There were significant group (block) effects on several measurements (Table 6),
and indicated spatial variations in environmental conditions where different groups were
buried. All measurement varied with time, except C:N ratio of soil (Table 6). C:N ratio
of soil was quiet stable with time. Treatment had different effect at different level
(extractant, soil and soil microbes, Table 6, Fig. 13~16). Latrine treatment increased all
the N concentration in extractant, but no effect on extractable C. Only microbial
biomass N increased after latrine added in, but the microbial biomass C and microbial
C:N ratio not changed. Conversely, soil C content was increased with latrine treatment,
but N content and C:N ratio of soil were no changed. After one month (September to
October) of incubation, latrine significantly increased the entire extractable N in upper
horizon, and nitrate and inorganic N in lower horizon soil (Table 6, Fig. 13). Generally,
double amount of fecal pellets (treatment D) had greater effects than single amount
(treatment S) in elevating soil nitrogen, except organic N of upper horizon. After two
months of incubation, extractable inorganic N of both horizons still differed among
treatments, likely caused by the increasing nitrate from nitrification (Fig. 13 C, D, E,
and F). No latrine effect was found in the upper horizon after three months, but the
nitrate in the lower horizon was still higher than control even after four months. Overall,
extractable inorganic N was approximately 1% of soil total nitrogen. Even extractable
TN was just 2~3% of soil total nitrogen. Whereas, microbial biomass N was 6~8 % of
soil total nitrogen content, more then that of extractable TN (Table 5). Microbial
biomass C was also a magnitude higher than extractable carbon (Table 5). Thus, the N
and C in microbial biomass were important in alpine soil. There was a sudden decline in
microbial biomass C in both horizons in January 2009 (the 4th month of incubation), but
not in microbial biomass N (Fig. 15 A, B, C, and D). C:N ratio of microbes remained
relative constant in the first 2 months, yet dropped in the later 2 months (Fig. 16 E and
F).
The fecal pellets on the soil surface lost little weights between the first and second
months. At the end of incubation, fecal pellets maintained > 60% initial weight (Fig. 17).
N content had little increase (F = 6.85, d.f. = 3, p = 0.013, Fig. 18A), but C contents did
not change significantly during incubation (F = 3.68, d.f. = 3,p = 0.06, Fig. 18B).
Fifty-five percent of the initial total nitrogen in fecal pellets remained at the end of
incubation, and nearly all the lost nitrogen was gone within the first month (September
~ October, Fig. 19A). The slight increases in TN at the second and third months were
likely caused by microbial immobilization. Similarly, most of the loss of total carbon
occurred in the first month (Fig. 19B). The sharp decline occurred between December
and January might have been related to the decline of microbial biomass C in soil. The
ash content of fecal pellets was 11.05 %.
After one month of incubation, the nitrogen lost from latrines was 7.65 ± 0.26 mg
(47.6%) and 15.36 ± 0.27 mg (47.8%) for the single and double fecal pellet treatments,
respectively. The increased of extractable TN in upper horizon was 6.94 ± 0.57 mg
(62%) and 15.64 ± 0.45 mg (140%) for the single and double treatments than control,
respectively. No any significantly increasing was found in addition to extractable N. The
total increased amount of N at upper horizon approximates the amount released from
fecal pellets.
Table 5. The proportions of extractable N and C to the total N and C in field incubation on homogenized soil. There were 3 treatments: Control (no vole fecal pellets), Single (added 2.5 g of fresh vole fecal pellets), and Double (D, added 5.0 g of fresh vole fecal pellets).
