Advertent domestication attenuates the influence of propagule pressure

doi:10.21203/rs.3.rs-2753589/v1

Download PDF

Research Article

Advertent domestication attenuates the influence of propagule pressure

https://doi.org/10.21203/rs.3.rs-2753589/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

The process of domestication affects fitness following return to the wild. For the invasion of non-native species, however, captive rearing is thought to increase propagule pressure, the quantity and rate that individuals are introduced. Invasion success for domesticated species may represent a balance between survival and propagule pressure. Survival is likely affected by selective breeding (advertent versus inadvertent selection) and predator populations, which contribute to biotic resistance, the ability of communities to resist invasion. Ornamental species are subjected to deliberate selection (advertent) for phenotypic traits such as coloration, as is the case with the green swordtail (Xiphophorus hellerii). Wild-type varieties are also produced without deliberate selection (inadvertent) and resemble their wild counterparts. In Florida, swordtails are produced in aquaculture and propagule pressure is high, yet few colorful individuals are encountered in the wild. Here we examined how invasion success is influenced by the interactive effects of biotic resistance, selective breeding, and propagule pressure. We used outdoor ponds to examine intraspecific variation in invasion success for five swordtail varieties across increasing biotic resistance. Propagule pressure over 14 weeks was varied proportional to trade volume for the five varieties. Biotic resistance increased with community complexity and affected swordtail survival and reproduction. In control ponds the number of fish for each variety followed propagule size. Despite lower propagule pressure, the wild-type variety increased relative to the advertently produced varieties, but only in ponds with greater biotic resistance. These results suggest propagule pressure is attenuated by increased biotic resistance and deliberate breeding.

Advertent selection

Biotic resistance

Inadvertent selection

Gambusia holbrooki

Micropterus salmoides

Xiphophorus hellerii

Domestication is a process whereby organisms become adapted to the rearing environment (Price 2002; Lorenzen et al. 2012). Genetic changes during domestication can be attributed to genetic drift, inbreeding, and selection, which can be divided according to whether selection is intentional due to selective breeding (advertent) or unintentional (inadvertent), a byproduct of the rearing environment (Hutchings and Fraser 2008). Advertent selection is the cornerstone of rearing for food and ornamental purposes (Gjedrem et al. 2012), and in aquaculture settings, selective breeding is used to augment desirable traits such as growth, fecundity, and vibrant colors, with potential positive and negative fitness consequences upon return to the wild. Even when advertent selection is non-existent and inadvertent selection is limited in scope, e.g., a reduced number of generations, artificial rearing can lead to poor survival in the wild (Araki et al. 2007, 2008; Hutchings and Fraser 2008; Christie et al. 2012). For the invasion of domesticated species, by comparison, emphasis is placed on the importance of increased propagule pressure due to production and trade (Magalhães and Jacobi 2013), a widely recognized contributor to invasion success (Lockwood et al. 2005; Colautti et al. 2006a; Simberloff 2009). Increased propagule pressure along with maladaptation might represent a balance; understanding this balance is necessary for predicting invasion success of domesticated species.

Propagule pressure, the quantity (i.e., propagule size) and rate (i.e., propagule number) individuals are introduced to the wild (Simberloff 2009), can be considered a “null model” for species invasions (Colautti et al. 2006b), with deviations from expectations due to species traits and the abiotic or biotic environment (Hill and Tuckett 2018). Food and ornamental aquaculture relies on the production of non-native species, which should increase propagule pressure (Tuckett et al. 2017). There is a long history of rearing tropical ornamental species in Miami, Tampa Bay, and elsewhere in the state of Florida, USA. This industry produces a large number of varieties and species, some of which escape from ornamental fish farms (Tuckett et al. 2017). Compared to the number of species reared (Hill and Yanong 2016), however, relatively few ornamental fish are established in Florida (Schofield and Loftus 2015), despite appearing to exhibit the requisite characteristics to persist, including tolerance to relatively cold winters (Tuckett et al. 2016, 2021b). This suggests invasion limitation by biological communities (i.e., biotic resistance), representing a possible explanation for failed invasions and deviation from expectations based on propagule pressure.

Biotic resistance theory has a long history of testing and has been commonly identified as a relationship between diversity and invasibility (Elton 1958; Levine and D’Antonio 1999); other interpretations emphasize strongly-interacting species (Henriksson et al. 2015; Tuckett et al. 2021a), especially by native predators (Rogers et al. 2016; Yorisue et al. 2019). These latter studies emphasize the increased probability of including a species that is a strong resistor to invasion as diversity increases (Henriksson et al. 2015). Many ornamental species are selectively bred for specific phenotypic characteristics such as bright coloration, patterns, and fin shape or size, all of which may increase predation vulnerability and biotic resistance to invasion (Hill et al. 2011, 2014; Thompson et al. 2012; Howard et al. 2015). Thus, biotic resistance may have a strong effect on which ornamental species become established, particularly when strong resisting species are encountered.

The green swordtail (Xiphophorus hellerii) is native to Central America and exhibits a dullish coloration often with a red stripe and is distinguished by a long “sword-like” caudal fin appendage. In Florida, the green swordtail a widely produced species with ~ 100 recognized varieties in trade that are intentionally bred for various vibrant color patterns (Tuckett et al. 2021b). “Wild-type” variety swordtails are also produced and, due to the lack of deliberate selection, exhibit phenotypic traits similar to their wild counterparts. Trade volume and local production is dominated by the highly selected colorful swordtail varieties. It is then a valid assumption that these selected varieties experience greater levels of propagule pressure compared to the wild-type variety at production facility effluents.

