

Hansen’s method
FICTIPSPGLSPMM
SF and MI > 1SF and MI > 5
r = 0.23r = −0.10r = 0.07r = −0.21
r2 = 0.05r2 = 0.01r2 = 0.08
α = 15.4
Offspring size h2 =
0.22
SF and MI h2 = 0.86
SF h2 = 0.81
MI h2 = 0.87

FIC, Felsenstein’s independent contrasts; TIPS, non-phylogenetic approach; PGLS, phylogenetic generalized least squares method assuming an Ornstein-Uhlenbeck model of evolution; PMM, phylogenetic mixed model; SF, superfetation; MI, matrotrophy index; r, correlation coefficient; r2, proportion of the variation among species in offspring size explained by superfetation and matrotrophy; α, a measure of the strength of selection pulling offspring size towards optimal values; h2, phylogenetic heritability.

Similarly, the evolution of superfetation accompanied by extensive matrotrophy did not explain the interspecific variation in offspring size (Hansen’s r2 = 0.01) (Table 2). Again here, the estimated optimal values were opposite to our prediction. The optimal offspring size for species with superfetation and MI values > 5 was −0.26, whereas that for all other species was 0.05.

The evolutionary relationship between degrees of superfetation and matrotrophy treated as a combined continuous variable (i.e. the species-specific scores in the first principal component) and offspring size was very weak as indicated by the small, non-significant correlation coefficients obtained from FIC, PGLS, and PMM (0.23, 0.07, and −0.21, respectively; P > 0.13 in all cases) (Table 2). A non-phylogenetic approach also indicated a weak correlation (TIPS r = −0.10, P = 0.50). According to PGLS, the combination of superfetation and matrotrophy explained only 8% of the interspecific variation in offspring size (r2 = 0.08). All these analyses conducted on the subset of 33 species that were preserved in ethanol yielded equivalent results (Table S2).

Phylogenetic signal in superfetation, matrotrophy, and offspring size

PGLS estimated a large value of the parameter α (15.4), which indicates that the species-specific offspring sizes have been pulled strongly towards optimal values. Therefore, the evolution of this trait cannot be explained very well by the phylogenetic history alone. Consistent with this result, the phylogenetic heritability of offspring size was low (h2 = 0.22), retaining only weak evidence of the phylogeny. In contrast, the combination of superfetation and matrotrophy has tracked the phylogeny more closely, resulting in a remarkably high phylogenetic heritability (h2 = 0.86) (Table 2). Separately, superfetation and MI values also had high phylogenetic heritabilities (h2 = 0.81 and h2 = 0.87, respectively). We obtained similar estimates of phylogenetic signal for offspring size, superfetation, and matrotrophy from the subset of 33 species that were preserved in ethanol (Table S2).

DISCUSSION

Superfetation and matrotrophy have not facilitated the evolution of larger offspring

We have demonstrated here that the combination of superfetation and matrotrophy does not exhibit an evolutionary relationship with offspring size, regardless of whether superfetation and matrotrophy were treated as a dichotomous variable (presence or absence of superfetation and moderate or extensive matrotrophy) or as a continuous variable (degrees of superfetation and matrotrophy combined into a single measure). Thus, neither of our two hypotheses was supported by our results. In fact, the estimated optimal sizes at birth for species with superfetation and moderate or extensive matrotrophy were smaller compared to those for species without superfetation, species with superfetation and lecithotrophy, and species with superfetation and incipient matrotrophy (although this effect was weak). This result was contrary to expectations. Pires et al. (2011b) found a similar result using only six species of the genus Poeciliopsis, all of which exhibit superfetation. Larger MI values were associated with smaller offspring. Reznick & Miles (1989), in a preliminary examination that did not account for phylogenetic relatedness, also found that superfetation (which they noted was strongly linked to matrotrophy) was not associated with larger offspring.

Our estimates of phylogenetic heritability also confirmed that offspring size has evolved at a different pace in comparison to superfetation and matrotrophy, rather than in a correlated way as we expected. Offspring size has evolved flexibly at a faster pace, leaving behind little trace of the phylogeny. In contrast, the combination of superfetation and matrotrophy has evolved at a rate that has been proportional to the time elapsed between speciation events. When we analyzed the degree of superfetation and matrotrophy indexes separately, we confirmed that both traits evolved for the most part by Brownian motion. This is the main reason why most of the interspecific variation in the degrees of superfetation and matrotrophy can be explained by the phylogenetic history. However, we highlight that the particular combination of superfetation and extensive matrotrophy has arisen recently and independently in three distinct species (Heterandria formosa, Poeciliopsis prolifica, and P. turneri) with little evidence of the presence of this particular reproductive mode in ancestral nodes. In contrast, within the subgenus Micropoecilia, superfetation and extensive matrotrophy arose in the common ancestor of Poecilia (Micropoecilia) bifurca, P. (M.) branneri, and P. (M.) parae and these three species inherited this reproductive mode (Fig. 3). Apparently, these distinct and relatively recent evolutionary events that resulted in the co-occurrence of superfetation and extensive matrotrophy did not promote parallel changes in offspring size.

We note here an important caveat. Superfetating species appear to produce smaller broods (Reznick & Miles, 1989; Pollux et al., 2009), thus potentially providing additional space within the reproductive tract if at least moderate matrotrophy is present. However, the potential for this additional space within the reproductive tract, that we hypothesized could be used to produce larger offspring, depends on total fecundity remaining similar to that of non-superfetating and lecithotrophic species. If the selective forces that favor the presence of superfetation and matrotrophy also promote increased fecundity, then the additional space could be used to produce more embryos (by means of increasing the number of simultaneous broods) rather than larger offspring.

