Ok, perfect.
- In relation to NDied, I have to calculate the number of died seedlings
as a new column since the model you proposed uses the number of survived
and died seedlings, true?
- Do I have to use the same model to evaluate seed emergence?
- Are you agree that this is a nested experiment design?
Thank you very much,
---
Manuel Esteban Lucas Borja
Universidad de Castilla La Mancha
Escuela T?cnica Superior de Ingenieros Agr?nomos y de Montes
Departamento de Ciencia y Tecnolog?a Agroforestal y Gen?tica
Campus Universitario s/n,
C.P. 02071, Albacete (Spain)
T?lf.; 967599200 ext. 2818
Mail: ManuelEsteban.Lucas at uclm.es
Web:http://www.uclm.es/profesorado/manuelestebanlucas/
El 30/03/2015, a las 15:15, Thierry Onkelinx <thierry.onkelinx at inbo.be>
escribi?:
Yes you need the number of seeds because the distribution is discrete.
E.g. with 4 seeds only 0%, 25%, 50%, 75% or 100% can survive.
I notice a typo in my code. You need to specify the distribution as well.
library(lme4)
model <- glmer(cbind(NSurvived, NDied) ~ Site * Condition * Protection +
(1|Plot), data = your.dataset, family = binomial)
ir. Thierry Onkelinx
Instituut voor natuur- en bosonderzoek / Research Institute for Nature and
Forest
team Biometrie & Kwaliteitszorg / team Biometrics & Quality Assurance
Kliniekstraat 25
1070 Anderlecht
Belgium
To call in the statistician after the experiment is done may be no more
than asking him to perform a post-mortem examination: he may be able to say
what the experiment died of. ~ Sir Ronald Aylmer Fisher
The plural of anecdote is not data. ~ Roger Brinner
The combination of some data and an aching desire for an answer does not
ensure that a reasonable answer can be extracted from a given body of data.
~ John Tukey
2015-03-30 15:09 GMT+02:00 Manuel Esteban Lucas Borja <
ManuelEsteban.Lucas at uclm.es>:
Dear Thierry,
Thank you so much. Please see in attachment the .txt file (sorry about
the xls fiel).
Following your recommendation, I have to use the number of seeds and not
the percentage, true?
Sincerely,
---
Manuel Esteban Lucas Borja
Universidad de Castilla La Mancha
Escuela T?cnica Superior de Ingenieros Agr?nomos y de Montes
Departamento de Ciencia y Tecnolog?a Agroforestal y Gen?tica
Campus Universitario s/n,
C.P. 02071, Albacete (Spain)
T?lf.; 967599200 ext. 2818
Mail: ManuelEsteban.Lucas at uclm.es
Web:http://www.uclm.es/profesorado/manuelestebanlucas/
El 30/03/2015, a las 14:57, Thierry Onkelinx <thierry.onkelinx at inbo.be>
escribi?:
Dear Manuel,
You can solves this with a mixed model. Something like
library(lme4)
model <- glmer(cbind(NSurvived, NDied) ~ Site * Condition * Protection +
(1|Plot), data = your.dataset)
Note that the list only allows several types of attachments. xls files
are striped.
You might want to do some reading (e.g. Zuur et al, 2009) or consult a
local statistician.
Best regards,
ir. Thierry Onkelinx
Instituut voor natuur- en bosonderzoek / Research Institute for Nature
and Forest
team Biometrie & Kwaliteitszorg / team Biometrics & Quality Assurance
Kliniekstraat 25
1070 Anderlecht
Belgium
To call in the statistician after the experiment is done may be no more
than asking him to perform a post-mortem examination: he may be able to say
what the experiment died of. ~ Sir Ronald Aylmer Fisher
The plural of anecdote is not data. ~ Roger Brinner
The combination of some data and an aching desire for an answer does not
ensure that a reasonable answer can be extracted from a given body of data.
~ John Tukey
2015-03-30 14:31 GMT+02:00 Manuel Esteban Lucas Borja <
ManuelEsteban.Lucas at uclm.es>:
Dear all,
I am just starting with R program and I would be very grateful if you
could help me about how to analyze the data. I am involved in a Pinus
nigra regeneration experiment in Cuenca Mountains (Spain). The aim of the
study is to evaluate the effect of fire, soil preparation (scalping) and
control (without treatment) on seed emergence and seedling survival in two
sites (pure and mixed pinus nigra stands). Also seed predation effect will
be evaluated using traps to protect the seeds and seedlings.
Different sowing points (subplots 25x25 cm) were sown using 6 pinus
nigra seeds previously collected in the field. Variable response is seed
emergence rate (%) and seedling survival rate (%) at the end of the
experiment (1 year)
The experimental design is as follow:
- Three factors: Site, condition and seed protection.
- Levels of each factor:
- Two sites (a mixed stand and a pure stand)
- Three conditions (postfire, scalping and control)
- Protection of seeds (yes or not)
Five plots were established at each condition and two subplots were
established at each plot (A and B). Each subplot is considered as sowing
points being one protected against predators and the other not.
2 sites x 3 conditions x 2 seed protection x 5 plots x 2 sowing points=
120 (n=observations)
Please see attached .xls file to explore the data.
As far as I know, the experimental design can be considerer as nested
since the factor ?condition? is nested in the site factor (control level is
not the same for mixed stand or pure stand). All the factors are fixed.
On these context my questions is how to analyze the data? And how to
perform the analysis using the R program? I am thinking to use GML models
but maybe using an ANOVA being condition factor nested in the site factor
is easier.
Thank you so much. Any help is more than welcome.
Manuel
---
Manuel Esteban Lucas Borja
Universidad de Castilla La Mancha
Escuela T?cnica Superior de Ingenieros Agr?nomos y de Montes
Departamento de Ciencia y Tecnolog?a Agroforestal y Gen?tica
Campus Universitario s/n,
C.P. 02071, Albacete (Spain)
T?lf.; 967599200 ext. 2818
Mail: ManuelEsteban.Lucas at uclm.es
Web:http://www.uclm.es/profesorado/manuelestebanlucas/