Previous research has shown the green swordtail exhibits persistent populations in roadside ditches and small streams in the Tampa Bay area of Florida (Tuckett et al. 2016; Lawson et al. 2017). With ongoing propagule pressure as swordtails escape from ornamental aquaculture facilities (Tuckett et al. 2017), the expectation would be a greater abundance of the selected varieties relative to that of the wild-type variety. To the contrary, feral green swordtail populations in Florida are dominated by the wild-type variety (Tuckett et al. 2017, 2021b). In addition, as distance from the propagule source (aquaculture effluent) increases, the green swordtail population composition becomes increasingly dominated by the wild type (Tuckett et al. 2017, 2021b; Hill and Tuckett 2018). We suspect that biotic resistance increases when escaped fish move away from the aquaculture facilities which exerts a greater influence on swordtail population composition and success. We propose that green swordtails selected for vivid color patterns have reduced fitness compared to the wild-type variety due to increased predation vulnerability. This results in biotic resistance effectively mitigating invasion success as would otherwise be expected from substantial propagule pressure.

For this study, we tested the hypothesis that biotic resistance (i.e., the presence and interactions of native predatory fish) and the effects of advertent selection would overcome propagule pressure to affect the survival and relative abundance of non-native domesticated swordtail varieties in ponds. Biotic resistance was represented by an additive design of increasing species diversity ranging from a swordtail only control to treatments with multiple native predatory species. This design is similar to what a feral swordtail would experience moving at increasing distance from an aquaculture facility as community complexity (both habitat and species) increases. Differences in swordtail varietal propagule pressure was characterized by differing stocking rates of each variety proportional to their production volume in the Florida ornamental fish industry. The influence of advertent selection was inferred by comparing the relative success of the wild-type variety with the selected varieties. We predicted that swordtail survival would decrease with increasing biotic resistance. As an indication of reduced fitness, we expected that the wild-type variety would increase its abundance relative to the other varieties advertently selected for bright coloration due to their increased predation vulnerability.

Survival and population growth of swordtail varieties were examined across increasing community complexity in experimental outdoor ponds (each treatment, n = 4). Treatments included control (swordtails only), simple communities with swordtails and a poeciliid congener (eastern mosquitofish; Gambusia holbrooki; hereafter low) to increasing complexity with weak piscivores (swordtails + mosquitofish + sunfishes Lepomis spp.; medium) and a strong piscivore (swordtails + mosquitofish + sunfishes + largemouth bass Micropterus salmoides; high). This additive sequence of treatments is similar to what a non-native fish would experience when moving from an aquaculture effluent to more natural environments at increasing distance (Hill and Tuckett 2018). Thus, our goal was to mimic natural systems by including species commonly fund near aquaculture effluents and not simply increase diversity as would be expected from a study examining diversity-invasibility.

Pond preparation began in early August 2014 by applying an aquatic herbicide, to ensure ponds began with a similar vegetative community. A fish toxicant was applied on 08-Aug-2014 using recommended dosages to ensure all fish were removed prior to experimentation (Finlayson et al. 2010; Lazur et al. 2013). Ponds were then filled with aerated well water and also groundwater infiltration. Mosquitofish were stocked on 20-Aug-2014 (1000 individuals of various size per pond). A subsample of 100 individuals from the total stock were individually weighed and measured prior to stocking; mosquitofish length (total length; TL) averaged 29.57 mm (SD = 6.93) and weight averaged 0.35 g (SD = 0.29). Average pond size (wetted area) was 168 m²; thus, mosquitofish were stocked at an average density of 6.05 / m², a lower density than local streams and ditches and lower than that used in a previous experimental study (Thompson et al. 2012). However, mosquitofish typically utilize vegetative habitats, which will be more common around the pond perimeter (Moyle and Nichols 1973; Pyke 2008). Pond size and the resulting mosquitofish beginning density was similar among treatment groups (pond size: F_3,12 = 0.537; P = 0.666; density: F_2,9 = 0.583; P = 0.638). Ponds lacking a top predator (medium treatments) were stocked with ten bluegills (Lepomis macrochirus), one redbreast sunfish (Lepomis auritus), and one spotted sunfish (Lepomis punctatus), approximating the density and proportion of sunfishes near sources of production (Hill and Tuckett 2018). Ponds with the top predator (high), received the same compliment of bluegill, redbreast sunfish, and spotted sunfish, but also received two largemouth bass. Initial sunfish size (TL) was similar among ponds (nested ANOVA, pond nested within treatment; F_6,88 = 0.784; P = 0.533) and between medium (average = 127.46 mm) and high treatments (mean = 123.94 mm; F_1,88 = 0.530; P = 0.784). Similarly, mean bass TL among ponds varied from 258 to 304, but was similar among the four ponds (F_3,4 = 0.365; P = 0.783).

Periodic propagule pressure was manipulated using weekly stockings of swordtails beginning on 28-Aug-2014; stocking continued for an additional 13 weeks for a total of 14 stocking dates. Twenty swordtails were stocked weekly, and each pond received 280 total swordtails. The proportion of stocked swordtails mirrored trade data and presumably escapement from aquaculture facilities. The top six varieties sold by a local wholesaler were the velvet wag (33.7%), velvet (21.9), neon (18.6), assorted (17.2; not classified to variety), pineapple (14.3), and wild-type variety (11.5). Velvet wag and velvet varieties were unavailable from producers during the required timeframe and were replaced with similarly vibrant varieties, marigold and painted. Seven marigold, four painted, four neon, three pineapple, and two wild type were stocked weekly in each pond (Fig. 1). Thus, propagule pressure was varied among varieties but was held constant among all treatments. A subsample of ten fish from the five varieties were measured for TL. Fish varied in size from 35 to 60 mm (mean = 44.84 ± 6.57 SD) and size was similar among varieties (F_4,45 = 1.002; P = 0.417). All swordtail varieties, including the wild-type variety, were sourced from the same local producer using similar rearing protocols. Thus, the wild-type experienced similar rearing conditions as the more colorful varieties but was not deliberately bred by the producer for color patterns. Water quality parameters and temperature were monitored weekly, but always stayed within acceptable levels (Online Resource).