Treat NH4+ NO3- Inorganic N Extractable organic N Extractable TN Nmic Extractable C Cmic
Sep Initial 1.70 0.03 1.72 1.07 2.80 7.70 0.34 3.03
Oct
Control 0.68 0.05 0.73 0.86 1.59 8.80 0.40 3.54
Single 1.20 0.23 1.43 1.07 2.50 8.43 0.39 3.52
Double 1.90 0.53 2.44 1.20 3.64 7.77 0.36 3.56
Nov
Control 0.49 0.09 0.58 0.82 1.40 7.50 0.41 3.35
Single 0.16 0.66 0.82 0.77 1.59 7.62 0.43 3.17
Double 0.18 0.96 1.14 0.89 2.03 7.14 0.43 3.05
Dec
Control 0.24 0.18 0.42 0.86 1.28 6.79 0.34 3.00
Single 0.07 0.22 0.30 0.84 1.14 6.65 0.36 3.09
Double 0.08 0.41 0.50 0.80 1.30 6.50 0.35 3.19
Jan
Control 0.26 0.14 0.41 0.67 1.07 7.74 0.47 2.17
Single 0.07 0.12 0.19 0.66 0.85 8.23 0.47 2.03
Double 0.09 0.23 0.32 0.68 1.00 7.60 0.50 1.78
Table 6. Results of two-way randomized complete block design ANOVAs of field incubation on homogenized soil. There were 3 treatments: Control (no vole fecal pellets), Single (added 2.5 g of fresh vole fecal pellets), and Double (D, added 5.0 g of fresh vole fecal pellets). Each treatment had 24 replicates, with a total of 72 incubation pipes. The 72 pipes were divided into six groups of 12 pipes, and each group had 4 pipes from a treatment. Effects tested include group, treatment (Txt), and retrieving time
Soil Layer Upper (A horizon) Lower (B horizon)
Effects Group Treatment Time Time × Txt Time × Group Group Treatment Time Time × Txt Time × Group
d.f. =5 d.f. =2 d.f. =3 d.f. =6 d.f. =10 d.f. =5 d.f. =2 d.f. =3 d.f. =6 d.f. =10
F p F p F p F p F p F p F p F p F p F p
Ammonium 0.94 0.496 10.8 0.003 112 <.0001 15 <.0001 22.49 0.868 2.64 0.09 30.17 <.0001 237 <.0001 6.39 0.001 1.29 0.285
Nitrate 111 <.0001 3.77 0.035 30.7 <.0001 7.91 <.0001 1.14 0.379 2.32 0.121 168 <.0001 200 <.0001 24.3 <.0001 2.67 0.017
Inorganic N 3.92 0.031 79.2 <.0001 83.4 <.0001 7.32 0.001 1.03 0.459 4.52 0.021 210 <.0001 60 <.0001 21.5 0.001 1.62 0.146
Extractable
organic N 1.55 0.261 8.4 0.007 18 0.001 2.4 0.076 0.84 0.627 1.31 0.335 8.67 0.007 17.9 <.0001 0.86 0.547 0.41 0.959
Extractable TN 2.56 0.097 58.8 <.0001 103 <.0001 7.6 0.001 0.86 0.609 5.06 0.014 136 <.0001 19.5 <.0001 6.69 0.001 0.81 0.656
Extractable C 0.95 0.489 0.59 0.572 21 0.0004 1.68 0.189 0.59 0.572 1.62 0.242 2.06 0.178 171 <.0001 0.65 0.69 1.45 0.208
Microbial
biomass N 1.96 0.171 8.12 0.008 7.53 0.01 0.28 0.937 1.06 0.438 1.13 0.407 1.27 0.322 14.8 0.001 0.99 0.467 0.91 0.564
Microbial
biomass C 6.95 0.005 0.38 0.693 33.8 <.0001 1.27 0.326 2.57 0.021 9.96 0.001 2.2 0.161 214 <.0001 1.15 0.381 2.43 0.028
C:N of microbes 6.21 0.007 0.35 0.71 28.9 0.0001 1.35 0.293 0.76 0.703 9.61 0.001 0.15 0.867 115 <.0001 0.63 0.706 1.28 0.289
TN 0.4 0.839 2.32 0.149 4.62 0.037 1.18 0.365 0.82 0.652
TC 1.2 0.375 4.67 0.037 9.28 0.006 0.78 0.599 1.25 0.309
C:N 1.99 0.166 0.27 0.772 0.91 0.48 2.02 0.122 1.65 0.138
(A) Upper horizon (B) Lower horizon
Fig. 13. Concentrations (mean±1se, n = 6) of ammonium, nitrate, and overall inorganic N in the upper (A, C, E), and lower (B, D, F) soil horizons. C, control; S, single fecal pellets; D, double fecal pellets.
Different alphabets indicate significant difference between treatments within a given month (Duncan pairwise comparisons).
(A) Upper horizon (B) Lower horizon
Fig. 14. Concentrations (mean±1se, n = 6) of extractable organic N, extractable total N, and extractable total carbon in the upper (A, C, E) and lower (B, D, F) soil horizons. The treatments are C, control; S, single fecal pellets; D, double fecal pellets. Different alphabets indicate significant difference between treatments within a given month (Duncan pairwise comparisons).
