Flea community
The flea community included the following species: Ctenophthalmus apertus apertus (n = 2), C. a. gilcolladoi (n = 1879), C. baeticus baeticus (n = 11), Leptopsylla taschenbergi amitina (n = 460), Nosopsyllus fasciatus (n = 1903) and Rhadinopsylla beillardae (n = 8); in addition, two specimens were identified at genus level only (Ctenophthalmus spp.). Two species dominated the small mammal flea community: N. fasciatus and C. a. gilcolladoi (frequency = 44.6% and 44.1%, respectively), followed by L. taschenbergi (10.8%). Patterns of flea infestation differed between small mammal host species (Table 1 and Figure 1; complementary information in the additional Figure S1). The most abundant flea species infesting common voles were N. fasciatus and C. a. gilcolladoi accounted for 98.5% of all fleas identified (48.5% and 50.0%, respectively). By contrast, L. taschenbergi was the most abundant flea infesting mice (56% and 48% of all fleas for A. sylvaticus and M. spretus, respectively). Regarding shrews, we found that individuals were only infested by C. a. gilcolladoi and N. fasciatus. Other Ctenophthalmus spp. different from C. a. gilcolladoi (C. a. apertus, C. baeticus and Ctenophthalmus spp.) were seldom found in M. arvalis (in one, eight and two animals respectively). The flea R. beillardae was occasionally identified in the most abundant rodent species (two fleas in tow M. arvalis, five fleas in two A. sylvaticus, and one flea in one M. spretus).
Overall, flea prevalence on small mammal hosts averaged 51.6% (CI = 49.5-53.7), and mean intensity was 3.66 fleas per infested host (SE = ± 0.15; range = 0-68), resulting a mean abundance of 1.89 (SE = ± 0.09). Detailed information on prevalence, mean abundance, and mean intensity of each flea species and host is provided in Table 1. Fleas were highly aggregated on their small mammal hosts (Table 1; D-Index values close to 1). Variance to mean ratios were also indicative of a marked parasite aggregation, with greater ratios for those fleas typically more abundant on a given host (C. a. gilcolladoi and N. fasciatus for M. arvalis; L. taschenbergi for A. sylvaticus and M. spretus).
Table 1. Parasitological parameters of the flea community on all small mammal hosts studied.
Host species
[n total1/ n alive2]
|
Flea
species
|
n. identified fleas
[n hosts3]
|
Fleas
intensity range
|
Prevalence %
(95% CI) 4
|
Mean
intensity
(±SE)
|
Mean
abundance
(±SE)
|
Variance / Mean ratio
|
Discrepancy index
Mean (95% CI) 5
|
Microtus arvalis
[1380/941]
|
CAG
|
1731 [539]
|
1-18
|
39.1 (36.5-41.7)
|
3.21 (0.14)
|
1.25 (0.09)
|
9.8
|
0.81* (0.79-0.83)
|
NF
|
1681 [643]
|
1-30
|
46.6 (43.9-49.3)
|
2.61 (0.09)
|
1.22 (0.06)
|
4.19
|
0.73* (0.71-0.75)
|
LT
|
34 [29]
|
1-4
|
2.1 (1.4-3.0)
|
1.17 (0.01)
|
0.02 (0.01)
|
1.45
|
0.98* (0.97-0.99)
|
Apodemus sylvaticus
[522/238]
|
CAG
|
116 [75]
|
1-7
|
14.4 (11.5-17.7)
|
1.55 (0.06)
|
0.22 (0.03)
|
2.2
|
0.90 (0.87-0.92)
|
NF
|
182 [116]
|
1-7
|
22.2 (18.7-26.0)
|
1.57 (0.07)
|
0.35 (0.04)
|
1.97
|
0.84 (0.81-0.87)
|
LT
|
387 [140]
|
1-23
|
26.8 (23.1-30.8)
|
2.76 (0.17)
|
0.74 (0.09)
|
5.21
|
0.85 (0.82-0.87)
|
Mus spretus
[304/49]
|
CAG
|
13 [11]
|
1-3
|
3.6 (1.8-6.4)
|
1.18 (0.09)
|
0.04 (0.01)
|
1.42
|
0.97 (0.95-0.98)
|
NF
|
29 [21]
|
1-5
|
6.9 (4.3-10.4)
|
1.38 (0.14)
|
0.10 (0.02)
|
1.88
|
0.94 (0.92-0.97)
|
LT
|
39 [18]
|
1-7
|
5.9 (3.5-9.2)
|
2.17 (0.22)
|
0.13 (0.04)
|
3.24
|
0.96 (0.94-0.97)
|
Crocidura russula
[42/6]
|
CAG
|
7[2]
|
2-5
|
4.8 (0.6-16.2)
|
3.50 (0.83)
|
0.17 (0.13)
|
4.07
|
0.94 (0.86-0.95)
|
NF
|
7 [5]
|
1-3
|
11.9 (4.0-25.6)
|
1.40 (0.40)
|
0.17 (0.08)
|
1.73
|
0.89 (0.79-0.95)
|
Microtus lusitanicus/
duodecimcostatus
[4/3]
|
CAG
|
12 [3]
|
2-6
|
-
|
-
|
-
|
-
|
-
|
NF
|
2 [1]
|
2
|
-
|
-
|
-
|
-
|
-
|
Mustela nivalis
[2/2]
|
NF
|
2 [2]
|
1
|
-
|
-
|
-
|
-
|
-
|
1 Number of total hosts captured. 2 Number of hosts brought alive to the lab and euthanized, infested or uninfested. 3 Number of hosts infested with all fleas identified. 4 95% Confidence Interval by Clopper-Pearson. 5 95% Confidence Interval by bootstrap method. * Sample too big for bootstrap confident limits; the percentile method was used instead.