We found no evidence of an effect of superfetation and matrotrophy on offspring size at the interspecific level. However, several studies have demonstrated that these three traits may vary widely among populations within particular species (Trexler, 1985; Riesch, Martin & Langerhans, 2013; Frías-Alvarez et al., 2014). Therefore, a large offspring size as a result of increased levels of superfetation and matrotrophy at the intraspecific level is still a possibility that deserves future field and experimental tests.

Offspring size as a local adaptation

If offspring size is not influenced by the combination of superfetation and moderate or extensive matrotrophy, then other factors must have caused the observed variation in size at birth among species of the family Poeciliidae. Our results indicated a low phylogenetic signal in offspring size because there was more interspecific variation than expected by Brownian motion. Therefore, the evolution of offspring size in poeciliid fishes appears to be driven primarily by local adaptation to particular environmental conditions.

Several studies in poeciliids demonstrated adaptive responses of offspring size to local conditions. For example, Leips & Travis (1999) found that offspring of Heterandria formosa from high-density populations were 45% larger than offspring from low-density populations. High population density is characterized by high levels of competition and, under these conditions, large offspring are more likely to survive to maturity (Schrader & Travis, 2012). This same positive effect of population density and intraspecific competition on offspring size has been observed in the guppy, Poecilia reticulata (Bashey, 2006, 2008).

Predation is also an important selective force for the evolution of offspring size (Gordon et al., 2009; Schrader & Travis, 2012). The size at birth of Poecilia reticulata differs between populations inhabiting two types of predation environments. Female P. reticulata from populations sympatric with predators that feed on small juvenile fish produce larger newborns relative to females from populations that do not experience this predation pressure. Larger newborns reach adult sizes faster than smaller newborns and spend less time at body sizes susceptible to higher predation rates. In contrast, when coexisting with predators that prey on large adult fish, females of P. reticulata increase their reproductive output by producing many smaller offspring (Reznick, 1982; Reznick & Endler, 1982; Reznick et al., 1990). Two species in the genus Brachyrhaphis (B. episcopi and B. rhabdophora) show a similar pattern of predator-mediated divergence in offspring size (Johnson & Belk, 2001; Jennions & Telford, 2002). Additional studies in Gambusia hubbsi and Poecilia vivipara also have suggested local selective effects of predation on offspring size (Downhower, Brown & Matsui, 2000; Gomes & Monteiro, 2007). Apparently, high rates of cannibalism on juveniles may also select for larger offspring sizes as observed in Poeciliopsis monacha (Thibault, 1974; Weeks & Gaggiotti, 1993).

Additionally, toxic compounds in the water (e.g. hydrogen sulfide) could be selective agents for offspring size. Species such as Poecilia sulphuraria and Gambusia eurystoma that inhabit water bodies with high concentrations of hydrogen sulfide, produce relatively large newborns. This is presumably because a large body size results in a relatively low surface/volume ratio, which in turn implies that less surface area per volume of body tissue is exposed to the toxin. In addition, larger offspring likely have lower metabolic rates and, hence, lower oxygen consumption compared to smaller offspring (Riesch et al., 2010a). Some studies have documented effects of water salinity on offspring size with contradictory results. In Gambusia affinis, G. holbrooki, and Heterandria formosa water bodies with high salinities appear to promote smaller offspring (Stearns & Sage, 1980; Brown-Peterson & Peterson, 1990; Alcaraz & García-Berthou, 2007; Martin et al., 2009). In contrast, in Poecilia vivipara high salinity is associated with larger offspring (Gomes & Monteiro, 2007). The mechanisms by which salinity favors large offspring in some species and small offspring in other species remain unknown. In addition to factors that promote adaptive changes, offspring size might also exhibit phenotypic plasticity as a result of food availability (Grether & Kolluru, 2011). In Brachyrhaphis episcopi, females produce larger offspring during the wet season when food is more abundant (Jennions et al., 2006). All these studies have demonstrated that in poeciliid fishes, offspring size may exhibit adaptive and plastic responses to a wide array of environmental conditions, thereby explaining the fluid evolutionary changes that we have detected here.

Adaptive value of superfetation and matrotrophy

The combination of superfetation and moderate or extensive matrotrophy has not facilitated the evolution of larger offspring in poeciliid fishes. Several authors have suggested an evolutionary link between superfetation and matrotrophy and have proposed other potential benefits of their concurrent presence (Reznick & Miles, 1989; Pollux et al., 2009, 2014; Trexler & DeAngelis, 2010; Meredith et al., 2011). First, matrotrophy may increase total fecundity because initial egg size is small and females may fertilize a relatively large number of eggs without a large initial energetic investment. In contrast, lecithotrophic species are limited in the initial number of eggs that females can fertilize because a large amount of resources are needed to produce fully yolked eggs (Trexler, 1997; Trexler & DeAngelis, 2003, 2010; Marsh-Matthews et al., 2005; Marsh-Matthews, 2011). Thus, total fecundity may be limited in lecithotrophic species. However, matrotrophic females need constant food supply to actively transfer nutrients to developing embryos. Otherwise, matrotrophy could become a disadvantage if enough resources are not available to sustain developing offspring. Hence, matrotrophy could increase total fecundity in environments where food resources are constant and abundant. According to a theoretical model developed by Trexler & DeAngelis (2003, 2010), in these predictable and stable environments, the previous presence of superfetation would facilitate the evolution of matrotrophy. Recently, some studies have conducted empirical tests of this ‘resource-availability’ hypothesis (Trexler, 1997; Marsh-Matthews & Deaton, 2006; Banet & Reznick, 2008; Banet, Au & Reznick, 2010; Pollux & Reznick, 2011; Riesch et al., 2013).