Ponds were randomly harvested over a three-day period, from 01-Dec-2014 to 03-Dec-2014, by first reducing the pool volume ~ 75% by pumping water through a screened intake; pond pumping concentrates fish and aids in capture. Fishes were captured using a fine-mesh seine net (1.59 mm) with three passes. A 1.59 mm mesh seine was used because it captures all but the smallest mosquitofish and swordtails. All captured fish except bass and sunfish were euthanized following institutional protocols and subsequently frozen for later analysis. Adult bass and sunfish were measured (TL) and removed to an on-site detention pond. Relative growth in TL was calculated for bass (relative growth = TL final – TL initial / TL initial). Juvenile sunfish were retained, weighed, and measured. Because of the large numbers of mosquitofish, up to 100 individuals for each pond were subsampled for length (TL) and weight (g); the remaining individuals were batch weighed to estimate numbers at the conclusion of the trial. All swordtails were measured, identified to variety, and sex was determined based on the presence of a gonopodium.

Predation Experiments

Two separate predation experiments were performed to determine bass prey selection: 1) between the swordtail and mosquitofish and 2) among swordtail varieties. Experiment #1 was conducted in an indoor re-circulating tank system. The tanks were concrete (221 cm x 79 cm x 58 cm) with a water depth of 30 cm. Water parameters fell within acceptable levels (Online Resource). All tanks had artificial vegetation as a refuge (645 stems m^− 2) which consisted of black plastic strips tied to a plastic screen that covered 50% of the tank bottom (Savino and Stein 1989). Twenty each of marigold swordtails (TL = 32 ± 6 mm) and mosquitofish (TL = 33 ± 4 mm) were stocked concurrently into the tanks. A low density of mosquitofish was used because previous research showed that effects of mosquitofish (i.e., fin nipping of ornamental fish) are density-dependent (Thompson et al. 2012). Prey were stocked concurrently to provide a similar acclimation period but also because the effects of mosquitofish can be delayed when stocked concurrently with other small-bodied fishes (Thompson et al. 2012). A single bass per tank (TL = 235 ± 28 mm) was stocked 1 hour after the prey. Trials (n = 12) lasted 24 hours. All fish were counted at the end of the trials.

Experiment #2 used the same experimental system exhibiting similar water quality parameters. Ten each of marigold, neon, painted, pineapple, and wild-type swordtails were stocked concurrently into the tanks. Swordtails for experiment #2 were taken from the same source as the pond study and thus exhibited similar size variation. Bass (TL = 317 ± 35 mm) were stocked as above and included eight trials, each lasting 24 hours. Bass in experiment #2 were not used in experiment #1. All swordtails were counted at the conclusion of the trials. For both experiments the “case 2” approximation of Chesson’s α was used as a selection index (Chesson 1983; see Hill et al. 2011), which accounts for the depletion of prey through consumption without replacement:

where m is the number of prey categories, r_i0 is the number of prey type, i is the diet, and n_i0 is the number of prey type i in the environment. Values of α can range between 0 indicating complete avoidance and 1 indicate complete preference. Random prey selection in this experiment (1/m) equaled 0.5 for experiment #1 and 0.2 for experiment #2.

Data Analysis

Our general approach for the analysis of count data was to first fit the data to the Poisson distribution. Alternant distributions were applied to overdispersed data (quasi-Poisson; negative binomial) and the model fit was evaluated using multiple criteria. For the pond study, we first analyzed counts of fish within the artificial community. The number of mosquitofish at the conclusion of the experiment with a GLM (Poisson distribution), but our data were overdispersed (McCullagh and Nelder 1989), whereupon we fit the data to the negative binomial distribution. Mosquitofish size (logarithm transformed TL; total n = 1296) was analyzed with nested GLM (normal distribution; identity link) with factors treatment (low, medium, and high) and pond nested within treatment. Counts of surviving sunfish (pooled among species) and also young of the year (reproduction in the ponds) were both analyzed with GLM (Poisson distribution). Similar to the mosquitofish counts, swordtail counts across levels of biotic resistance were analyzed with GLM (negative binomial distribution).

For the analysis of variation in success of the five varieties across pond treatments, some individuals could not be identified to a specific variety. Most of these individuals were found in control ponds (324 of 339) due to reproduction, which could have been due to hybridization between varieties. Varietal survival (square root transformed # remaining/# stocked) was analyzed with a mixed-model ANOVA with fixed effects treatment, variety, and interaction. Pond was included as a random factor nested within treatment. The control treatment was analyzed separate from the other three treatments because of dramatic differences in variance and to streamline the analysis for the fixed effect, variety. For the control treatment we analyzed both the relative success (square root transformed # remaining/# stocked) and the total counts to determine if propagule pressure increased numbers in the experimental ponds. In both analyses, pond was included as a random factor. Tukey HSD was used to analyze pairwise differences.

Selectivity values were compared using the paired t-test (experiment #1) or within-subjects ANOVA on transformed selectivity values (log (x + 1); experiment #2) because selection values within each tank were not independent. The within-subjects ANOVA included the fixed effect of variety and tank was included as the random effect. For experiment #2, two bass ate more than half of the swordtail prey; one of these bass consumed all 50 swordtails within a 24-hour period, significantly more than expected. Because of a potential saturation effect of predation where the predator could switch to a less desirable or visible prey as the preferred prey becomes scarce (Allen and Greenwood 1988), we excluded both of these trials. Thus, our final replicate number was just six.

