| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Life Science Department, Bar-Ilan University, Ramat Gan, Israel, and Interuniversity Institute for Marine Science, P.O. Box 469, Eilat, Israel
2 Department of Pathology, Stanford Medical School, Stanford University, Stanford, California 94305-5323
* To whom correspondence should be addressed. E-mail: furman{at}mail.biu.ac.il
 |
Abstract |
|---|
 TOP Abstract IntroductionMaterials and MethodsResultsDiscussionLiterature Cited |
|---|
|
Introduction |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionLiterature Cited |
|---|
Studies of fused allogeneic colonies of the protochordate Botryllus schlosseri have shown that the eggs of one partner may be retained and brooded by the other partner over several reproductive cycles (Sabbadin and Zaniolo, 1979). In laboratory studies, fusion between allogeneic colonies of B. schlosseri leads to costs rather than benefits in terms of several fitness parameters (Rinkevich and Weissman, 1987, 1992a, b); indeed an inevitable result of such fusion is the death and resorption of all zooids (colonial units) of one colony, and the survival of the zooids of the other colony for up to many weeks after fusion (Rinkevich and Weissman, 1992a, b). In nature, however, resorption is not the inevitable conclusion of fusion prior to the onset of reproductive competence (Chadwick-Furman and Weissman, 1995a). Further, the resorbed partner also may parasitize the germ cell line of the resorbing partner in chimeras under both laboratory and field conditions (Pancer et al., 1995; Stoner and Weissman, 1996; Stoner et al., 1999). Thus, the genetic composition of chimeric colonies in nature may be more complex than previously observed in the laboratory.
Previously we reported on seasonal variation in life history traits of B. schlosseri colonies in a field population in Monterey Bay, California (Chadwick-Furman and Wiessman, 1995b). Here we determine natural frequencies of allogeneic contact in the same field population, and assess the resulting impacts on life-history traits in this colonial ascidian. We also describe the morphology and stability of chimeric colonies under natural field conditions.
|
Materials and Methods |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionLiterature Cited |
|---|
|
To test the effects of allogeneic contact on life-history traits in B. schlosseri, we set up three treatments using each of four cohorts of newly settled offspring from field-collected colonies. The four cohorts settled on 19 May 1990, 3 July 1990, 15 October 1990, and 25 January 1991 (after Chadwick-Furman and Weissman, 1995b). In each cohort, newly settled, one-system colonies of 1–7 zooids each (= one circular system of zooids, Fig. 1a), were either (1) isolated on plates, (2) placed in incompatible pairs that rejected each other, or (3) placed in compatible pairs that fused. We distinguished between fusible and incompatible pairs by placing the small, one-system colonies into contact and observing the outcome. We used only colony pairs that established contact during the one-system stage of development. In each cohort, offspring from 10 field-collected colonies were assigned randomly to each of the three treatments (n = 8–36 newly settled colonies per treatment). A total of 274 colonies from all cohorts were monitored.
Each experimental colony or pair of colonies was placed on a glass plate measuring 5.0 x 7.5 cm and allowed to attach firmly for 1 week in the laboratory (after Rinkevich and Weissman, 1992a; Chadwick-Furman and Weissman, 1995b). Colonies that did not firmly attach to the plates, or that appeared damaged, were removed from the study at this point. All well-attached colonies were then transferred to the marina field site, in an area where abundant colonies of B. schlosseri grow naturally on fouling surfaces. About every 7 days, depending on the time of year, all the zooids in each colony passed through an asexual growth cycle (hereafter termed "cycle"). During each cycle, the zooids produced buds, then shrank and were replaced by their buds; thus a new asexual generation of zooids was formed in each colony. To examine cycle-related life-history traits, every 4–7 days we collected all the experimental colonies, observed them under a dissecting microscope in the laboratory, and returned them to the field within a few hours (after Chadwick-Furman and Weissman, 1995a, b).
We examined the following life-history parameters for each colony: (1) growth rate of somatic tissues, as measured by the number of clonal units (zooids, Fig. 1) produced per cycle; (2) age and size at sexual maturity, defined as the beginning of egg production; and (3) sexual reproductive output, as measured by the number of eggs produced by each zooid during each cycle, the number of cycles in which eggs were produced (# clutches), and the total number of eggs produced by each colony throughout its lifespan (fecundity) (after Sabbadin and Zaniolo, 1979; Sabbadin and Astorri, 1988; Chadwick-Furman and Weissman, 1995b). We assigned zooids in chimeric colonies to genotype on the basis of morphological and developmental characters, such as their relative positions in the chimera, the number of buds produced, and in some cases, color patterns (after Chadwick-Furman and Weissman, 1995a; Yund et al., 1997). Since colonies were observed every 4–7 days, we counted directly the number of buds produced by each zooid at each cycle, and thus accurately assigned each new budded zooid to original colony genotypes in chimeras.