Second, matrotrophy and superfetation could reduce the locomotory cost imposed by viviparity. During pregnancy, females gain mass and volume, which negatively affects escape response and swimming performance due to increased drag forces on the female body (Plaut, 2002; Ghalambor, Reznick & Walker, 2004; Langerhans & Reznick, 2010). Matrotrophy could reduce this reproductive burden because mature ova are small and, thus, initial space requirements are small (Thibault & Schultz, 1978; Marsh-Matthews, 2011). Superfetation might provide additional advantages because it further reduces reproductive allocation by allowing females to divide the total number of developing embryos into smaller broods (Thibault & Schultz, 1978; Zúñiga-Vega et al., 2010). Therefore, both matrotrophy and superfetation could result in thinner bodies without a reduction in the total number of offspring produced because females bear a combination of small early-staged embryos and fewer large late-staged embryos compared to lecithotrophic and non-superfetating species that must bear many large embryos (Miller, 1975; Thibault & Schultz, 1978; Zúñiga-Vega et al., 2007, 2010; Pollux et al., 2009; Pires et al., 2011a). Selective environments where a streamlined body shape is needed such as high-velocity water systems or habitats where fish must swim fast to escape from abundant predators might promote the evolution of both reproductive strategies. Zúñiga-Vega et al. (2007) provided evidence that supports this hypothesis using Poeciliopsis turrubarensis as a model system. However, they focused only on the adaptive value of superfetation without any emphasis on spatial variation in matrotrophy. Additional tests of this ‘locomotor cost’ hypothesis are needed, analyzing the effects of the concurrent presence of matrotrophy and superfetation on body streamlining and fitness in these particular selective environments.

Third, moderate and extensive matrotrophy have coevolved with specialized morphological structures that facilitate the active transfer of nutrients from mother to embryos (i.e. placentas; Turner, 1940; Jollie & Jollie, 1964; Knight et al., 1985; Grove & Wourms, 1994; Kwan et al., 2015). This enhanced physiological communication through placental tissues could facilitate the rise of parent-offspring conflicts during pregnancy (i.e. individual embryos demand a greater investment and mothers attempt to optimize the allocation to each offspring; Trivers, 1974; Zeh & Zeh, 2000, 2008; Crespi & Semeniuk, 2004; Schrader & Travis, 2008). Parent-offspring conflicts during embryo development may drive a shift from pre-copulatory female mate choice to post-copulatory mechanisms of sexual selection (Pollux et al., 2014). The capacity to control the amount of resources that are devoted to each offspring through matrotrophy and placentas would allow females to allocate more resources to those offspring that carry the best paternal genes. In addition, superfetation could further enhance post-copulatory female selection by facilitating multiple paternity because each brood may be fertilized by a different male (Travis, Trexler & Mulvey, 1990; Zane et al., 1999; Soucy & Travis, 2003). This potential to produce temporally overlapping, mixed-paternity broods might also increase genetic variability of the offspring (Macías-Garcia & González-Zuarth, 2005). Pollux et al. (2014) used phylogenetic comparative methods and a large data set (94 species) to suggest that the correlated evolution of superfetation and matrotrophy in the family Poeciliidae has allowed placental females to gain control over paternity, thereby relying in a less extent on pre-copulatory mechanisms of sexual selection.

Additional theoretical models have attempted to explain the fitness benefits of superfetation and matrotrophy, although considering them separately (Thibault, 1974; Downhower & Brown, 1975; Travis et al., 1987; Zúñiga-Vega et al., 2010; Pires et al., 2011b; Bassar, Auer & Reznick, 2014). However, their correlated evolution strongly suggests that the adaptive benefits of superfetation depend, at least to some degree, on the concurrent presence of matrotrophy and/or vice versa (Reznick & Miles, 1989; Pollux et al., 2009, 2014; Trexler & DeAngelis, 2010; Meredith et al., 2011). Therefore, further theoretical and empirical studies must focus on the combination of both traits when examining their potential adaptive value.

Supplementary Material

Supp TableS1-S2

ACKNOWLEDGEMENTS

This research was supported by the Mexican Research Council (Consejo Nacional de Ciencia y Tecnología, CONACyT) through a doctorate scholarship awarded to COT (368782/245650) and through the grant SEP-CONACyT-129675. AGOK was supported by a Common Themes in Reproductive Diversity training grant (NIH-NICD 5T32HD049336-10). This paper is a requisite for COT to obtain the Ph.D. degree in the Posgrado en Ciencias Biológicas of Universidad Nacional Autónoma de México. We thank Alecandria Ader, Alejandro Molina-Moctezuma, Jerald Johnson, Joseph Travis, Matthew Schrader, Mark Belk, Michael Jennions, Patricia Frías-Alvarez, and Rüdiger Riesch for providing us with their data sets and Norma Moreno-Mendoza and Maricela Villagrán-Santa Cruz for academic advice. We also thank Patricia Frías-Alvarez, Marcelo Pires, and the anonymous reviewers for their helpful comments.