(A) Upper horizon (B) Lower horizon
Fig. 15. Concentrations (mean±1se, n = 6) of microbial biomass N and C, and C:N ratio in the upper (A, C, E) and lower (B, D, F) soil horizons. The treatments are C, control; S, single fecal pellets; D, double fecal pellets. Different alphabets indicate significant difference between treatments within a given month (Duncan pairwise comparisons).
(A) (B)
Fig. 16. Concentrations (mean±1se, n = 6) of TN and TC, C:N ratio, and organic material in the upper soil horizons. The treatments are C, control; S, single fecal pellets; D, double fecal pellets. Different alphabets indicate significant difference between treatments within a given month (Duncan pairwise comparisons).
Fig. 17. Percentage (mean±1se, n=6) of initial weight remained in fecal pellets. Treatments were: S, single; D, double
0 September values were the initial values of fecal pellets for both treatments.
(B) (C)
Fig. 19. Percentage (mean±1se, n=6) of (A) nitrogen, and (B) carbon remained in fecal pellets.
Treatments were: S, single; D, double fecal pellets
Laboratory Incubation on Homogenized Soil
In the beginning of incubation (day 0), the concentrations of all nutrients (Fig. 21)
in leachates were not significantly different among treatments (ANOVA, NH4+ (water), F0.05,3,24 = 0.77, p = 0.52; NO3
-(water), F0.05,3,24 = 0.62, p = 0.61; total water soluble N, F0.05,3,24 = 1.04, p = 0.39; water soluble carbon, F0.05,3,24 = 0.50, p = 0.68) indicating that
all microcosms were homogenous. I used the values of control (Fig. 20) as the baseline
values of N and C dynamics of other treatments. Vole feces, time, and their interaction
all showed significant effects on N concentration, including ammonium(water),
nitrate(water), total water soluble N, and water soluble carbon in leachate. In contrary,
litter and feces-by-litter interaction had no significant effect on concentrations of N and
C in leachant (Table 7, Fig. 21~24). The results showed that dynamic patterns of N and
C concentration of leachates from the four treatments could be divided into two groups:
with latrine (F and F+L, referred to as latrine group hereafter) and without latrine (C
and L, referred to as no latrine group hereafter).
0
Fig. 20. N and C concentrations of control in leachate. All concentrations were referred to left y axis, except ammonium(water) was referred to right y axis The error bar represented ± 1se, n = 7.
Table 7. Results of repeated-measure two-way ANOVAs that examined the effects of vole feces and plant litter on N and C concentrations of rainfall leachant collected and CO2 evolution rate during the laboratory incubation experiment.
(A)
Fig. 21. The concentrations in leachates. (A) Ammonium(water). (B) Nitrate(water). (C) Water soluble N. (D) Water soluble C. The treatments were C, control; L, litter; F, feces; F+L, feces and litter. The error bar represented ± 1se, n = 7.
(A)
(B)
(C)
(D)
Concentration (ppm)
Fig. 22. The concentration difference between treatment and control, obtained by subtracting the average value of control from treatment. (A) Ammonium(water). (B) Notrate(water). (C) Water soluble N. (D) water soluble C. The treatments were L, litter; F, feces; F+L, feces and litter. The error bar represented ± se, n = 7.
(A)
Accumulation of total water soluble N
0
Fig. 23. The total amount of N and C accumulated in leachates. (A) Ammonium(water). (B) Nitrate(water). (C) Water soluble N. (D) Water soluble C. The treatments were C, control; L, litter; F, feces; F+L, feces and litter. The error bar represented ± se, n = 7.
(A)
Relative water soluble N leaching rate
-3.0
Relative water soluble C leaching rate
-0.4
Fig. 24. The relative leaching rate of treatment than control. (A) Ammonium(water). (B) Nitrate(water). (C) Water soluble N. (D) Water soluble C. Relative leaching rate obtained by subtracting the accumulation value of control from treatment, and then divided by the sampling days. The treatments were L, litter; F, feces; F+L, feces and litter. The error bar represented ± se, n = 7.
Ammonium(water) concentration was low in leachate, but its response to treatment
was fast, increased almost immediately in latrine group, reached peak on day 10, then
declined with time until day 42 when it dropped to the same level as the control (Fig.
22A). The accumulation of leached ammonium(water) between latrine and no latrine
group started to show differences on day 6, and at the end the latrine group accumulated
nearly 130 % (2.3 mg) more ammonium in leachate than did no latrine group (Fig. 23A).