Abbreviations: SE, standard error; CAG, Ctenophthalmus apertus gilcolladoi; NF, Nosopsyllus fasciatus; LT, Leptopsylla taschenbergi.
Figure 1. Flea abundance frequencies in the main small mammal host species (Rodents and Insectivores).
Abbreviations: CAA, Ctenophthalmus apertus apertus; CAG, Ctenophthalmus apertus gilcolladoi; CB, Ctenophthalmus baeticus; NF, Nosopsyllus fasciatus; LT, Leptopsylla taschenbergi; RB, Rhadinopsylla beillardae; * Sample size too small (n = 1).
Variation of flea parasitological parameters according to season, crop type and host sex
Flea prevalence, mean intensity and mean abundance were highest in M. arvalis (68.2%, CI = 65.7-70.6; 4.14, SE = ± 0.19; and 2.83, SE = ± 0.13, respectively) and A. sylvaticus (45.6%, CI = 413-50.0; 2.93, SE = ± 0.20; and 1.34, SE = ± 0.11, respectively), and noticeably lower in M. spretus (16.1%, CI = 12.2-20.7; 1.78, SE = ± 0.21; and 0.29, SE = ± 0.48, respectively) and C. russula (14.3%, CI = 5.40-28.5; 2.33, SE = ± 0.67; and 0.33, SE = ± 0.15, respectively). The highest number of fleas per host was harboured by M. arvalis [range 1-68], followed by A. sylvaticus [1-29], while M. spretus [1-5] and C. russula [1-7] had fewer fleas per host. For further analyses of infestation patterns, we focused on the main flea species (C. a. gilcolladoi, N. fasciatus and L. taschenbergi) and the most frequently captured small mammal hosts (M. arvalis, A. sylvaticus and M. spretus) (Table 2).
Variation in C. a. gilcolladoi parameters on M. arvalis were significantly explained by month (prevalence: Χ2 = 26.82, P < 0.001; intensity: Χ2 = 46.35, P < 0.001; abundance: Χ2 = 34.53, P < 0.001) and crop type (prevalence: Χ2 = 5.25, P = 0.073*; intensity: Χ2 = 9.85, P = 0.007; abundance: Χ2 = 14.65, P < 0.001), but not by host sex. Post-hoc tests (Tukey) indicated that C. a. gilcolladoi was less frequent and severe on voles in July than March, with intermediate values in November. Moreover, voles caught in alfalfa had lower flea prevalence, abundance and intensity than those from fallows. In A. sylvaticus, C. a. gilcolladoi parameters differed between sexes (prevalence: Χ2 = 6.45, P = 0.011; intensity: Χ2 = 8.10, P = 0.004; abundance: Χ2 = 9.44, P = 0.002), being greater for male than female hosts. The only exception was mean intensity that reached higher values in animals trapped during March than in July. In M. spretus, neither variable explained C. a. gilcolladoi prevalence variation. The small sample size for this species did not allows us to analyse intensity or abundance.
Regarding N. fasciatus, we found in voles the same pattern than in C. a. gilcolladoi, with differences in the three parameters between crop types (prevalence: Χ2 = 24.15, P < 0.001; intensity: Χ2 = 30.45, P < 0.001; abundance: Χ2 = 43.28, P < 0.001) and months (prevalence: Χ2 = 26.82, P < 0.001; intensity: Χ2 = 13.23, P = 0.001; abundance: Χ2 = 34.53, P < 0.001) in the three parameters. Infestation with N. fasciatus is more frequent and severe during July, and lower levels of flea infestation are found in voles from alfalfas. In A. sylvaticus, N. fasciatus abundance did not differ between sexes, but varied between months (prevalence: Χ2 = 18.48, P <0.001; intensity: Χ2 = 5.92, P = 0.052*; abundance: Χ2 = 18.14, P < 0.001), with a higher infestation rate during July and a reduced intensity and abundance during November. The infestation with this flea varied between months in M. spretus although the effect was only marginally significant (prevalence: Χ2 =5.36, P = 0.069*; intensity: NA; abundance: Χ2 = 5.48, P = 0.065*). The highest prevalence rate was found in July but the mean abundance dropped in March. In M. spretus, N. fasciatus was also more abundant on males than females (Χ2 = 2.74, P = 0.0098*). The small sample size for this species did not allow to model intensity variation.
