Downloaded from rsbl.royalsocietypublishing.org on October 30, 2014
Developmental constraints in cave beetles Alexandra Cieslak, Javier Fresneda and Ignacio Ribera Biol. Lett. 2014 10, 20140712, published 29 October 2014
Supplementary data
"Data Supplement" http://rsbl.royalsocietypublishing.org/content/suppl/2014/10/27/rsbl.2014.0712.DC1.ht ml
References
This article cites 10 articles, 2 of which can be accessed free
Subject collections
Articles on similar topics can be found in the following collections
http://rsbl.royalsocietypublishing.org/content/10/10/20140712.full.html#ref-list-1
evolution (791 articles) taxonomy and systematics (81 articles)
Email alerting service
Receive free email alerts when new articles cite this article - sign up in the box at the top right-hand corner of the article or click here
To subscribe to Biol. Lett. go to: http://rsbl.royalsocietypublishing.org/subscriptions
Downloaded from rsbl.royalsocietypublishing.org on October 30, 2014
Evolutionary biology
Developmental constraints in cave beetles rsbl.royalsocietypublishing.org
Alexandra Cieslak1,2, Javier Fresneda3 and Ignacio Ribera2 1
Museo Nacional de Ciencias Naturales (MNCN, CSIC), Madrid, Spain Institute of Evolutionary Biology (CSIC-Universitat Pompeu Fabra), Passeig Maritim de la Barceloneta 37-49, Barcelona 08003, Spain 3 Museu de Cie`ncies Naturals (Zoologia), Barcelona, Spain 2
Research Cite this article: Cieslak A, Fresneda J, Ribera I. 2014 Developmental constraints in cave beetles. Biol. Lett. 10: 20140712. http://dx.doi.org/10.1098/rsbl.2014.0712
Received: 3 September 2014 Accepted: 5 October 2014
Subject Areas: evolution, taxonomy and systematics Keywords: developmental constraints, larval development, life cycle, subterranean environment
Author for correspondence: Alexandra Cieslak e-mail:
[email protected]
Electronic supplementary material is available at http://dx.doi.org/10.1098/rsbl.2014.0712 or via http://rsbl.royalsocietypublishing.org.
In insects, whilst variations in life cycles are common, the basic patterns typical for particular groups remain generally conserved. One of the more extreme modifications is found in some subterranean beetles of the tribe Leptodirini, in which the number of larval instars is reduced from the ancestral three to two and ultimately one, which is not active and does not feed. We analysed all available data on the duration and size of the different developmental stages and compared them in a phylogenetic context. The total duration of development was found to be strongly conserved, irrespective of geographical location, habitat type, number of instars and feeding behaviour of the larvae, with a single alteration of the developmental pattern in a clade of cave species in southeast France. We also found a strong correlation of the size of the first instar larva with adult size, again regardless of geographical location, ecology and type of life cycle. Both results suggest the presence of deeply conserved constraints in the timing and energy requirements of larval development. Past focus on more apparent changes, such as the number of larval instars, may mask more deeply conserved ontogenetic patterns in developmental timing.
1. Introduction Insects are extremely diverse in their morphological and physiological traits [1], and many of these modifications affect the life cycle, allowing fine-tuned adaptation to most environments [2]. Despite this diversity, there are developmental constraints and ontogenetic patterns that characterize general types of life cycles common to all members of a lineage. In holometabolous insects, the basic pattern is characterized by embryonic development, followed by several larval stages alternating growth and moulting. After the last larval moult, the metamorphic transition is completed in the pupal stage, resulting in the formation of an adult with a morphology substantially different from the previous stages (see [3] for an overview). Variations on this basic pattern are rare, occurring for instance in cases of heterochrony, in which either reproductive maturity is reached before transforming into an adult or the insect becomes an adult that retains some larval characters [4,5]. In addition, many groups show extreme stability in larval development, with a fixed number of larval instars before pupation [6]. Detailed knowledge of the life cycle and its modifications is restricted to relatively few species, and it is far from clear how often and how easily life cycles diverge from the general pattern. A spectacular case of life cycle modification is the reduction in the number of larval instars observed in different groups of subterranean insects, leading in its most extreme forms to a non-feeding larva that does not moult until pupation [7,8]. In the subterranean species of Leptodirini beetles, these modifications of larval development have been documented for several species in different geographical areas and have been traditionally assumed to have evolved multiple times independently, even in the case of closely related species [8,9]. However, with the use of a comprehensive molecular phylogeny, it has recently been shown that the number of independent origins of life cycle modification was low, with a single one in the Pyrenees, the
& 2014 The Author(s) Published by the Royal Society. All rights reserved.