Pond Study

The stocked fish community developed over the course of the experiment and reproduction was evident for mosquitofish, sunfishes, and swordtails. Mosquitofish reproduction, as evidenced by the presence of neonates, was observed in all 12 stocked ponds within one week. Final mosquitofish numbers exceeded the number stocked in all treatment ponds and the number of mosquitofish was greatest in low (mean = 6533.49 ± 1610.80 SE; χ²_2,7 = 7.81; P = 0.020; Online Resource 1), followed by high (5426.46 ± 1337.95) and medium treatments (2599.34 ± 641.16), which exhibited over 50% fewer individuals than the low or high treatment. However, post-hoc multiple comparisons among treatments were not significant (P > 0.050). Mosquitofish TL varied among ponds (F_9,1284 = 32.62; P < 0.001) and among all three treatments (F_2,1284 = 140.35; P < 0.001; Online Resource 1). Mosquitofish size was greatest in medium treatments (LSMean = 29.43 cm ± 0.34 SE) followed by low (24.76 ± 0.35); the smallest fish were found in high treatments (21.09 ± 0.33). The number of stocked sunfish surviving to the conclusion of the experiment was low (pooled among the three species), especially in medium treatments without bass (mean = 3.50 ± 0.79 SE); however, the number of surviving sunfish did not differ between high (6.00 ± 0.79) or medium treatments (χ²_1,5 = 2.57; P = 0.109). Bluegill reproduction was evident in all eight medium and high ponds and the number of juvenile bluegill was greater without bass (χ²_1,5 = 1105.89; P < 0.001; medium = 717.10 ± 194.31 SE; high = 178.41 ± 194.31). Bass survival was high, with seven of eight stocked fish surviving to the conclusion of the experiment (88%). Relative growth for individual bass (TL) varied from 10–26%.

The number of swordtails was greatest in control (GLM, negative binomial distribution; χ²_3,10 = 244.14; P < 0.001; Fig. 2) followed by low and medium ponds. High treatments exhibited just ten total surviving swordtails among the four replicates; one pond exhibited zero surviving swordtails. The sex ratio of surviving swordtails (female:male) varied among treatments (F_3,11 = 6.62; P = 0.008), was highest for control (mean = 7.40 ± 0.90 SE), followed by low (3.43 ± 0.80), medium (3.37 ± 0.90), and high (mean = 1.25 ± 1.27 SE). Sex ration for the control treatment differed from the remaining treatments, which were all similar (Tukey HSD). Swordtails were smaller in control ponds (mean = 31.98 cm ± 0.28 SE; F_3,2435 = 110.54; P < 0.001), followed by high (38.57 ± 3.71), low (40.57 ± 0.58), and medium (42.59 ± 0.69). Numerous individuals in the control treatments fell below the stocked size (Fig. 2).

Of the 2450 total swordtails captured (3360 total stocked), 13.8% were not classified to variety because they had characteristics of two varieties and were probable hybrids produced during the trial. The wild-type variety represented 7.1% of all stocked fish but increased to 11.7% of all harvested fish. As a proportion of the total swordtails captured, the wild type increased in three out of four control ponds (mean = 10.4%), four out of four low (17.9%) and medium (17.1%) ponds, and three out of three high ponds (50%; one high pond had zero surviving swordtails). In control ponds, the relative number of swordtails differed among varieties and was greatest for the painted followed by the marigold (F_4,12 = 5.97; P = 0.008; Fig. 3). Further, in control ponds, propagule size strongly increased the number of swordtails (F_1,15 = 11.55; P = 0.004; Fig. 3). For the remaining three treatments, the relative number of swordtails varied by treatment (F_2,9 = 80.82; P < 0.001; Fig. 4) and variety (F_4,36 = 18.11; P < 0.001); no treatment*variety interaction was evident (F_8,36 = 1.09; P = 0.390). The relative number of swordtails was greatest for the wild-type variety, which differed from all other varieties. The remaining four varieties were similar.

Prey Selection

For experiment #1, bass exhibited almost complete preference for marigold swordtails over mosquitofish (t_0.05,11 = 9.33; P < 0.001; Fig. 5). On average, each bass ate 9.6 swordtails and 1.7 mosquitofish. In experiment #2, bass selectivity values did not vary among swordtail varieties (within-subjects ANOVA; F_4,25 = 5.18; P = 0.168; Fig. 5). However, the wild-type variety exhibited negative selection (selected against; α < 0.2) and neon and painted varieties exhibited positive selection (α > 0.2). Error bars for marigold and pineapple varieties overlapped random varietal selection (α = 0.2).

In this study, biotic resistance, in the form of increasing community complexity, attenuated the influence of propagule pressure of ornamental swordtail varieties stocked in ponds. The control treatment followed null expectations where the absence of biotic resistance resulted in swordtail varietal relative abundance remaining similar to propagule pressure. Treatment effects of increasing biotic resistance were shown to reduce survival of all swordtail varieties. Negative effects of advertent selection were shown by the resulting abundance of the wild-type swordtail variety increasing relative to the advertently produced, colorful swordtail varieties. Similarly, the prey selection studies showed largemouth bass preferentially selected swordtails over mosquitofish and that the colorful swordtail varieties were more susceptible to predation.