REFERENCES

  • Alcaraz C, García-Berthou E. Life history variation of invasive mosquitofish (Gambusia holbrooki) along a salinity gradient. Biological Conservation. 2007;139:83–92. [Google Scholar]
  • Almeida-Silva PH, Mazzoni R. Life history aspects of Phalloceros anisophallos Lucinda, 2008 (Osteichthyes, Poeciliidae) from Córrego Andorinha, Ilha Grande (RJ, Brazil) Studies on Neotropical Fauna and Environment. 2014;49:1–8. [Google Scholar]
  • Arias AL, Reznick DN. Life history of Phalloceros caudiomaculatus: a novel variation on the theme of livebearing in the family Poeciliidae. Copeia. 2000;2000:792–798. [Google Scholar]
  • Banet AI, Au AG, Reznick DN. Is mom in charge? Implications of resource provisioning on the evolution of the placenta. Evolution. 2010;64:3172–3182. [Abstract] [Google Scholar]
  • Banet AI, Reznick DN. Do placental species abort offspring? Testing an assumption of the Trexler-DeAngelis model. Functional Ecology. 2008;22:323–331. [Google Scholar]
  • Bashey F. Cross-generational environmental effects and the evolution of offspring size in the Trinidadian guppy Poecilia reticulata . Evolution. 2006;60:348–361. [Abstract] [Google Scholar]
  • Bashey F. Competition as a selective mechanism for larger offspring size in guppies. Oikos. 2008;117:104–113. [Google Scholar]
  • Bassar RD, Auer SK, Reznick DN. Why do placentas evolve? A test of the life history facilitation hypothesis in two clades in the genus Poeciliopsis representing two independent origins of placentas. Functional Ecology. 2014;28:999–1010. [Google Scholar]
  • Brockelman WY. Competition, the fitness of offspring, and optimal clutch size. American Naturalist. 1975;109:677–699. [Google Scholar]
  • Brown-Peterson N, Peterson MS. Comparative life history of female mosquito fish, Gambusia affinis, in tidal freshwater and oligohaline habitats. Environmental Biology of Fishes. 1990;27:33–41. [Google Scholar]
  • Cheong RT, Henrich S, Farr JA, Travis J. Variation in fecundity and its relationship to body size in a population of the least killifish, Heterandria formosa (Pisces: Poeciliidae) Copeia. 1984;1984:720–726. [Google Scholar]
  • Cooley LR, Foighil DÓ. Phylogenetic analysis of the Sphaeriidae (Mollusca: Bivalvia) based on partial mitochondrial 16S rDNA gene sequences. Invertebrate Biology. 2000;119:299–308. [Google Scholar]
  • Crespi B, Semeniuk C. Parent offspring conflict in the evolution of vertebrate reproductive mode. American Naturalist. 2004;163:635–653. [Abstract] [Google Scholar]
  • Doadrio I, Perea S, Alcaraz L, Hernandez N. Molecular phylogeny and biogeography of the Cuban genus Girardinus Poey, 1854 and relationships within the tribe Girardinini (Actinopterygii, Poeciliidae) Molecular Phylogenetics and Evolution. 2009;50:16–30. [Abstract] [Google Scholar]
  • Downhower JF, Brown L. Superfetation in fishes and the cost of reproduction. Nature. 1975;256:345. [Google Scholar]
  • Downhower JF, Brown LP, Matsui ML. Life history variation in female Gambusia hubbsi . Environmental Biology of Fishes. 2000;59:415–428. [Google Scholar]
  • Felsenstein J. Phylogenies and the comparative method. American Naturalist. 1985;125:1–15. [Abstract] [Google Scholar]
  • Fraser EA, Renton RM. Observation on the breeding and development of the viviparous fish, Heterandria formosa . Quarterly Journal of Microscopical Science. 1940;81:479–520. [Google Scholar]
  • Frías-Alvarez P, Macías Garcia C, Vázquez-Vega LF, Zúñiga-Vega JJ. Spatial and temporal variation in superfoetation and related life history traits of two viviparous fishes: Poeciliopsis gracilis and P. infans . Naturwissenschaften. 2014;101:1085–1098. [Abstract] [Google Scholar]
  • Ghalambor CK, Reznick DN, Walker JA. Constraints on adaptive evolution: the functional trade off between reproduction and fast start swimming performance in the Trinidadian guppy (Poecilia reticulata) American Naturalist. 2004;164:38–50. [Abstract] [Google Scholar]
  • Gomes JL, Jr, Monteiro LR. Size and fecundity variation in populations of Poecilia vivipara Block & Schneider (Teleostei; Poeciliidae) inhabiting an environmental gradient. Journal of Fish Biology. 2007;71:1799–1809. [Google Scholar]
  • Gordon SP, Reznick DN, Kinnison MT, Bryant MJ, Weese DJ, Räsänen K, Millar NP, Hendry AP. Adaptive changes in life history and survival following a new guppy introduction. American Naturalist. 2009;174:34–45. [Abstract] [Google Scholar]
  • Grether GF, Kolluru GR. Evolutionary and plastics responses to resource availability. In: Evans JP, Pilastro A, Schlupp I, editors. Ecology, evolution of poeciliid fishes. Chicago: The University of Chicago Press; 2011. pp. 28–37. [Google Scholar]
  • Grove BD, Wourms JP. The follicular placenta of the viviparous fish, Heterandria formosa. I. Ultrastructure and development of the embryonic absorptive surface. Journal of Morphology. 1994;209:265–284. [Abstract] [Google Scholar]