Downloaded from rsbl.royalsocietypublishing.org on October 30, 2014
We obtained detailed data on the life cycle of 15 species from the literature, including average duration and size of individual developmental phases obtained from breeding experiments usually conducted through multiple generations under controlled conditions (electronic supplementary material). To place these species in a phylogenetic context, we used the phylogeny in [10], which includes 14 of these 15 species (the missing species is presumably related to another that is included) plus a comprehensive sampling of the main lineages of Leptodirini (electronic supplementary material). To compare the duration of the different stages, we grouped species according to different criteria: (i) phylogenetic: a derived clade of five species with 1-larval instar from southeast France versus the rest (showing the plesiomorphic condition [10]); (ii) geographical: species from the Pyrenees versus those from southeast France; and (iii) developmental: the species with 1-instar larvae versus those with 2- or 3- instar larvae. We compared the duration (in days) of the egg, larval and pupal stages and the total developmental time among species using a two-tailed ANOVA with the uncorrected data and using phylogenetically independent contrasts as implemented in the PDAP package run in MESQUITE v. 2.7 (http://mesquiteproject.org). We also used discriminant analysis with membership of one of the three groups as the independent variable and the duration of egg, larvae and pupae as predictors to identify misclassified species (electronic supplementary material, table S3). To compare larval and adult size, we used phylogenetically independent contrasts in the PDAP package, also using the phylogeny in [10].
3. Results The phylogenetic criterion was the best for comparing the duration of the different developmental stages, with no misclassified species in the discriminant analysis and the highest significances in pairwise PDAP comparisons (electronic supplementary material). Most species showed the ancestral,
2
(a)
(b)
egg larvae pupae
egg larvae pupae
plesiomorphic
1 instar SE France
days
200 150 100 50 0
Figure 1. Average values of the duration of the different developmental stages in eight species from the Pyrenees and southeast France, representatives of the plesiomorphic condition (a), and in a derived clade of five species from southeast France with a 1-larval instar only (b). Bars, absolute range of the values (in days). 8
1st instar length (mm)
7 6 5 4 3
0.7
2
0.3
1
0 0.1
0.3
0.5
0 0
1
2
5 3 4 adult length (mm)
6
7
8
Figure 2. Relationship between the size of the adult and the size of the freshly hatched larva. Insert, phylogenetically independent contrasts for the same data. Black circles, 1-instar cycle; grey, 2-instars; white, 3-instars. The dotted line marks the isometric relationship. See the electronic supplementary material for details on the species and the available data. plesiomorphic condition, with a relative duration of the larval stages—irrespective of the number of instars—of ca 60% of the total development (figure 1a). This was significantly longer than the embryonic and pupal stages, which took up ca 20% each. In the second group, however, embryonic, larval and pupal development were each of similar duration, thus taking a 33% relative duration of total development (figure 1b). This group comprised a clade of five species from southeast France with a 1-larval instar life cycle. The total duration of development (N ¼ 8, average ¼ 305 days, s.d. ¼ 27) was not significantly different between these two groups (electronic supplementary material, table S3). The