Invasion success should increase for species brought into commercial production by increasing propagule pressure (Duggan et al. 2006; Magalhães and Jacobi 2013), but this may be balanced by biotic resistance and the fitness costs due to domestication (Gering et al. 2019a). Looking across all treatments, our results suggest biotic resistance, domestication selection, and propagule pressure all determined the outcome of the final non-native swordtail community. It may seem obvious that top predators would significantly affect prey community structure, but the extent of these interactions within the context of non-native species invasion success is not well understood. Using swordtails in ponds as our model we can conclude that invasion outcomes for domesticated small-bodied ornamental fish will depend upon 1) varietal trade volume and propagule pressure, 2) the extent of domestication selection, and 3) the degree to which biotic resistance interacts with domesticated phenotypes.

Propagule pressure

Propagule pressure can have a dominant influence on invasion success (Lockwood et al. 2005; Colautti et al. 2006a; Simberloff 2009). When propagule pressure is unknown, commercial production and trade availability has been used as a surrogate to estimate propagule pressure (Rixon et al. 2005; Gertzen et al. 2008). Propagule pressure can also be inferred from the number of individuals captured near sources of production (Tuckett et al. 2017) or availability at the retail level (Strecker et al. 2011). In close proximity to ornamental aquaculture facilities in Florida, this trend holds true where colorful feral swordtails are plentiful and varietal relative abundance is proportional to trade volume and local retail availability (Tuckett et al. 2021b).

For this pond study, we replicated field observations of swordtail propagule pressure by weekly stockings of swordtail varieties proportional to their production volume. Thus, swordtail varieties with higher production volumes were stocked in greater numbers relative to the varieties with less production volume. The wild-type variety (lowest trade volume) was stocked at a rate of just two individuals per week in each pond compared to the marigold variety (highest trade volume) stocked at a rate of seven individuals per week per pond. Therefore, the colorful varieties (marigold, painted, neon, and pineapple) should have a numerical advantage and dominate the resulting swordtail assemblages at the conclusion of the experiment. Indeed, in the control ponds where biotic resistance was limited (no other fish species stocked) our results followed null expectations. Similarly, near discharge points of ornamental aquaculture facilities in the Tampa Bay area biotic resistance is limited with few native species, predators, or community complexity. Treatments with predators did not follow null expectations where final swordtail populations did not follow trends expected based on propagule pressure.

Previous research indicates that as distance from the production facility increases, fish community complexity increases but swordtail varietal diversity and swordtail abundance decreases (Hill and Tuckett 2018; Tuckett et al. 2021b). Our results showed a similar trend, where the wild-type variety increased abundance relative to the selected varieties and associated propagule pressures in all treatments. In the highest treatment, all swordtail populations were significantly reduced (eliminated in one replicate) indicating significant susceptibility to biotic resistance (Hill and Tuckett 2018). It would be of interest to further test what level of propagule pressure would be required to overcome the influence of biotic resistance, which may be a large number for brightly colored ornamental fish.

Domestication

Advertent selection for enhanced color incurred negative survival costs and significant and recognizable phenotypic variation in color patterns due to advertent selection is readily observed in many ornamental species (Hill et al. 2014, 2017; Tuckett et al. 2016). This should have consequences for fitness and ultimately invasion success. Beyond selective breeding, even inadvertent domestication selection can lead to maladaptation and the evidence is particularly strong for salmonids (Araki et al. 2007; Hutchings and Fraser 2008; Christie et al. 2012). For example, despite decades of deliberate stocking, White and colleagues (2018) found limited introgression of hatchery-raised brook trout (Salvelinus fontinalis) into wild populations. Much less is known about how advertent selection influences invasion success. However, compared to rearing for other aquaculture segments, maladaptation due to advertent selection may be particularly pronounced in ornamental species. As the number of domesticated species continues to rise, rapidly for some groups (Duarte et al. 2007), this may be an increasingly important avenue of research for ornamental species and others.

For swordtails, advertent selection has led to broad intraspecific diversity and a large number of distinctive varieties (Tuckett et al. 2021b), many of which have experienced decades of selection, although the provenance of these varieties is generally unknown. We expected colorful varieties to be strongly affected by biotic resistance due to bass and mosquitofish predation. These advertently produced varieties decreased relative to the wild type in ponds with increased community complexity, suggesting fitness costs associated with enhanced color. Eleven varieties of swordtail have been captured in streams and ditches of the Tampa Bay area (Tuckett et al. 2017). Colorful fish are often captured near sources of production, the incidence and diversity of which declines with increasing distance from the effluent. These results may also be generalizable to other ornamental species. Several species within this family, guppy (Poecilia reticulata), southern platyfish (Xiphophorus maculatus), and variable platyfish (Xiphophorus variatus), also have a large number of recognized varieties and trade is dominated by the highly-domesticated, colorful varieties (Tuckett et al. 2021b). Several varieties of both southern and variable platyfish have also been captured in this region, with the southern platyfish exhibiting a similar attenuated incidence and diversity pattern as the swordtail (Tuckett et al. 2017). For intraspecific variation at its most extreme and recognizable, transgenic GloFish® exhibit profound fitness costs compared to the wild type due to their fluorescent colors, increasing vulnerability to predators (Hill et al. 2011; Howard et al. 2015). Enhanced color has obvious fitness consequences; however, traits such as age at maturity and growth rate are often under advertent selection in hatchery settings (Hutchings and Fraser 2008), but the consequences of phenotypic manipulation of these traits have not been fully resolved.