  • Gunn JS, Thresher RE. Viviparity and the reproductive ecology of clinid fishes (Clinidae) from temperate Australian waters. Environmental Biology of Fishes. 1991;31:323–344. [Google Scholar]
  • Hansen TF. Stabilizing selection and the comparative analysis of adaptation. Evolution. 1997;51:1341–1351. [Abstract] [Google Scholar]
  • Harvey PH, Pagel MD. The comparative method in evolutionary biology. Oxford: Oxford University Press; 1991. [Google Scholar]
  • Haynes JL. Standardized classification of poeciliid development for life-history studies. Copeia. 1995;1995:147–154. [Google Scholar]
  • Housworth EA, Martins EP, Lynch M. The phylogenetic mixed model. American Naturalist. 2004;163:84–96. [Abstract] [Google Scholar]
  • Jennions M, Telford S. Life-history phenotypes in populations of Brachyrhaphis episcopi (Poeciliidae) with different predator communities. Oecologia. 2002;132:44–50. [Abstract] [Google Scholar]
  • Jennions MD, Wong BB, Cowling A, Donnelly C. Life-history phenotypes in a live-bearing fish Brachyrhaphis episcopi living under different predator regimes: seasonal effects? Environmental Biology of Fishes. 2006;76:211–219. [Google Scholar]
  • Johnson JB, Belk MC. Predation environment predicts divergent life-history phenotypes among populations of the livebearing fish Brachyrhaphis rhabdophora . Oecologia. 2001;126:142–149. [Abstract] [Google Scholar]
  • Jollie WP, Jollie LG. The fine structure of the ovarian follicle of the ovoviviparous poeciliid fish, Lebistes reticulatus. I. Maturation of follicular epithelium. Journal of Morphology. 1964;114:479–501. [Abstract] [Google Scholar]
  • Knight FM, Lombardi J, Wourms JP, Burns JR. Follicular placenta and embryonic growth of the viviparous four eyed fish (Anableps) Journal of Morphology. 1985;185:131–142. [Abstract] [Google Scholar]
  • Kwan L, Fris M, Rodd FH, Rowe L, Tuhela L, Panhuis TM. An examination of the variation in maternal placentae across the genus Poeciliopsis (Poeciliidae) Journal of Morphology. 2015 [Abstract] [Google Scholar]
  • Langerhans RB, Gifford ME, Domínguez Domínguez O, Garcí Bedoya D, DeWitt TJ. Gambusia quadruncus (Cyprinodontiformes: Poeciliidae): a new species of mosquitofish from east central México. Journal of Fish Biology. 2012;81:1514–1539. [Abstract] [Google Scholar]
  • Langerhans RB, Reznick DN. Ecology and evolution of swimming performance in fishes: predicting evolution with biomechanics. In: Domenici P, Kapoor BG, editors. Fish locomotion: an eco-ethological perspective. Boca Ratón: Science Publishers; 2010. pp. 200–248. [Google Scholar]
  • Leips J, Richardson JM, Rodd FH, Travis J. Adaptive maternal adjustments of offspring size in response to conspecific density in two populations of the least killifish, Heterandria formosa . Evolution. 2009;63:1341–1347. [Abstract] [Google Scholar]
  • Leips J, Travis J. The comparative expression of life history traits and its relationship to the numerical dynamics of four populations of the least killifish. Journal of Animal Ecology. 1999;68:595–616. [Google Scholar]
  • Lombardi J, Wourms JP. The trophotaenial placenta of viviparous goodeid fish. I. Ultrastructure of the internal ovarian epithelium, the maternal component. Journal of Morphology. 1985;184:277–292. [Abstract] [Google Scholar]
  • Lydeard C, Wooten MC, Meyer A. Molecules, morphology, and area cladograms: a cladistic and biogeographic analysis of Gambusia (Teleostei: Poeciliidae) Systematic Biology. 1995;44:221–236. [Google Scholar]
  • Lynch M. Methods for the analysis of comparative data in evolutionary biology. Evolution. 1991;45:1065–1080. [Abstract] [Google Scholar]
  • Macías-Garcia C, González-Zuarth CA. Reproductive behavior of viviparous fish and intersexual conflict. In: Uribe MC, Grier HJ, editors. Viviparous fishes. Homestead: New Life Publications; 2005. pp. 289–302. [Google Scholar]
  • Maddison WP, Maddison DR. Mesquite: a modular system for evolutionary analysis. Version, 2.1. 2009 Available at: http://mesquiteproject.wikispaces.com.
  • Marsh-Matthews E. Matrotropy. In: Evans JP, Pilastro A, Schlupp I, editors. Ecology and evolution of poeciliid fishes. Chicago: The University of Chicago Press; 2011. pp. 28–37. [Google Scholar]
  • Marsh-Matthews E, Brooks M, Deaton R, Tan H. Effects of maternal and embryo characteristics on post-fertilization provisioning in fishes of the genus Gambusia . Oecologia. 2005;144:12–24. [Abstract] [Google Scholar]
  • Marsh-Matthews E, Deaton R. Resources and offspring provisioning: a test of the Trexler-DeAngelis model for matrotrophy evolution. Ecology. 2006;87:3014–3020. [Abstract] [Google Scholar]
  • Martin SB, Hitch AT, Purcell KM, Klerks PL, Leberg PL. Life history variation along a salinity gradient in coastal marshes. Aquatic Biology. 2009;8:15–28. [Google Scholar]