Regarding L. taschenbergi, we found that this flea was more abundant on males than on females in M. arvalis (prevalence: Χ2 = 3.47, P = 0.062*; intensity: Χ2 = 4.56, P = 0.032; abundance: Χ2 = 4.50, P = 0.034). In both mouse species, we found a significant effect of both sex (A. sylvaticus prevalence: Χ2 = 16.68, P < 0.001; intensity: Χ2 = 7.28, P = 0.007; abundance: Χ2 = 16.69, P < 0.001; M. spretus prevalence: Χ2 = 4.30, P = 0.038; intensity: NA; abundance: Χ2 = 4.80, P = 0.029) and month (A. sylvaticus prevalence: Χ2 = 54.53, P < 0.001; intensity: Χ2 = 56.34, P < 0.001; abundance: Χ2 = 97.93, P < 0.001; M. spretus: prevalence: Χ2 = 7.15, P = 0.028; intensity: NA; abundance: Χ2 = 9.01, P = 0.011). The infestation was more prevalent and severe in July and among males than females in both host species. Furthermore, crop type explained variation in intensity and abundance A. sylvaticus (intensity: Χ2 = 7.86, P = 0.020; abundance: Χ2 = 6.15, P < 0.046), with a greater intensity those from cereal compared with fallows.
Table 2. Summary of GLMM best model results for the parasitological parameters on the main rodent host species.
Co-infections
The majority of hosts were infested with one or two flea species (63.2% and 34.5% respectively). Few hosts (2.3%) harboured three flea species (Table 3). Higher co-infection rates were found in M. arvalis (38.8%) and A. sylvaticus (34.6%) than in M. spretus (13.3%) and C. russula (16.7%). Co-infection patterns with two flea species were diverse among hosts. Associations composed of C. a. gilcolladoi-N. fasciatus were most commonly found in M. arvalis (90.3%), with very few co-infections with L. taschenbergi. In fact, this flea was collected alone in almost 90% of the cases. L. taschenbergi-N. fasciatus associations prevailed in the mice hosts (A. sylvaticus = 45.9%; M. spretus = 66.7%), although this predominance in A. sylvaticus was similar to the other co-infections (C. a. gilcolladoi-N. fasciatus: 30.3% and C. a. gilcolladoi-L. taschenbergi: 23.9%; Figure 1). These results were in agreement with the Fager index values obtained for pairs of flea species on these hosts (Table 3).
Co-infection rates did not differ between sexes in M. arvalis (Χ2 = 2.87, P = 0.090) or in M. spretus (G = 2.90, P = 0.886), but did in A. sylvaticus (Χ2 = 5.89, P = 0.015), with fewer co-infections in females than males. Considering hosts infested with two or three species, we found no differences between sexes (M. arvalis: Χ2 = 2.67, P = 0.102; A. sylvaticus: G = 2.63, P = 0.105). In terms of co-infection assemblies, male and female wood mice presented similar values of co-infection for all flea pairs (G = 0.87, P = 0.649). In voles, however, we found a male-biased N. fasciatus-L. taschenbergi co-infection, which was more frequent in males than in females (Χ2 = 9.62, P = 0.008).
Table 3. Flea co-infection rates on the main hosts and Fager index for the commonest flea associations.
Host
|
Flea species per host
prevalence% [n host]
|
Fager Index [n host]
|
1
|
2
|
3
|
CAG-LT
|
CAG-NF
|
LT-NF
|
M. arvalis
|
61.2 [534]
|
37.2 [324]
|
1.6 [14]
|
0.046 [13]
|
0.538 [318]
|
0.063 [21]
|
Male
|
64.1 [270]
|
34.0 [143]
|
1.9 [8]
|
0.060 [8]
|
0.498 [139]
|
0.098 [16]
|
Female
|
58.5 [264]
|
40.1 [181]
|
1.3 [6]
|
0.029 [5]
|
0.574 [179]
|
0.029 [5]
|
A. sylvaticus
|
65.1 [155]
|
28.7 [68]
|
5.9 [14]
|
0.242 [26]
|
0.346 [33]
|
0.391 [50]
|
Male
|
60.2 [97]
|
31.7 [51]
|
8.1 [13]
|
0.273 [22]
|
0.424 [28]
|
0.442 [40]
|
Female
|
76.3 [58]
|
22.4 [17]
|
1.3 [1]
|
0.148 [4]
|
0.169 [5]
|
0.267 [10]
|
M. spretus
|
86.7 [39]
|
13.3 [6]
|
0
|
0.069 [1]
|
0.063 [1]
|
0.205 [4]
|
Male
|
83.3 [30]
|
16.7 [6]
|
0
|
0.083 [1]
|
0.077 [1]
|
0.250 [4]
|
Female
|
100.0 [9]
|
0
|
0
|
0
|
0
|
0
|
C. russula
|
83.3 [5]
|
16.7 [1]
|
0
|
0
|
0.286 [1]
|
0
|
Male
|
75.0 [3]
|
25.0 [1]
|
0
|
0
|
0.400 [1]
|
0
|
Female
|
100.0 [2]
|
0
|
0
|
0
|
0
|
0
|
Abbreviations: 1, one flea species; 2, two co-occurrence flea species; 3, three co-occurrence flea species. For flea species abbreviation (CAG, LT, NF) see figure 1.