Despite the potential for fitness tradeoffs, domestication is not an evolutionary dead end and many highly-domesticated species have become successful invasives (Gering et al. 2019b). While our results suggest the pattern of artificial selection affects survival, domestication may not universally lead to reduced fitness, and hybridization is a possible explanation for invasion success. For example, plants selected for rapid growth and early flowering, and upon hybridization with a wild counterpart, exhibit increased fitness compared to wild populations (Mercer et al. 2007). The swordtail has successfully feralized in west-central Florida (Tuckett et al. 2016; Lawson et al. 2017). The origin of these feral swordtail populations is unclear, although some authors consider ornamental swordtails to be exoferal hybrids (D’Amore et al. 2019), perhaps originating from research in the 1930s (Gordon 1937), and potentially contributing to their invasiveness. Established swordtail populations also exhibit evidence of feralization and associated phenotypic change (Tuckett et al. 2021b). In Florida, feral swordtails exhibit greater ability to tolerate cold temperature and an established population in Hawaii exhibits increased boldness compared to captive individuals (Tuckett et al. 2016; D’Amore et al. 2019). Ultimately, feral populations in Florida are phenotypically similar to the inadvertently produced variety used in this study.

Biotic resistance

Biotic resistance is often identified as a positive relationship between diversity and invasibility (Elton 1958; Levine and D’Antonio 1999), a theory that has declining empirical support (Jeschke et al. 2012), which could be due to study design and emphasis. Alternate approaches emphasize particular species and interactions (Henriksson et al. 2015). We expected increasing community complexity to lead to greater biotic resistance, which was largely borne out. We also expected treatments with competitors and predators to alter the relative success of swordtail varieties. Despite evidence that propagule pressure can surmount the resistive effects of native species (Von Holle and Simberloff 2005), field observations (Tuckett et al. 2021b) and the present study suggest biotic resistance attenuated the relative propagule pressure of swordtail varieties. The wild-type variety consistently increased relative to the other four varieties in all three treatments with native competitors and predators, despite the lower weekly propagule size. While identification of the phenotypes associated with reduced survival were not specifically delineated, the striking colors were undoubtedly important, which is consistent with a previous study (Hill et al. 2011).

We focused on ubiquitous and dominant native species because they occur in layered communities as the ditches and streams receiving aquaculture effluents become larger systems capable of supporting larger, predatory fish species (Hill and Tuckett 2018). We note a very strong effect of mosquitofish on swordtail abundance and previous studies have identified two mechanisms, consumption of neonates and aggression directed at adults (Hill et al. 2011; Thompson et al. 2012; Tuckett et al. 2021a). Increasing fish diversity by adding reproducing sunfish had little additive effect on numbers, but resulted in larger average body size of prey. While opportunistic feeders, bluegill gape, especially for the abundant juvenile bluegill, might limit the size of consumed swordtails to neonates and juveniles. In comparison, bass addition strongly affected swordtails in our additive approach. The number of swordtails with mosquitofish and both mosquitofish and sunfish were indistinguishable, but swordtail number declined to just 2.5 with bass. The strength of bass predation was likely compounded by prey selection as bass strongly preferred swordtail prey to the native mosquitofish. Largemouth bass are thought to discriminate among colors (Kawamura and Kishimoto 2002), particularly for the colors red and green. For species and varieties destined for the ornamental pet trade, species are often selectively bred for enhanced coloration and red colors are common. Our results support alternate approaches to testing biotic resistance, including identification of species (mosquitofish and bass) as strong resistors. It is not simply the number of species which limits invasion, but increasing diversity will likely increase the probability of including strongly-interacting species (Wardle 2001).

For an industry that rears hundreds of species and varieties (Hill and Yanong 2016), it is important to consider how intraspecific variation modifies invasion risk. While risk screens and assessments have been performed below the species level (e.g., Hill et al. 2014), in practice, it may be difficult to fully account for intraspecific diversity. Issues arise due to the lack of information below the species level, and often for the species themselves (Lyons et al. 2019). A pattern emerges where data on the biology, ecology, and invasion history of a species is borrowed from a congener and little is known until after the species is already established (Lyons et al. 2019). Thus, it may be difficult to fully account for these intraspecific differences.

Domesticated plants and animals were some of the world’s first “invasive species”, with human activity responsible for their creation, introduction, and spread. The process of domestication continues today at an ever accelerating rate and scale, particularly for aquatic organisms (Duarte et al. 2007; Teletchea 2016), such that domestication is one of the most important forces affecting the evolutionary trajectory of species inside and outside captivity (Turcotte et al. 2017). The importance of domestication selection suggests more emphasis should be placed on evaluating the conditions under which domestication takes place and the potential for variation in outcomes.

Acknowledgements

We thank the Florida Tropical Fish Farms Association and fish farmers who supplied the fish used in this study. We also thank current and former staff at the Florida Department of Agriculture and Consumer Services Division of Aquaculture, including Joe Clayton, Kal Knickerbocker, Serina Rocco, and Portia Sapp. Taylor Lipscomb provided the images.

Funding: Funding was provided by the University of Florida/IFAS School of Forest Resources and Conservation, Tropical Aquaculture Laboratory and a grant from the Florida Department of Agriculture and Consumer Services’ Division of Aquaculture.

Competing interests: The authors have no financial or non-financial interests in the material in this article.

Ethics approval: The authors followed all relevant institutional guidance for the care and use of vertebrate animals (UF IACUC #201408492).

Availability of data and materials: The datasets and analyses for this study are available from the corresponding author upon request.

Author contributions: QMT, JLR, LLL, and JEH conceived and designed the experiments. QMT, JLR, and LLL performed the experiments. QMT and JC analyzed the data. QMT and LLL wrote the manuscript; all authors provided edited the manuscript.