  • Martins EP. Conducting phylogenetic comparative studies when the phylogeny is not known. Evolution. 1996;50:12–22. [Abstract] [Google Scholar]
  • Martins EP. COMPARE, version 4.6 b. 2004 Available at: http://www.indiana.edu/~martinsl/compare.
  • Martins EP, Garland T., Jr Phylogenetic analyses of the correlated evolution of continuous characters: a simulation study. Evolution. 1991;45:534–557. [Abstract] [Google Scholar]
  • Martins EP, Hansen TF. Phylogenies and the comparative method: a general approach to incorporating phylogenetic information into the analysis of interspecific data. American Naturalist. 1997;149:646–667. [Google Scholar]
  • Meffe GK. Offspring size variation in eastern mosquitofish (Gambusia holbrooki: Poeciliidae) from contrasting thermal environments. Copeia. 1990;1990:10–18. [Google Scholar]
  • Meredith RW, Pires MN, Reznick DN, Springer MS. Molecular phylogenetic relationships and the evolution of the placenta in Poecilia (Micropoecilia) (Poeciliidae: Cyprinodontiformes) Molecular Phylogenetics and Evolution. 2010;55:631–639. [Abstract] [Google Scholar]
  • Meredith RW, Pires MN, Reznick DN, Springer MS. Molecular phylogenetic relationships and the coevolution of placentotrophy and superfetation in (Poecilia) (Poeciliidae: Cyprinodontiformes) Molecular Phylogenetics and Evolution. 2011;59:148–157. [Abstract] [Google Scholar]
  • Meyer MK, Schories S, Schartl M. Description of Priapella lacandonae sp. n.-a new poeciliid fish from the río Tulijá basin, Grijalva system, Chiapas, Mexico (Teleostei: Cyprinodontiformes: Poeciliidae) Vertebrate Zoology. 2011;61:91–97. [Google Scholar]
  • Miller RR. Five new species of Mexican poeciliid fishes of the genera Poecilia Gambusia, and Poeciliopsis . Occasional Papers of the Museum of Zoology University of Michigan. 1975;672:1–44. [Google Scholar]
  • Mojica CL, Meyer A, Barlow GW. Phylogenetic relationships of species of the genus Brachyrhaphis (Poeciliidae) inferred from partial mitochondrial DNA sequences. Copeia. 1997;1997:298–305. [Google Scholar]
  • Molina-Moctezuma A. Poeciliopsis baenschi. Bachelor’s Thesis, Universidad Nacional Autónoma de México; 2011. Influencia de la depredación sobre las características de historias de vida y la dinámica poblacional de. [Google Scholar]
  • Molina-Moctezuma A. Relación entre la disponibilidad de recursos y el nivel de transferencia de nutrientes entre madres y embriones en dos especies de peces vivíparos. Master’s Thesis, Universidad Nacional Autónoma de México; 2015. [Google Scholar]
  • Morales-Cazan A, Albert JS. Monophyly of Heterandriini (Teleostei: Poeciliidae) revisited: a critical review of the data. Neotropical Ichthyology. 2012;10:19–44. [Google Scholar]
  • Mossman HW. Classics revisited: comparative morphogenesis of the fetal membranes and accessory uterine structures. Placenta. 1991;12:1–5. [Abstract] [Google Scholar]
  • Pires MN, Arendt J, Reznick DN. The evolution of placentas and superfetation in the fish genus Poecilia (Cyprinodontiformes: Poeciliidae: subgenera Micropoeciliia and Acanthophacelus) Biological Journal of the Linnean Society. 2010;99:784–796. [Google Scholar]
  • Pires MN, Banet AI, Pollux BJA, Reznick DN. Variation and evolution of reproductive strategies. In: Evans JP, Pilastro A, Schlupp I, editors. Ecology and evolution of poeciliid fishes. Chicago: The University of Chicago Press; 2011a. pp. 28–37. [Google Scholar]
  • Pires MN, Bassar RD, McBride KE, Regus JU, Garland T, Reznick DN. Why do placentas evolve? An evaluation of the life history facilitation hypothesis in the fish genus Poeciliopsis . Functional Ecology. 2011b;25:757–768. [Google Scholar]
  • Pires MN, McBride KE, Reznick DN. Interpopulation variation in life history traits of Poeciliopsis prolifica: implications for the study of placental evolution. Journal of Experimental Zoology Part A: Ecological Genetics and Physiology. 2007;307:113–125. [Abstract] [Google Scholar]
  • Plaut I. Does pregnancy affect swimming performance of female Mosquitofish, Gambusia affinis? Functional Ecology. 2002;16:290–295. [Google Scholar]
  • Pollux BJA, Meredith RW, Springer MS, Reznick DN. The evolution of the placenta drives a shift in sexual selection in livebearing fish. Nature. 2014;513:233–236. [Abstract] [Google Scholar]
  • Pollux BJA, Pires MN, Banet AI, Reznick DN. Evolution of placentas in the family Poeciliidae: an empirical study of macroevolution. Annual Review of Ecology, Evolution, and Systematics. 2009;40:271–289. [Google Scholar]
  • Pollux BJA, Reznick DN. Matrotrophy limits a female’s ability to adaptively adjust offspring size and fecundity in fluctuating environments. Functional Ecology. 2011;25:747–756. [Google Scholar]
  • Ptacek MB, Breden F. Phylogenetic relationships among the mollies (Poeciliidae: Poecilia: Mollienesia group) based on mitochondrial DNA sequences. Journal of Fish Biology. 1998;53:64–81. [Google Scholar]