application of alternative criteria—either geographical (Pyrenees versus southeast France) or type of development (species with 1-larval versus 2- and 3-larval instars)—resulted in heterogeneous groups with respect to the duration of different developmental stages, with lower or no significant differences between them (electronic supplementary material). There was a strong correlation between the total length of the freshly hatched larva and the adult, irrespective of developmental type and geographical origin of species (regression using phylogenetically independent contrasts, N ¼ 12, r 2 ¼ 0.92, p , 0.000001; figure 2). The exclusion of an outlier species with large body size (Leptodirus hochenwartii) only slightly
Biol. Lett. 10: 20140712
2. Material and methods
250
rsbl.royalsocietypublishing.org
geographical area for which most data are available [10]. In this lineage, there was a single transition during the Late Oligocene–Early Miocene from an ancestral 3-instar cycle [11] in species living in forest litter or under stones to a 2-instar cycle in species with fully subterranean habits (either in deep or shallow subterranean environments), in which the active larva still feeds. From within this lineage with a 2-instar cycle, there was then a single transition to a 1-instar, non-feeding larva in the Lower Miocene, in a lineage found today only in the deep subterranean environment of caves [10]. In this work, we analyse in detail the available data on the duration and size of the different developmental stages of subterranean leptodirine beetles in the phylogenetic context provided by Cieslak et al. [10]. We find that the total duration of development is strongly conserved in these insects, despite variation in the number of instars and feeding behaviour of larvae; this developmental duration differs only in a clade of cave species from southeast France. We also find a strong correlation between first instar larval and adult body sizes, again regardless of number of larval moults. Both results suggest the presence of deeply conserved constraints in the timing and energy requirements of larval development, and at the same time higher variability in traits traditionally given greater biological relevance, such as the number of larval instars [9,10].
Downloaded from rsbl.royalsocietypublishing.org on October 30, 2014
4. Discussion
Data accessibility. All primary data are provided in the electronic supplementary material. Acknowledgements. We thank C. Bordeaux, A. Casale and A. Faille for multiple discussions on the evolution of the subterranean fauna; and A. Minelli, D. T. Bilton and three anonymous referees for comments on the manuscript.
Funding statement. This work was partially funded by the Spanish Government project nos. CGL2007-61943 and CGL2006-11403 to A.C.
References 1. 2. 3.
4.
5.
Grimaldi D, Engel AS. 2005 Evolution of the insects. Cambridge, UK: Cambridge University Press. Taylor F, Karban R. 1986 The evolution of insect life cycles. New York, NY: Springer. Scott SM, Dingle H. 1990 Developmental programmes and adaptive syndromes in insect life-cycles. In Insect life cycles: genetics, evolution and co-ordination (ed. F Gilbert), pp. 69–85. New York, NY: Springer. South A, Strangler-Hall K, Jeng ML, Lewis SM. 2011 Correlated evolution of female neoteny and flightlessness with male spermatophore production in fireflies (Coleoptera: Lampyridae). Evolution 65, 1099–1113. (doi:10.1111/j.1558-5646.2010.01199.x) Cronin AL, Molet M, Doums C, Monnin T, Peeters C. 2013 Recurrent evolution of dependent colony foundation across eusocial insects. Annu. Rev. Entomol. 58, 37 –55. (doi:10.1146/annurev-ento120811-153643)
6.
