Caenorhabditis nematodes colonize ephemeral resource patches in neotropical forests

Abstract

Factors shaping the distribution and abundance of species include life-history traits, population structure, and stochastic colonization-extinction dynamics. Field studies of model species groups help reveal the roles of these factors. Species of Caenorhabditis nematodes are highly divergent at the sequence level but exhibit highly conserved morphology, and many of these species live in sympatry on microbe-rich patches of rotten material. Here, we use field experiments and large-scale opportunistic collections to investigate species composition, abundance, and colonization efficiency of Caenorhabditis species in two of the world's best-studied lowland tropical field sites: Barro Colorado Island in Panamá and La Selva in Sarapiquí, Costa Rica. We observed seven species of Caenorhabditis, four of them known only from these collections. We formally describe two species and place them within the Caenorhabditis phylogeny. While these localities contain species from many parts of the phylogeny, both localities were dominated by globally distributed androdiecious species. We found that Caenorhabditis individuals were able to colonize baits accessible only through phoresy and preferentially colonized baits that were in direct contact with the ground. We estimate the number of colonization events per patch to be low.

Keywords: Caenorhabditis; dispersal; nematode; population biology; species description.

FIGURE 1

Collection sites for Caenorhabditis species used in this study. Caenorhabditis were collected at two localities: Barro Colorado Island, Panamá, and La Selva, Sarapiquí, Costa Rica. (a–c) Distribution of species collected from opportunistic sampling from each locality by year. Each marker represents a patch positive for that species. Patches may be plotted multiple times if species co‐occurred on the same patch. Patches are jittered to prevent overpotting. (d) A field of rotting Spondias mombin substrates (e,f). Rarefaction curve of the chao2 incidence‐based estimator for both localities. The solid line represents the predicted species richness the dotted line represents an extrapolation of species richness. The gray area is the 95% confidence interval.

FIGURE 2

Phylogeny of 36 Caenorhabditis species, with D. coronatus and D. pachys forming an outgroup, based on 1931 single‐copy orthologs each shared between 80% of the species. (a) Phylogeny inferred using a coalescent approach that takes gene trees as input (substitution models for each gene tree selected automatically). Branch lengths in substitutions per site were estimated using the LG substitution model with gamma‐distributed rate variation among sites (LG + Γ) while fixing the phylogeny to the coalescent tree topology. Species incorporated into the phylogeny for the first time are bolded. Posterior probabilities are 1.0 unless noted. (b) Alternative topology using a supermatrix approach that uses concatenated alignments of all orthologs as input under an LG + Γ model. Bootstrap support is 100 unless noted.

FIGURE 3

Species are patchily distributed among rotting Gustavia superba flowers. (a) 10 × 10 meter plots were systematically sampled at each of four focal trees. At each plot, four flowers were collected from two or three 1‐meter quadrats. Each box represents a flower; each color represents the species present on that flower (b). The distribution of C. briggsae colonization events per flower under a simple Poisson model (mean = 1.08).

FIGURE 4

Colonization rates vary in response to bait composition and accessibility. (a) Table describing the six types of baits used in the experiment, the observations for each of the baits, and the counts of Caenorhabditis‐positive baits. (b) Baits were set up at each of seven sites across BCI. Each site consisted of 30 baits arranged in groups of six in the corners and center of each site. (c) The six types of agar bait showed different rates of colonization by nematodes. The blue line is linear regression of Caenorhabditis on non‐Caenorhabditis colonization rates across bait types.

FIGURE 5

Nematodes colonized 30 baits across six experimental plots, each containing a randomized grid of 4 replicates of each of 6 types of bait differing only in accessibility (143 baits all together with one lost). Accessibility ranged from no barrier to being accessible via 0.01 mm pores. Colonization varied significantly by bait accessibility. C. tropicalis and C. briggsae both colonized baits isolated from the environment and accessible only by phoresy while O. tipulae was only found to colonize baits making direct contact with the ground.

FIGURE A1

DIC light micrographs of C. krikudae n. sp. strain QG3050 female (a,b,d–f) and male (c,g–j) anatomy. (a) Whole adult showing two gonad arms. (b) Left side view of head with cephalic sensilla in focus. (c) alae, (d) vulva, (e) spermatheca, (f) female tail, arrow points to phasmid. (g) Spicule and gubernaculum. (h) Stoma, arrow shows the border between gymnostom and stegostom. Male tail in (i) lateral and (j) dorsal view. Scale bar in (a) is 100 μm; b, g, and h 5 μm; c, d, i, and j 10 μm; e and f 15 μm.

FIGURE A2

Comparison of stoma, male tail and spicules in species related to C. krikudae n. sp. Stoma and male tail morphology differ profoundly in these species. The phylogenetic relationships shown on the bottom are based on this study and Dayi et al. (2021). The position of C. sonorae is based on SSU rRNA sequence data only.

FIGURE A3

DIC light micrographs of C. agridulce n. sp. strain QG555 female (a–e) and male (f–j) anatomy. (a) Vulval region in right side view, showing one embryo in the uterus and one older embryo next to the female. An oocyte is passing through the posterior spermatheca. (b) Pharynx in left side view. (c) tail in right side view. (d) Stoma viewed from a sub‐dorsal right angle; arrow points to cross‐section of the flap on the lip, arrowhead to border between gymnostom and stegostom. (e) Stoma in right side view, arrow points to the dorsal metastegostom; one subventral denticle is also in focus. (f) Male stoma in right side view. (g) Male tail in ventral view; the dorsally directed GP7 on the right side and most GPs on the left side are out of focus. (h) Male tail in right side view. (i) spicules and gubernaculum in situ showing (left side) the precloacal lip (arrow) and its horseshoe‐shaped bulge in cross‐section. (j) Spicules and gubernaculum pushed out of the animal and flattened. Arrow points to the characteristic paw‐shaped spicule tip. ph = phasmid. Scale bars in a, b, c, g, and h 20 μm, in all other images 10 μm.

FIGURE A4

Spicules (drawings) and stomata of species in the Angaria group in comparison, and phylogenetic relationships based on this study and an analysis of partial sequences of 15 protein‐coding genes and rRNA genes (Karin Kiontke, unpublished analysis). The spicule tip is more similar in the respective sister species: it is slightly enlarged in C. angaria and C. castelli, narrow in C. dolens and C. quiockensis, and paw‐shaped in C. agridulce n. sp. and C. sp. 8. The stoma is relatively short in all species, but only in C. angaria and C. castelli is the stegostom unusually short. In all other species, the stegostom and gymnostom contribute at least equally to the stomatal tube; the gymnostom is shorter than the stegostom in C. sp. 8. All species display flaps at the lips. The metastegostom carries a projection with a bifid tip (compare Baldwin et al., 1997) in all species. In the light microscope, this structure appears shorter in C. angaria and C. castelli than in the other species.