  • Reznick DN. The impact of predation on life history evolution in Trinidadian guppies: genetic basis of observed life history patterns. Evolution. 1982;36:1236–1250. [Abstract] [Google Scholar]
  • Reznick DN, Bryga H, Endler JA. Experimentally induced life-history evolution in a natural population. Nature. 1990;346:357–359. [Google Scholar]
  • Reznick DN, Endler JA. The impact of predation on life history evolution in Trinidadian guppies (Poecilia reticulata) Evolution. 1982;36:160–177. [Abstract] [Google Scholar]
  • Reznick DN, Hrbek T, Caura S, De Greef J, Roff D. Life history of Xenodexia ctenolepis: implications for life history evolution in the family Poeciliidae. Biological Journal of the Linnean Society. 2007b;92:77–85. [Google Scholar]
  • Reznick DN, Mateos M, Springer MS. Independent origins and rapid evolution of the placenta in the fish genus Poeciliopsis . Science. 2002;298:1018–1020. [Abstract] [Google Scholar]
  • Reznick DN, Meredith R, Collette BB. Independent evolution of complex life history adaptations in two families of fishes, live bearing halfbeaks (Zenarchopteridae, Beloniformes) and Poeciliidae (Cyprinodontiformes) Evolution. 2007a;61:2570–2583. [Abstract] [Google Scholar]
  • Reznick DN, Miles DB. A review of life history patterns in Poeciliid fishes. In: Meffe GK, Snelson FF, editors. Ecology and evolution of livebearing fishes (Poeciliidae) Englewood Cliffs: Prentice Hall; 1989. pp. 125–148. [Google Scholar]
  • Riesch R, Colston TJ, Joachim BL, Schlupp I. Natural history and life history of the Grijalva gambusia Heterophallus milleri Radda, 1987 (Teleostei: Poeciliidae) Aqua. 2011;17:95–102. [Google Scholar]
  • Riesch R, Martin RA, Bierbach D, Plath M, Langerhans RB, Arias-Rodriguez L. Natural history, life history, and diet of Priapella chamulae Schartl, Meyer & Wilde 2006 (Teleostei: Poeciliidae) Aqua. 2012;18:95–102. [Google Scholar]
  • Riesch R, Martin RA, Langerhans RB. Predation’s role in life-history evolution of a livebearing fish and a test of the Trexler-DeAngelis model of maternal provisioning. American Naturalist. 2013;181:78–93. [Abstract] [Google Scholar]
  • Riesch R, Plath M, García de León FJ, Schlupp I. Convergent life-history shifts: toxic environments result in big babies in two clades of poeciliids. Naturwissenschaften. 2010a;97:133–141. [Abstract] [Google Scholar]
  • Riesch R, Plath M, Schlupp I. Toxic hydrogen sulfide and dark caves: life-history adaptations in a livebearing fish (Poecilia mexicana, Poeciliidae) Ecology. 2010b;91:1494–1505. [Abstract] [Google Scholar]
  • Roellig K, Menzies BR, Hildebrandt TB, Goeritz F. The concept of superfetation: a critical review on a ‘myth’ in mammalian reproduction. Biological Reviews. 2011;86:77–95. [Abstract] [Google Scholar]
  • Schrader M, Travis J. Testing the viviparity driven conflict hypothesis: parent offspring conflict and the evolution of reproductive isolation in a poeciliid fish. American Naturalist. 2008;172:806–817. [Abstract] [Google Scholar]
  • Schrader M, Travis J. Assessing the roles of population density and predation risk in the evolution of offspring size in populations of a placental fish. Ecology and Evolution. 2012;2:1480–1490. [Europe PMC free article] [Abstract] [Google Scholar]
  • Schultz RJ. Reproductive mechanism of unisexual and bisexual strains of the viviparous fish Poeciliopsis . Evolution. 1961;15:302–325. [Google Scholar]
  • Scrimshaw NS. Superfetation in poeciliid fishes. Copeia. 1944;1944:180–183. [Google Scholar]
  • Shields PA, Carlson SR. Effects of formalin and alcohol preservation on lengths and weights of juvenile sockeye salmon. Alaska Fishery Research Bulletin. 1996;3:81–93. [Google Scholar]
  • Smith CC, Fretwell SD. The optimal balance between size and number of offspring. American Naturalist. 1974;108:499–506. [Google Scholar]
  • Soucy S, Travis J. Multiple paternity and population genetic structure in natural populations of the poeciliid fish, Heterandria formosa . Journal of Evolutionary Biology. 2003;16:1328–1336. [Abstract] [Google Scholar]
  • Stearns SC, Sage RD. Maladaptation in a marginal population of the mosquito fish, Gambusia affinis . Evolution. 1980;34:65–75. [Abstract] [Google Scholar]
  • Stearns SC. Life-history tactics: a review of the ideas. Quarterly Review of Biology. 1976;51:3–47. [Abstract] [Google Scholar]
  • Stearns SC. The evolution of life-history traits in mosquitofish since their introduction to Hawaii in 1905: rates of evolution, heritabilities, and developmental plasticity. American Zoologist. 1983;23:65–75. [Google Scholar]
  • Thibault RE. Genetics of cannibalism in a viviparous fish and its relationship to population density. Nature. 1974;251:138–140. [Google Scholar]
  • Thibault RE, Schultz RJ. Reproductive adaptations among viviparous fishes (Cyprynodontiformes: Poeciliidae) Evolution. 1978;32:320–333. [Abstract] [Google Scholar]
