Here's something I coded that should do the job :)

#!/usr/bin/env perl
# Perl script to retrieve the 3'-UTR sequence for 12 eutherian mammals
# Coded by Steve Moss (gawbul [at] gmail [dot] com)
# http://stevemoss.ath.cx/

# make things easier
use strict;
use warnings;

#import EnsEMBL and BioPerl modules
use Bio::EnsEMBL::Registry;
use Bio::SeqIO;
use Data::Dumper;

# setup array of all epo mammals
my @epo_mammals = ("homo_sapiens", "gorilla_gorilla", "pan_troglodytes", "pongo_abelii", "macaca_mulatta", "callithrix_jacchus", "mus_musculus", "rattus_norvegicus", "equus_caballus", "canis_familiaris", "sus_scrofa", "bos_taurus");

# connect to EnsEMBL
my $registry = 'Bio::EnsEMBL::Registry';
$registry->load_registry_from_db(    -host => 'ensembldb.ensembl.org',
                                    -user => 'anonymous');

# use Bio::SeqIO for write sequence to STDOUT
my $outseq = Bio::SeqIO->new(    -fh => \*STDOUT,
                            -format => 'FASTA');

# loop over the mammals
foreach my $mammal (@epo_mammals) {
    # get gene adaptor
    my $gene_adaptor = $registry->get_adaptor($mammal, 'Core', 'Gene');

    # fetch all genes
    my @gene_ids = @{$gene_adaptor->list_stable_ids()};

    # traverse through genes
    foreach my $gene_id (@gene_ids) {
        # get gene
        my $gene = $gene_adaptor->fetch_by_stable_id($gene_id);

        # get canonical transcript
        my $transcript = $gene->canonical_transcript();

        # check gene has a transcript associated
        unless (defined $transcript) {
            next;
        }

        # get 3'-UTR - returns a Bio::Seq object
        my $tputr = $transcript->three_prime_utr();

        # check transcript has a 3'-UTR annotated
        unless (defined $tputr) {
            next;
        }

        # print to STDOUT
        print $outseq->write_seq($tputr) . "\n";
    }
}

This currently prints the sequence in FASTA format to STDOUT, but you could replace -fh => \STDOUT,* with -file => ">filename", to write to a file instead.

Also, the default Bio::Seq object returned by the transcript object's three_prime_utr method, contains only the sequence and the associated transcript stable_id as the display_id. I would probably create a custom Seq object with gene stable_id, transcript stable_id, and the start and end coordinates of the 3'-UTR. As the three_prime_utr method only returns the Bio::Seq object, it is necessary to determine the start and end of the 3'-UTR based on the start of the transcript and the start of the coding region respectively (or the end of both, if it is on the alternate strand).

Did you look at the Ensembl API Tutorial? There is an example just about retrieving 3' UTRs: http://www.ensembl.org/info/docs/api/core/core_tutorial.html#genes

You can just copy & paste the example in the tutorial and change the name of the species. Note that the example fetches all the UTR in the genes in a chromosomal region: if you want to specify a single gene, you can just use a gene adaptor instead of a slice adaptor.

If you have any specific problem in using the APIs, you can get better answers by editing your question and specifying what is causing you trouble.

I have a script in R to fetch 3UTR via biomaRt: http://ygc.name/2011/03/02/easiest-utr-sequence/