Allen J, Greenwood J (1988) Frequency-Dependent Selection By Predators. Philos Trans R Soc London Ser B-Biological Sci 319:485–503
Araki H, Berejikian BA, Ford MJ, Blouin MS (2008) Synthesis: Fitness of hatchery-reared salmonids in the wild. Evol Appl 1:342–355. https://doi.org/10.1111/j.1752-4571.2008.00026.x
Araki H, Cooper B, Blouin MS (2007) Genetic effects of captive breeding cause a rapid, cumulative fitness decline in the wild. Sci (80-) 318:100–103. https://doi.org/10.1126/science.1145621
Chesson J (1983) The Estimation and Analysis of Preference and Its Relatioship to Foraging Models. Ecology 64:1297–1304. https://doi.org/10.2307/1937838
Christie MR, Marine ML, French RA, Blouin MS (2012) Genetic adaptation to captivity can occur in a single generation. Proc Natl Acad Sci 109:238–242. https://doi.org/10.1073/pnas.1111073109
Colautti RI, Grigorovich IA, MacIsaac HJ (2006a) Propagule pressure: a null model for biological invasions Robert. Biol Invasions 8:1023–1037. https://doi.org/10.1007/s10530-005-3735-y
Colautti RI, Grigorovich IA, MacIsaac HJ (2006b) Propagule pressure: a null model for biological invasions. Biol Invasions 8:1023–1037. https://doi.org/10.1007/s10530-005-3735-y
D’Amore DM, Popescu VD, Morris MR (2019) The influence of the invasive process on behaviours in an intentionally introduced hybrid, Xiphophorus helleri–maculatus. Anim Behav 156:79–85. https://doi.org/10.1016/j.anbehav.2019.08.009
Duarte CM, Marbá N, Holmer M (2007) Rapid domestication of marine species. Sci (80-) 316:382–383. https://doi.org/10.1126/science.1138042
Duggan IC, Rixon CAM, MacIsaac HJ (2006) Popularity and propagule pressure: Determinants of introduction and establishment of aquarium fish. Biol Invasions 8:377–382. https://doi.org/10.1007/s10530-004-2310-2
Elton CS (1958) The ecology of invasions by animals and plants. Methuen and Co., London, UK
Finlayson BJ, Schnick R, Skaar D et al (2010) Planning and Standard Operating Procedures for the Use of Rotenone in Fish Management; Rotenone SOP Manual. American Fisheries Society, Bethesda, Maryland
Gering E, Incorvaia D, Henriksen R et al (2019a) Maladaptation in feral and domesticated animals. Evol Appl 0–2. https://doi.org/10.1111/eva.12784
Gering E, Incorvaia D, Henriksen R et al (2019b) Getting back to nature: Feralization in animals and plants. Trends Ecol Evol 34:1137–1151. https://doi.org/10.1016/j.tree.2019.07.018
Gjedrem T, Robinson N, Rye M (2012) The importance of selective breeding in aquaculture to meet future demands for animal protein: A review. Aquaculture 350–353:117–129. https://doi.org/10.1016/j.aquaculture.2012.04.008
Gordon M (1937) Heritable color variations in the Mexican swordtail-fish: Aquarium species as the drosophila of fish genetics. J Hered 28:223–230. https://doi.org/10.1093/oxfordjournals.jhered.a104368
Henriksson A, Yu J, Wardle DA, Englund G (2015) Biotic resistance in freshwater fish communities: Species richness, saturation or species identity? Oikos 124:1058–1064. https://doi.org/10.1111/oik.01700
Hill JE, Kapuscinski AR, Pavlowich T (2011) Fluorescent transgenic Zebra danio more vulnerable to predators than wild-type fish. Trans Am Fish Soc 140:1001–1005. https://doi.org/10.1080/00028487.2011.603980
Hill JE, Lawson LL, Hardin S (2014) Assessment of the Risks of Transgenic Fluorescent Ornamental Fishes to the United States Using the Fish Invasiveness Screening Kit (FISK). Trans Am Fish Soc 143:817–829. https://doi.org/10.1080/00028487.2014.880741
Hill JE, Tuckett QM (2018) Abiotic and biotic contributions to invasion resistance for ornamental fish in west-central Florida. USA Hydrobiologia 817:363–377. https://doi.org/10.1007/s10750-017-3496-5
Hill JE, Tuckett QM, Hardin S et al (2017) Risk screen of freshwater tropical ornamental fishes for the conterminous United States. Trans Am Fish Soc 146:927–938. https://doi.org/10.1080/00028487.2017.1312523
Hill JE, Yanong RPE (2016) Freshwater ornamental fish commonly cultured in Florida. University of Florida IFAS Extension. Circular 54
Howard RD, Rohrer K, Liu Y, Muir WM (2015) Mate competition and evolutionary outcomes in genetically modified zebrafish (Danio rerio). Evol (N Y) 69:1143–1157. https://doi.org/10.1111/evo.12662
Hutchings JA, Fraser DJ (2008) The nature of fisheries-and farming-induced evolution. Mol Ecol 17:294–313. https://doi.org/10.1111/j.1365-294X.2007.03485.x
Jeschke J, Gómez Aparicio L, Haider S et al (2012) Support for major hypotheses in invasion biology is uneven and declining. NeoBiota 14:1–20. https://doi.org/10.3897/neobiota.14.3435
Kawamura G, Kishimoto T (2002) Color vision, accommodation and visual acuity in the largemouth bass. 1041–1046
Lawson KM, Tuckett QM, Ritch JL et al (2017) Distribution and status of five non-native fish species in the Tampa Bay drainage (USA), a hot spot for fish introductions. BioInvasions Rec 6:393–406
Lazur AM, Cichra CE, Watson C (2013) The Use of Lime in Fish Ponds
Levine JM, D’Antonio CM (1999) Elton revisited: a review of evidence linking diversity and invasibiltiy. Oikos 87:15–26. https://doi.org/Doi 10.2307/3546992