  • Travis J, Farr JA, Henrich S, Cheong RT. Testing theories of clutch overlap with the reproductive ecology of Heterandria formosa . Ecology. 1987;68:611–623. [Google Scholar]
  • Travis J, Trexler JC, Mulvey M. Multiple paternity and its correlates in female Poecilia latipinna (Poeciliidae) Copeia. 1990;1990:722–729. [Google Scholar]
  • Trexler JC. Variation in the degree of viviparity in the sailfin molly, Poecilia latipinna. Copeia. 1985;1985:999–1004. [Google Scholar]
  • Trexler JC. Resource availability and plasticity in offspring provisioning: embryo nourishment in sailfin mollies. Ecology. 1997;78:1370–1381. [Google Scholar]
  • Trexler JC, DeAngelis DL. Resource allocation in offspring provisioning: an evaluation of the conditions favoring the evolution of matrotrophy. American Naturalist. 2003;162:574–585. [Abstract] [Google Scholar]
  • Trexler JC, DeAngelis DL. Modeling the evolution of complex reproductive adaptations in poeciliid fishes: matrotrophy and superfetation. In: Uribe MC, Grier HJ, editors. Viviparous fishes II. Homestead: New Life Publications; 2010. pp. 231–240. [Google Scholar]
  • Trivers RL. Parent-offspring conflict. American Zoologist. 1974;14:249–264. [Google Scholar]
  • Turner CL. Reproductive cycles and superfetation in poeciliid fishes. Biological Bulletin. 1937;72:145–164. [Google Scholar]
  • Turner CL. Pseudoamnion, pseudochorion, and follicular pseudoplacenta in poeciliid fishes. Journal of Morphology. 1940;67:59–89. [Google Scholar]
  • Turner JS, Snelson FF., Jr Population structure, reproduction and laboratory behavior of the introduced Belonesox belizanus (Poeciliidae) in Florida. Environmental Biology of Fishes. 1984;10:89–100. [Google Scholar]
  • Weeks SC, Gaggiotti OE. Patterns of offspring size at birth in clonal and sexual strains of Poeciliopsis (Poeciliidae) Copeia. 1993;1993:1003–1009. [Google Scholar]
  • Weldele ML, Zúñiga-Vega JJ, Johnson JB. Life history of Gambusia vittata (Pisces: Poeciliidae) Southwestern Naturalist. In press [Google Scholar]
  • Wourms JP. Viviparity: the maternal fetal relationships in fishes. American Zoologist. 1981;21:473–515. [Google Scholar]
  • Wourms JP, Grove BD, Lombardi J. The maternal-embryonic relationship in viviparous fishes. In: Hoar WS, Randall DJ, editors. Fish physiology: the physiology of developing fish. New York: Academic Press; 1988. pp. 1–134. [Google Scholar]
  • Zane L, Nelson WS, Jones AG, Avise JC. Microsatellite assessment of multiple paternity in natural populations of a live-bearing fish, Gambusia holbrooki . Journal of Evolutionary Biology. 1999;12:61–69. [Google Scholar]
  • Zeh DW, Zeh JA. Reproductive mode and speciation: the viviparity-driven conflict hypothesis. Bioessays. 2000;22:938–946. [Abstract] [Google Scholar]
  • Zeh JA, Zeh DW. Viviparity driven conflict. Annals of the New York Academy of Sciences. 2008;1133:126–148. [Abstract] [Google Scholar]
  • Zúñiga-Vega JJ, Macías-Garcia C, Johnson JB. Hypotheses to explain the evolution of superfetation in viviparous fishes. In: Uribe MC, Grier HJ, editors. Viviparous fishes II. Homestead: New Life Publications; 2010. pp. 241–254. [Google Scholar]
  • Zúñiga-Vega JJ, Reznick DN, Johnson JB. Habitat predicts reproductive superfetation and body shape in the livebearing fish Poeciliopsis turrubarensis . Oikos. 2007;116:995–1005. [Google Scholar]
  • Zúñiga Vega JJ, Suárez Rodríguez M, Espinos Pérez H, Johnson JB. Morphological and reproductive variation among populations of the Pacific molly Poecilia butleri . Journal of Fish Biology. 2011;79:1029–1046. [Abstract] [Google Scholar]

Full text links 

Read article at publisher's site: https://doi.org/10.1111/bij.12662

Read article for free, from open access legal sources, via Unpaywall: https://academic.oup.com/biolinnean/article-pdf/116/4/787/16714217/bij12662.pdfUnpaywall

Citations & impact 

Impact metrics

2
Citations
Jump to Citations

Smart citations by scite.ai
Smart citations by scite.ai include citation statements extracted from the full text of the citing article. The number of the statements may be higher than the number of citations provided by EuropePMC if one paper cites another multiple times or lower if scite has not yet processed some of the citing articles.
Explore citation contexts and check if this article has been supported or disputed.
https://scite.ai/reports/10.1111/bij.12662

Supporting
Mentioning
Contrasting
0
18
0

Article citations

Similar Articles 

To arrive at the top five similar articles we use a word-weighted algorithm to compare words from the Title and Abstract of each citation.

Funding 

Funders who supported this work.

Common Themes in Reproductive Diversity (1)

Mexican Research Council (2)

NICHD NIH HHS (1)

Posgrado en Ciencias Biológicas of Universidad Nacional Autónoma de México