All statistical analyses were performed using STATA, version 7.0 (Statacorp, 2001). Effects of allogeneic contact treatment on life-history traits were examined only within each cohort, since between-cohort comparison of life-history traits were made previously (Chadwick-Furman and Weissman, 1995b). For life-history traits that were examined on a per-cycle basis (i.e., number of zooids produced per cycle and number of eggs per zooid per cycle, see above), we measured the value for each cycle within a colony, but we present only the mean of these values for each colony. Thus, at one-way model was used in analyzing these traits. Log-transformed values of all life-history traits had approximately equal variances between treatment groups within each cohort, so ANOVA tests were applied to the data.
|
Results |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionLiterature Cited |
|---|
Morphology and growth
Colony morphology was similar in all cohorts and experimental treatments. All colonies were flat and disc-shaped when small, with closely spaced groups of zooids (Fig. 1a). As they grew, some of the colonies developed irregular outlines, but the zooid systems remained compact and close together (Fig. 1b, c). In fused chimeric colonies, the area of fusion became barely visible over time, and some zooid systems straddled the area of initial fusion (Fig. 1b). The zooids of all genotypes in fused colonies appeared to grow constantly and to coexist in chimeras during their entire lifespan (Fig. 1b). We did not observe any shrinkage or somatic resorption of one genotype by another in chimeras. Until the time of chimeric colony senescence and death, robust blastozooids from all partners appeared to coexist within a single fused colony (Fig. 1b).
Colonies that contacted noncompatible partners underwent rejection reactions that persisted along an extensive border of contacting tissues (Fig. 1c). As colonies grew, the contact area expanded along this border, and the number of points of rejection increased. Up to 15 points of rejection were observed during each sampling period throughout the lifespan of rejecting colonies. All rejecting colonies maintained a long, continuous border throughout their lifespans, until one of them senesced and died (Fig. 1c). Pairs of rejecting colonies were compact, grew actively, and neither retreated nor grew away from each other.
Colonies grew until they reached the edges of the glass culture plate, then grew around the plate edges, and continued to spread over the back sides of the plate. None of the colonies filled all of the space available on both sides of the plate (Fig. 1).
Juvenile colonies grew exponentially, regardless of treatment (Fig. 2). During January and October, exponential growth began after a lag time of 3–5 cycles (= 32 to 64 days, Fig. 2). In all cohorts, colonies in the isolated treatment reached the largest maximum size (Fig. 2). This pattern persisted even in the October cohort, in which some isolated colonies experienced partial predation during cycle 9 that reduced their size to almost zero, after which they recovered and became the largest colonies in the cohort (Fig. 2). Growth rate slowed upon commencement of sexual reproduction in all cohorts (Fig. 2).
|
|
|
|
Variation in the size of colonies at first reproduction followed the same pattern as did colony growth rate (Table 2), but the differences between groups were magnified (compare Figs. 3a and c). In both the January and May cohorts, isolated colonies, which grew relatively rapidly as juveniles (Fig. 3a), were significantly larger at maturity than were fused and rejected colonies (Fig. 3c, Table 2). Where there were significant differences, isolated colonies were, on average, 1.5–2.5 times larger at sexual maturity than rejected colonies, and 2–3 times larger than colonies that fused to form chimeras (Fig. 3c).
The number of eggs produced per zooid per cycle (= reproductive effort) varied widely between colonies within each treatment, and did not vary between treatments, except in the May cohort (Table 1, Fig. 3d). For colonies born during May, isolated individuals produced significantly more eggs per zooid per cycle than did colonies in either of the allogeneic contact treatments (Table 2).
The total number of egg clutches produced by each colony was affected by treatment in only the two cohorts that overwintered, those born during October and January (Table 1, Fig. 3d). In both cases, colonies that were isolated from contact produced significantly more egg clutches than did those in either of the two allogeneic contact treatments (on average, 2–3 times more clutches, Fig. 3e, Table 2).
The lifetime fecundity of colonies varied significantly with treatment in all cohorts (Table 1). The combined effects of relatively rapid somatic growth, large size at maturity, and a large number of egg clutches in the isolated colony treatment (Fig. 3a–e) resulted in much higher lifetime fecundity in isolated than in either the fused or rejected treatments (Fig. 3f, Table 2). The mean fecundity of isolated colonies ranged from 1.8 to 2.5 times that of fused or rejected colonies in summer cohorts (May and July). In the winter cohorts (January and October), the mean fecundity of isolated colonies was more than 5–10 times that of fused or rejected colonies. Fused colonies that were born in October did not produce eggs at all (Fig. 3f).