Relative ammonification rates increased dramatically in the latrine group during day
2~18 (Fig. 24A), indicating that ammonification was vigorous at the early incubation
stage, and slow down afterward. Plant litter had little effect on ammonium concentration,
(Table 7, Fig. 22A and 24A).
Nitrate(water) concentration had a slightly slower response than that of
ammonium(water). The increase in concentration started on day 10 in all treatments (L, F,
& F+L), about the same time when ammonium(water) concentration peaked (Fig. 22A and
B). Nitrate(water) concentration peaked on day 18, and declined with time until day 50
when it reached the same level as the control. Interestingly, nitrate concentration had a
much greater response to the litter treatment than ammonium(water) (Fig. 22A and B). The
accumulation of nitrate(water) in the latrine group became significantly higher than that in
no latrine group on day 22 (Fig. 24B). At the end of incubation, F+L, F, and L
treatments had accumulated 30.8 mg (40%), 27.6 mg (35%), and 5.07 mg (7%) higher
concentration of nitrate in the leachant than control, respectively. Relative nitrification
rates of L and F+L treatments had the same pattern, decreasing between day 2~8 and
increasing between day 8~26. Relative nitrification rate of F treatment increased from
day 0 to 26 (Fig. 26A). Relative nitrification rate of all treatments declined after day 26.
Nitrate(water) made up more than 95% of total soluble nitrogen. Consequently, the
response of total soluble nitrogen was very similar to that of nitrate(water) (Fig. 21C, 22C,
23C & 24C). At the end of incubation, the accumulation of total soluble nitrogen in F+L,
F, and L treatments had accumulated 33.91 mg (40%), 29.75 mg (35%), and 3.93 mg
(5%) higher concentration of total soluble nitrogen than control, respectively (Fig. 23C).
Water soluble carbon responded immediately to F+L treatment, while the response
to F treatment did not become clear until day 6. The responses to both treatments
peaked on day 8 and remained relatively stable afterward. The response to L treatment
did not show difference from the control throughout the incubation (Fig. 21D and 22D).
From start till the end, the concentration of water soluble carbon of the latrine group
was higher than no latrine group, indicating a constantly high decomposition rate of the
latrine group. The accumulation of water soluble organic carbon of latrine group
became significantly higher than no latrine group on day 12 (Fig. 23D). Latrine group
leached more than 20% (14 mg) carbon than no latrine group at the end. Relative
leaching rate of F and L treatments had similar patterns, both were lower than control
before day 4. Afterward, relative leaching rates of F treatment became higher than
control, but those of L treatment remained the same as control. Only the F+L treatment
had a greater leaching rate than control from start to the end of incubation (Fig. 24D).
CO2 evolution rates were not different among 4 treatments before incubation
started (Fig. 25, ANOVA, F = 0.01, p = 0.998). The rates increased dramatically in
latrine group to 2~3 folds of those of no latrine group. The rates of no latrine group
remained at approximately 4 mg/min throughout the incubation. Rates of the F+L
treatment declined sooner (on day 10), and remained lower than F until day 34. Rates of
the F treatment declined on day 20, and joined F+L treatment on day 34. CO2 evolution
rates of latrine group remained 2-fold higher than no latrine group after day 34,
indicating that microbial activities were still high toward the end of incubations (Fig.
25).
(A)
Fig. 25. (A) CO2 evolution rates. (B) Difference between each treatment and control. The treatments were C, control; L, litter; F, feces; F+L, feces and litter. All treatments were at a steady state after 36 days. The error bar represented ± 1se, n = 7.
The initial chemical constituents of soil were shown in Table 8. At the end of 62
days incubation, there was no difference among treatments in nitrate, extractable TN,
and N & C content (Fig. 26 & 27). Feces had effects on upper layer soil in ammonium,
inorganic N, and extractable C (Table 9). The effects on lower layer were ammonium,
extractable organic N, and microbial biomass C & N. But the concentrations of
ammonium and inorganic N were lower in latrine than no latrine group (Fig. 26A & C).
Table 8. The initial chemical constituents (mg/g) of soil used in laboratory incubation. (Mean±1 se, n=4).