Lockwood JL, Cassey P, Blackburn T (2005) The role of propagule pressure in explaining species invasions. Trends Ecol Evol 20:223–228. https://doi.org/10.1016/j.tree.2005.02.004
Lorenzen K, Beveridge MCM, Mangel M (2012) Cultured fish: integrative biology and management of domestication and interactions with wild fish. 639–660. https://doi.org/10.1111/j.1469-185X.2011.00215.x
Lyons TJ, Tuckett QM, Hill JE (2019) Data quality and quantity for invasive species: A case study of the lionfishes. Fish Fish faf 12374. https://doi.org/10.1111/faf.12374
Magalhães ALB, Jacobi CM (2013) Asian aquarium fishes in a Neotropical biodiversity hotspot: Impeding establishment, spread and impacts. Biol Invasions 15:2157–2163. https://doi.org/10.1007/s10530-013-0443-x
McCullagh P, Nelder JA (1989) Generalized Linear Models, 2nd edn. Chapman and Hall/CRC Press, London, UK
Mercer KL, Andow DA, Wyse DL, Shaw RG (2007) Stress and domestication traits increase the relative fitness of crop-wild hybrids in sunflower. Ecol Lett 10:383–393. https://doi.org/10.1111/j.1461-0248.2007.01029.x
Moyle PB, Nichols RD (1973) Ecology of some native and introduced fishes of the Sierra Nevada foothills of Central California.Copeia478–490
Price EO (2002) Animal domestication and behavior. CABI Publishing, Wallingford
Pyke GH (2008) Plague minnow or mosquito fish? A review of the biology and impacts of introduced Gambusia species. Annu Rev Ecol Evol Syst 39:171–191. https://doi.org/10.1146/annurev.ecolsys.39.110707.173451
Rogers TL, Byrnes JE, Stachowicz JJ (2016) Native predators limit invasion of benthic invertebrate communities in Bodega Harbor, California, USA. Mar Ecol Prog Ser 545:161–173. https://doi.org/10.3354/meps11611
Savino JF, Stein RA (1989) Behavior of fish predators and their prey: habitat choice between open water and dense vegetation. Environ Biol Fishes 24:287–293. https://doi.org/10.1007/BF00001402
Schofield PJ, Loftus WF (2015) Non-native fishes in Florida freshwaters: a literature review and synthesis. Rev Fish Biol Fish 25:117–145. https://doi.org/10.1007/s11160-014-9373-7
Simberloff D (2009) The role of propagule pressure in biological invasions. Annu Rev Ecol Evol Syst 40:81–102. https://doi.org/10.1146/annurev.ecolsys.110308.120304
Strecker A, Campbell P, Olden J (2011) The aquarium trade as an invasion pathway in the Pacific Northwest. Fisheries 36:74–85
Teletchea F (2016) Is fish domestication going too fast? Nat Resour 7:399–404. https://doi.org/10.4236/nr.2016.76034
Thompson KA, Hill JE, Nico LG (2012) Eastern mosquitofish resists invasion by nonindigenous poeciliids through agonistic behaviors. Biol Invasions 14:1515–1529. https://doi.org/10.1007/s10530-012-0176-2
Tuckett QM, Deacon AE, Fraser DF et al (2021a) Unstable intraguild predation causes establishment failure of a globally invasive species. Ecology e03411
Tuckett QM, Ressel KN, Ritch JL et al (2021b) Domestication and feralization influence the distribution and phenotypes of escaped ornamental fish. Biol Invasions 23:1033–1047. https://doi.org/10.1007/s10530-020-02415-1
Tuckett QM, Ritch JL, Lawson KM et al (2016) Variation in cold tolerance in escaped and farmed non-native green swordtails (Xiphophorus hellerii) revealed by laboratory trials and field introductions. Biol Invasions 18:45–56. https://doi.org/10.1007/s10530-015-0988-y
Tuckett QM, Ritch JL, Lawson KM, Hill JE (2017) Landscape-scale survey of non-native fishes near ornamental aquaculture facilities in Florida, USA. Biol Invasions 19:223–237. https://doi.org/10.1007/s10530-016-1275-2
Turcotte MM, Araki H, Karp DS et al (2017) The eco-evolutionary impacts of domestication and agricultural practices on wild species. Philos Trans R Soc B Biol Sci 372:20160033. https://doi.org/10.1098/rstb.2016.0033
Von Holle B, Simberloff D (2005) Ecological resistance to biological invasion overwhelmed by propagule pressure. Ecology 86:3212–3218.
https://doi.org/10.1890/1
Wardle DA (2001) Experimental demonstration that plant diversity reduces invasibility – evidence of a biological mechanism or a consequence of sampling effect? Oikos 95:161–170
White S, Miller W, Dowell S et al (2018) Limited hatchery introgressions into wild brook trout (Salvelinus fontinalis) populations despite reocurring stocking. Evol Appl 11:1567–1581. https://doi.org/10.1111/eva.12646
Yorisue T, Ellrich JA, Momota K (2019) Mechanisms underlying predator-driven biotic resistance against introduced barnacles on the Pacific coast of Hokkaido, Japan. Biol Invasions 21:2345–2356. https://doi.org/10.1007/s10530-019-01980-4

Tuckettetal.Pondsupplement.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Advertent domestication attenuates the influence of propagule pressure

Status:

Version 1

Abstract

Figures

Introduction

Methods

Predation Experiments

Data Analysis

Results

Pond Study

Prey Selection

Discussion

Propagule pressure

Domestication

Biotic resistance

Declarations

References

Supplementary Files

Status:

Version 1