Colony longevity and survivorship
Colonies in all treatments and cohorts had short, subannual lifespans (Fig. 4). Within each cohort, colonies in all treatments reached sexual maturity at about the same age (Fig. 3b), reproduced sexually for a few cycles, and then all died within a few cycles of each other (Fig. 4). The percentage of colonies that survived to reproduce was high and did not vary significantly among treatments in the January and May cohorts (chi-square tests, 
20.05(2) = 5.99, G = 0.68 and 3.78 respectively, Fig. 5). In the July cohort, survivorship also was high, but did vary with treatment; rejected colonies had the lowest survivorship to maturity (G = 13.16). Colonies born in October had low survivorship that did not vary significantly with treatment, even though all colonies in the fused treatment died as juveniles (G = 5.04, Fig. 5).
|
|
When senescence began in the zooids of one genotype, it spread to the zooids of fused, but not rejected, partners (Fig. 1b, c). After one of the partners in a rejecting pair died, the other colony continued to live for a few cycles (Fig. 1c).
|
Discussion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionLiterature Cited |
|---|
We also show that colonies of Botryllus schlosseri in the wild frequently contact those of conspecifics and of other species of sessile invertebrates, so associated fitness costs may be a ubiquitous and important phenomenon in nature. Contact rates with the colonial ascidian Botrylloides violaceous were especially high at our site (see Results). Yet, because our survey of contact frequencies was based on a one-time observation, which inherently underestimates contacts throughout the life of a colony, lifetime contact rates between colonies of Botryllus schlosseri and other sessile organisms at Monterey are even higher than those presented here (see Results). Our limited manipulation of colonies in one of the cohorts of Botryllus schlosseri (born on 15 October 1990) indicates that xenogeneic interaction with Botrylloides violaceous results in a level of fecundity intermediate between those of isolated and allocontacted colonies [total number of eggs produced = 1383 + 769 (
+ SD), n = 9 xenocontacted colonies of Botryllus schlosseri that survived to maturity, N. E. Chadwick-Furman, pers. obs.; compare with October cohort in Fig. 3f]. Thus, xenogeneic contact appears to affect colony fecundity, but not as severely as allogeneic contact.
The reduced fitness of colonies following fusion or rejection may result from energetic or physiological costs associated with recognizing and reacting to non-self tissue. The process of interaction along the borders of rejecting colonies involves extensive tissue damage and resource demand on both colonies (Scofield and Nagashima, 1983; reviewed in Rinkevich, 1992). In addition, competition between somatic and germ cell lines within fused chimeras also may draw heavily on the physiological resources of the partners involved (Buss, 1982). Colonies that are isolated from allogeneic contact do not face these costs.
The lack of resorption observed here in field-raised chimeras of Botryllus schlosseri is in striking contrast to previous results from laboratory studies (Rinkevich and Weissman, 1987, 1992a, b; Pancer et al., 1995). The reduced chimeric stability of laboratory colonies has been demonstrated by growing genetically identical replicates of chimeras under field versus laboratory conditions (Chadwick-Furman and Weissman, 1995a). The present results show that, under field conditions in Monterey, the partners of a chimera appear to grow in a stable manner and do not undergo somatic resorption. Previous field studies indicate a high level of environmentally dependent plasticity in fitness-related life-history traits of B. schlosseri (Chadwick-Furman and Weissman, 1995a; Yund et al., 1997).
The reduced reproductive success of interacting versus isolated colonies (Fig. 3 and Table 1) reveals effects of allogeneic contacts on sexual reproduction as well as on somatic tissue production. Recruitment of larvae at Monterey in the springtime may be derived from a small number of parent colonies that overwinter (Carwile, 1989; Chadwick-Furman and Weissman, 1995b). Thus, the costs of interactions in this experiment would have resulted in reduced representation of the offspring of winter allo-contacting colonies in the summer bloom.
Allogeneic interactions do not alter the survivorship of colonies in most cohorts (Fig. 5). The longer lifespans (Figs. 3b and 4) and lower survivorship (Fig. 5) of colonies during winter, as compared to summer, appear to be due to a slowing of colony growth and development during low temperatures in the winter in Monterey Bay (Boyd et al., 1986; Chadwick-Furman and Weissman, 1995b). As found in past studies, whole-colony senescence causes the death of most colonies (Chadwick-Furman and Weissman, 1995a, b) and is genetically controlled (Rinkevich et al., 1992).