Minelli A, Brena C, Deflorian G, Maruzzo D, Fusco G. 2006 From embryo to adult - beyond the conventional periodization of arthropod development. Dev. Genes. Evol. 216, 373–383. (doi:10.1007/s00427-006-0075-6) 7. Glac¸on S. 1955 Remarques sur la morphologie et la biologie de quelques larves de Bathysciinae cavernicoles. C.R. Acad. Sci. Paris 240, 679–681. 8. Deleurance S. 1963 Recherchers sur les Cole´opte`res troglobies de la sous-famille Bathysciinae. Annu. Sci. Nat. Zool. Paris Ser. 12 1, 1–172. 9. Delay B. 1978 Milieu souterrain et e´cophysiologie de la reproduction et du de´veloppement des Cole´opte`res Bathysciinae hypoge´s. Me´m. Biosp. 5, 1–349. 10. Cieslak A, Fresneda J, Ribera I. 2014 Life-history specialization was not an evolutionary dead-end in Pyrenean cave beetles. Proc. R. Soc. B 281, 20132978. (doi:10.1098/rspb.2013.2978)
11. Minelli A, Fusco G. 2013 Arthropod post-embryonic development. In Arthropod biology and evolution (eds A Minelli, G Boxshall, G Fusco), pp. 91– 122. Berlin, Germany: Springer. 12. Suzuki Y, Koyama T, Hiruma K, Riddiford LM, Truman JW. 2013 A molt timer is involved in the metamorphic molt in Manduca sexta larvae. Proc. Natl Acad. Sci. USA 110, 12 518 –12 525. (doi:10. 1073/pnas.1311405110) 13. Minelli A. 2003 The development of animal form: ontogeny, morphology and evolution. Cambridge, UK: Cambridge University Press. 14. Tennessen JM, Thummel CS. 2008 Developmental timing: let-7 function conserved through evolution. Curr. Biol. 18, R707–R708. (doi:10.1016/j.cub.2008.07.013) 15. Strambi C. 1970 Evolution ponderale postembryonnaire chez trois cole´opte`res catopides. Ann. Spe´le´ol. 25, 275 –298.
3
Biol. Lett. 10: 20140712
We found a strong constraint in the total duration of developmental stages in subterranean leptodirines, irrespective of the number of larval instars, type of habitat (forest litter or subterranean) and geographical region. Even though we obtained all raw data from published sources, these data had never been used in comparative studies (neither in a phylogenetic context nor otherwise) and none of the developmental patterns we uncover have been proposed before. According to our results, the ancestral, plesiomorphic condition of the relative duration of the developmental stages remained unchanged through life cycle evolution in the Pyrenees in both epigean and subterranean species, and also through at least one transition to a life cycle with 2-larval instars in southeast France. It was, however, modified at the origin of one clade with 1-larval instar in southeast France, but even in this clade the total duration of development, from egg to adult, did not change. The origin of this constraint, and whether it is linked to the colonization of the subterranean environment, remains speculative, but may be related to the existence of programmed timers for the completion of metamorphosis [12] or the specific cell division rate and determination [13,14]. Our results show an increased evolutionary variability of the number of instars with respect to other developmental constraints, in agreement with the view that growth and development are decoupled from traditionally recognized strict boundaries in insect ontogeny, such as moulting [6,11]. In all species of Leptodirini, the reported size of the freshly hatched larvae was similar to that of the adult, which is unusual both in other groups of the same family (Leiodidae) and
in Coleoptera as a whole, in which first instar larvae are generally much smaller than the adult. The large size of the first instar larva suggests a general trend in Leptodirini to accumulate resources in the egg, reducing the contribution of larvae to the global input necessary to complete development. It has been suggested that such a non-feeding larva may allow a species to survive in more extreme environments, as the more mobile adult can forage over a larger area to provide enough resources for the whole life cycle [10]. In Leptodirini, the non-feeding, inactive larva is protected inside a cocoon (‘logette’) [7,8], and thus less vulnerable to predators, which may be an additional advantage [10]. The conservation of the size of first instar larvae with respect to adult size, irrespective of whether the larva feeds or not, is surprising and counterintuitive, but could be explained by differences in dry weight. The only available data report a dry weight of 41.3% in the recently hatched larvae of Troglodromus bucheti, with a 1-instar cycle and a non-feeding larva, compared to 18% for Speophyes lucidulus, with 2-instars and a feeding larva [15]. Through the ontogeny, the relative dry weight of the non-feeding larvae decreases due to water absorption through the integument in the saturated atmosphere of the cave, and just before pupation, the ratio of dry to total weight was shown to be similar in both species (18% versus 22%, respectively) [15]. In summary, our results reveal a number of surprising developmental constraints in this lineage of beetles acting through multiple evolutionary transitions both in habitat type and number of larval instars. These constraints are best explained by the existence of intrinsic factors related to cell-cycle regulation determining developmental rates [13].
rsbl.royalsocietypublishing.org
decreased the correlation (N ¼ 11, r 2 ¼ 0.71, p , 0.001). There was a negative allometric relationship (slope of the uncorrected regression line a ¼ 0.73, s.e.m. ¼ 0.06), with the first instar larvae being slightly larger than the adult for most of the species except for the largest. There was no significant correlation between the size of the adult and the size of the second instar larvae (N ¼ 6; r 2 ¼ 0.16, p . 0.3).