13.8± 3.15 138± 4.18 151± 6.12 25.4± 2.52 177± 7.73 267± 20.9
Microbial biomass N
Microbial
biomass C C:N of microbes TN (mg/g) TC (mg/g) C:N
Table 9. Results of non-parametric factorial test that examined the effects of vole feces and plant litter on the N and C concentration of soil, fecal pellets, and leaf litter after 62 days of laboratory
Ammonium Nitrate Inorganic N Extractable
organic N Extractable TN Extractable C
260± 4.52 993± 68.9 3.84± 0.29 6.18± 0.03 90.0± 0.28 14.5± 0.05
(A) (B)
Fig. 26. Concentration (mean±1se, n=3) of extractable C and N in soil at the end of incubation. Different alphabets indicate significant difference among treatments within a given soil layer (non-parametric factorial test (χ2). No difference in those figures without alphabets
(A) (B)
Fig. 27. C and N of soil microbes and soil. (A) Microbial biomass nitrogen. (B) Nitrogen content of soil.
(C) Microbial biomass carbon. (D) Carbon content of soil. (E) C:N ratio of microbial biomass.
(F) C:N ratio of soil. Values were mean±1se, n = 3. Different alphabets indicate significant difference among treatments within a given soil layer (non-parametric factorial test (χ2)
The weights of fecal pellets remained at the end were not different between F and
F+L treatments (Mann-Whitney U Test, U = 5.5, n = 3, p = 0.64), both 77% of initial
weights (Fig 28A). Leaf litters of L and F+L remained at the end (89.56 ± 0.53 % and
84.11 ± 1.02 %, respectively) were significantly different (Mann-Whitney U Test, U =
9.0, n = 3, p = 0.046). Leaf litter decomposed faster when vole fecal pellets were present.
At the end of incubation, N and C contents of fecal pellets were still higher than those
of leaf litter (Table 10, Fig. 28B & C). Fecal pellets still had 56% nitrogen remained,
and the amount reduced was not different between F and F+L treatments
(Mann-Whitney U Test, U = 6.50, n = 3, p = 0.38; loss 30.21 ± 0.93 mg N and 33.10 ±
2.59 mg N, respectively). Leaf litter still had 70% nitrogen remained, only 0.30 ±
0.01mg and 0.24 ± 0.03 mg were reduced in L and F+L treatments, respectively, no
significant between L and F+L treatment (Mann-Whitney U Test, U = 2.00, n = 3, p =
0.468). The amount and the percentage of nitrogen reduced were much higher for fecal
pellets than leaf litter (Table 10). On the other hand, fecal pellets still had 77% carbon
remained, and the amount reduced was not different between F and F+L treatments
(Mann-Whitney U Test, U = 5.00, n = 3, p = 0.82; loss 616.4 mg and 625.6 mg,
respectively). Whereas, leaf litter lost more carbon in F+L than L treatment
(Mann-Whitney U Test, U = 9.00, n = 3, p = 0.046).
Table 10. Weight, N, and C content of fecal pellets and leaf litter before and after incubation. Values are mean±1se. n=3. p value were calculated from Mann-Whitney U Test. The treatments were L, litter; F, feces; F+L, feces and litter.
Organic matter Fecal pellets Leaf litter
Treatment Initial F F+L Initial L F+L remaining. (B) Carbon content. (C) Nitrogen content. (D) C:N ratio. F / L, F was feces treatment for fecal pellets; L was litter treatment for leaf litter. The error bar represented ± 1se, n = 3.
Different alphabets indicate significant difference between treatments (Kruskal-Wallis test and post hoc comparisons by Dunn test)
Discussions
Taiwan vole population survey
Taiwan vole populations showed substantial spatial and temporal heterogeneity at
the study site from October 2005 to May 2009 (Fig. 3). There wasn’t any clear spatial or
temporal pattern. For example, population size at D plot sitting at the bottom of the
slope was consistently high though fluctuated between 10 to 35 individuals. Population
size at H plot sitting on the top of the slope also fluctuated greatly from 0 to 25
individuals. Sizes of both populations peaked in late 2007 during the three and half year
survey. Comparing with cyclic vole populations in North America, Europe, or Japan, the
magnitudes of fluctuations of Taiwan vole populations were relatively small. Wu (2007)
reported Taiwan vole population densities at a nearby alpine meadow at 10.3~12.5 voles
per hectare over 2 years. Wu’s sampling area (4 hectare) was larger than mine, and he
used a 20 m spacing grid as well as a different type of trap (Sherman single-capture live
traps), those factors may have contributed to the lower density estimates than mine in
his vole population.