At the time of our experiments, we did not have markers to identify the genotypes of blood cells, bud cells, or gametes in the fused colonies, and so we did not test for somatic or germ cell parasitism as a result of colony fusion. However, germ cell parasitism has been reported in this species (Rinkevich and Weissman, 1987; Sabbadin and Astorri, 1988; Pancer et al., 1995), and recent work has verified that it occurs in both male and female gametes capable of fertilization (Stoner and Weissman, 1996; Stoner et al., 1999). The process of germ cell parasitism, in which one partner in a chimera uses the somatic resources of the other to produce its own germ cells, may alter the relative fitness of fused genotypes in chimeras (Buss, 1982; Stoner and Weissman, 1996; Stoner et al., 1999; Weissman, 2000). However, because the fitness of fused pairs of genotypes was less than half that of isolated colonies in all cohorts (Fig. 3f), the reproductive output of all the genotypes combined in chimeric colonies was less than that of genotypes in isolated colonies. Thus, germ cell parasitism may alter the relative amount of fitness lost due to fusion in chimeric colonies, but cannot prevent an overall reduction in fitness due to fusion. Even if germ cell parasitism were extensive in the chimeras tested here, chimera formation causes reduced fecundity, regardless of which genotype dominates (Fig. 3f). In 30% of the field chimeras examined by Stoner and Weissman (1996) at the same Monterey marina site, little or no germ or somatic cell parasitism was found. Thus, the values presented here for genotype-specific measures of fitness (Fig. 3) may represent realistic estimates for at least some chimeras that retain a stable genetic composition in the wild.
A drawback of the present study is that we could not set up, as controls, undissected pairs of isogeneic colonies, to determine whether isogeneic contact affects fitness. Thus, an evaluation of the actual costs of allogeneic contact per se is problematic. However, set-up of this control group would have required dissecting apart and re-uniting systems from multi-system colonies, thus introducing further manipulation of all colonies in this experiment. As the colonies grew, they produced lobes of tissue that contacted along their edges and fused along the undulating margins of the colony in all treatments (Fig. 1b, c). Thus, if isogeneic contact affected fitness, it did so equally in all treatments here.
We show here that egg production in fused colonies is greatly reduced (Fig. 3f), possibly due to competition between the genetically different individuals that fused to make up that colony. Thus, one benefit of precise allorecognition in this species may be that it limits the unit of selection to chimeras composed of closely related kin (Grosberg and Quinn, 1986; Rinkevich and Weissman, 1987; Stoner and Weissman, 1996; Stoner et al., 1999). Because of the high polymorphism of the Fu/HC gene locus (that permits fusion rather than rejection to occur; Scofield et al., 1982), fused individuals in the wild most likely represent kin rather than a random assortment of genotypes (Grosberg and Quinn, 1986). In Botryllus schlosseri, the proportion of fusions occurring between siblings is higher than between nonsiblings (Scofield et al., 1982; Magor et al., 1999). Thus, the unit of genetic inheritance for chimeric colonies of fused siblings would be the outcome of germline competition between the mother colony and the diverse sperm that fertilized her. In addition to kin fusion, regulated by Fu/HC matching, kin cosettlement is encouraged by the limited dispersal of tadpole larvae from the maternal colony and nonrandom cosettlement according to shared Fu/HC genotype (Grosberg and Quinn, 1986). A common selected trait in these chimeras is allele-sharing at the Fu/HC locus (Weissman et al., 1990). Kin selection would act also on shared genes other than the selected Fu/HC types that are common to these siblings. Reproductive outcomes in these chimeras could be as simple as the direct gametic representation of the diverse blastozooid units in the chimera; or could be as complex as the outcomes of selective resorption or germ cell parasitism that generate skewing from that simple representation (Pancer et al., 1995; Stoner and Weissman, 1996; Stoner et al., 1999; Weissman, 2000).
No matter whether allogeneic colony contact results in fusion or rejection, if it leads to reduced fitness, as measured by growth and fecundity, with no increase in survivorship, why have these organisms developed and maintained an elaborate system of allorecognition? Perhaps, in this species, genetically based allorecognition is nonadaptive. It may be linked to other processes that are adaptive, and thus have evolved as a by-product of processes such as disease recognition (Buss and Green, 1985; Magor et al., 1999) or gametic compatibility (Scofield et al., 1982). However, the ability to recognize and reject nonrelated colonies, and to fuse only with closely related kin that share alleles at the Fu/HC locus, may be directly beneficial in that it reduces the costs of germ cell parasitism in colonies (Stoner et al., 1999).
The phenomena of cosettlement, fusion, and development of reproductive competence in chimeras are not limited to protochordates, and may be important selective factors in other sessile organisms as diverse as fungi, sponges, and cnidarians (reviewed in Buss, 1982; Rinkevich and Weissman, 1987; Pancer et al., 1995). Our findings that allogeneic contact, and especially chimera formation, reduce individual fitness under natural field conditions may have broad implications for the evolution of allorecognition systems.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
Literature Cited |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionLiterature Cited |
|---|
This article has been cited by other articles:
|
|
 |
|
  D. L. Ferrell Fitness Consequences of Allorecognition-Mediated Agonistic Interactions in the Colonial Hydroid Hydractinia [GM] Biol. Bull., June 1, 2004; 206(3): 173 - 187. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