Developmental constraints in cave beetles Alexandra Cieslak, Javier Fresneda and Ignacio Ribera
Electronic supplementary material Table S1. Species with life cycle data, and source of the information. Table S2. Data of the duration and size of the different stages of the life cycle. Table S3. Comparison of the duration of the different developmental stages for the species with detailed life cycle data. Figure S1. Phylogeny of Leptodirini, with the phylogenetic position of the species with life cycle data.
1
Table S1. Species with detailed life cycle data, with the reference to the work from which the data was obtained, the original species name used in it. No Genus
original species name
Species
Description
D
Ph
1 Antrocharis 2 Bathysciella
querilhaci jeanneli
Lespes, 1857 Abeille de Perrin, 1904
P P
1 1
3 Bathysciola
mystica
Fresneda & Fery, 2009
P
1
4 Cytodromus
dapsoides
Abeille de Perrin, 1876
SF
1
5 6 7 8
serullazi colasi serullazi hochenwartii
Peyerimhoff, 1904 Bonadona, 1955 Fagniez, 1914 Schmidt, 1832
SF SF SF B
1 1 -* 1
infernus
Dieck, 1970
P
1
Speonomus infernus
10 Parvospeonomus canyellesi
Lagar, 1974
P
1
Speonomus canyellesi
11 Parvospeonomus delarouzeei
Fairemaire, 1860
P
1
Speonomus delarouzeei
12 13 14 15
Bedel, 1878 Saulcy, 1982 Delarouzée, 1860 Sainte-Claire Deville, 1901
SF P SF SF
1 1 1 1
Diaprysius Isereus Isereus Leptodirus
9 Machaeroscelis
Royerella Speonomus Speophyes Troglodromus
tarissani longicornis lucidulus bucheti
Bathysciola schioedtei Kiesenwetter, 1850
life cycle origin of data data 3,10 data obtained in the Laboratoire Souterrain de Moulis 7 data obtained in the Laboratoire Souterrain de Moulis B. schioedtei has been found to be a complex of monophyletic, closely related species (12). 6,10 According to the distribution of the specimens breed in Moulis they should belong to B. mystica. Data obtained in the Laboratoire Souterrain de Moulis 8,10 data obtained in the Laboratoire Souterrain de Moulis 4 8 8 9,10
Troglodromus bucheti gaveti
15,16 1 14,16,17 9 13 11 5
data obtained in the Laboratoire Souterrain de Moulis data obtained in the Laboratoire Souterrain de Moulis data obtained in the Laboratoire Souterrain de Moulis data obtained in the Laboratoire Souterrain de Moulis data obtained in the Laboratoire Souterrain de Moulis; previously described from material found in caves given as 1 instar, but cycle not completed ** data obtained in the Laboratoire Souterrain de Moulis; previously described from material found in caves data obtained in the Laboratoire Souterrain de Moulis data obtained in the Laboratoire Souterrain de Moulis data obtained in the Laboratoire Souterrain de Moulis data obtained in the Laboratoire Souterrain de Moulis
*The only species not included in the analyses, I. serullazi, is geographically and morphologically similar to I. colasi, and most likely also phylogenetically close ** Although the cycle could not be completed, Parvospeonomus canyellesi was reported by Bellés et al. (1986) as having a 1-instar cycle due to the large size of the first instar larvae, the fact that it was not active and not feeding, and the presence of only one mature egg in the reproductive track of the females. The size of the first instar larvae is, however, within the normal relationship for the group (see main text and Cieslak et al. 2014), and cannot be used to predict the number of instars. There is also at least a known case of a species with a non-feeding 2-instar larva with a macrolecital egg (Dyaprysius serullazi, Table S2). D: Geographical distribution: B: Balkans, P: Pyrenees, SF: Southeast France. Ph: weather the species is included in the phylogeny or not (see figure S1).
2
Table S2. Duration (in days) and size of the different stages of development in the species for which there is available data, including the number of larval instars and the type of egg. The detailed number of specimens used for each of the breeding experiments is generally not reported, although in some cases it was noted that several hundred specimens were used. At present the data of the egg development of the different species is very incomplete and cannot be interpreted unequivocally. See Table S1 for the sources of the data. No Genus
Species
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
querilhaci 1 M 3.1 3.3 jeanneli 2 mM 3.65 mystica 3 m 2 2.4 dapsoides 1 M 3.75 4 serullazi 2 M 2.85 3.3 colasi 1 M serullazi 1 M 3.85 4 hochenwartii 1 M 6.6 6 infernus 2 mM 2.35 2.7 canyellesi 2? M 2 2.5 delarouzeei 2 mM 2.2 2.5 tarissani 1 3.9 longicornis 1 M 3 3.5 lucidulus 2 mM bucheti 1 M 3.5 3.5
Antrocharis Bathysciella Bathysciola Cytodromus Diaprysius Isereus Isereus Leptodirus Machaeroscelis Parvospeonomus Parvospeonomus Royerella Speonomus Speophyes Troglodromus
NoI ET SA S1 S2 S3 SP DE DL DP DT 3.5 4.1 2.7 3.3 2.4 3.7 3.4 3.1
100 124 165 56 210 59 325 134 123 116 192 117 94 117 328 4 117 99 116 332
3.8 2.5 3.5
2.8 63 167 74 304 2.5 51 170 50 3.8 119 118 4.1 66 184 67 60 143 56 3.8 94 101 108
271 317 258 303
NoI: number of instars. ET: Egg Type: M: macrolecithal, m: microlecithal, mM: micro-‐mesolecithal. SA: Average size of the adults (mm). S1-S3: Average size of the larval instars (mm). SP: Average size of the pupae (mm). DE: Duration of the egg development (days). DL: Duration of the larval development (days). DP: Duration of the pupal development (days). DT: Total duration of the development (days).
3
Table S3. Comparison of the duration of the different life stages (see Table S1 for the origin of the life cycle data). A) Species for which there is comprehensive data, with the pertinence to the groups tested (groups 1 to 3). dist., distribution (P, Pyrenees; F, southeastern France); No.I, number of larval instars. Duration of the stages in days. species gr1 gr2 gr3 dist. No.I egg larvae pupae total Cytodromus dapsoides 1 1 1 F 1 134 123 Isereus colasi 1 1 1 F 1 117 94 117 328 Isereus serullazi 1 1 1 F 1 117 99 116 332 Royerella tarissani 1 1 1 F 1 119 118 Troglodromus bucheti 1 1 1 F 1 94 101 108 303 P 1 100 124 Antrocharis querilhaci 2 2 1 P 1 66 184 Speonomus longicornis 2 2 1 67 317 Diaprysius serullazi 2 1 2 F 2 116 192 Speophyes lucidulus 2 1 2 F 2 60 143 56 258 Bathysciella jeanneli 2 2 2 P 2 165 P 3 56 210 Bathysciola mystica 2 2 2 59 325 P 2 63 167 Macharoscelis infernus 2 2 2 74 304 P 2 51 170 Parvospeonomus delarouzeei 2 2 2 50 271
B) ANOVA and PDAP (phylogenetically independent contrasts) results of the comparison of the duration of the different life stages. comparison stage gr1: 1-instar France egg vs ancestral larvae pupae total gr2: France vs egg Pyrenees larvae pupae total gr3: 1-instar vs egg 2&3-instars larvae pupae total
ANOVA n F average 1 12
